diff --git a/.rustfmt.toml b/.rustfmt.toml
new file mode 100644
index 000000000..250124b77
--- /dev/null
+++ b/.rustfmt.toml
@@ -0,0 +1,5 @@
+unstable_features = true
+
+use_small_heuristics = "max"
+imports_granularity = "Module"
+group_imports = "StdExternalCrate"
diff --git a/README.md b/README.md
index 13d35380a..b1498d0f5 100644
--- a/README.md
+++ b/README.md
@@ -41,3 +41,18 @@ the `content-type:application/json` and `content-type:application/x-ndjson` head
 ### Querying the engine via the website
 
 You can query the engine by going to [the HTML page itself](http://127.0.0.1:9700).
+
+
+## Contributing
+
+You can setup a `git-hook` to stop you from making a commit too fast. It'll stop you if:
+- Any of the workspaces does not build
+- Your code is not well-formatted
+
+These two things are also checked in the CI, so ignoring the hook won't help you merge your code.
+But if you need to, you can still add `--no-verify` when creating your commit to ignore the hook.
+
+To enable the hook, run the following command from the root of the project:
+```
+cp script/pre-commit .git/hooks/pre-commit
+```
diff --git a/benchmarks/benches/songs.rs b/benchmarks/benches/songs.rs
index e5da16a99..726190f77 100644
--- a/benchmarks/benches/songs.rs
+++ b/benchmarks/benches/songs.rs
@@ -6,33 +6,24 @@ use milli::update::Settings;
 use utils::Conf;
 
 fn base_conf(builder: &mut Settings) {
-    let displayed_fields = [
-        "id", "title", "album", "artist", "genre", "country", "released", "duration",
-    ]
-    .iter()
-    .map(|s| s.to_string())
-    .collect();
+    let displayed_fields =
+        ["id", "title", "album", "artist", "genre", "country", "released", "duration"]
+            .iter()
+            .map(|s| s.to_string())
+            .collect();
     builder.set_displayed_fields(displayed_fields);
 
-    let searchable_fields = ["title", "album", "artist"]
+    let searchable_fields = ["title", "album", "artist"].iter().map(|s| s.to_string()).collect();
+    builder.set_searchable_fields(searchable_fields);
+
+    let faceted_fields = ["released-timestamp", "duration-float", "genre", "country", "artist"]
         .iter()
         .map(|s| s.to_string())
         .collect();
-    builder.set_searchable_fields(searchable_fields);
-
-    let faceted_fields = [
-        "released-timestamp",
-        "duration-float",
-        "genre",
-        "country",
-        "artist",
-    ]
-    .iter()
-    .map(|s| s.to_string())
-    .collect();
     builder.set_filterable_fields(faceted_fields);
 }
 
+#[rustfmt::skip]
 const BASE_CONF: Conf = Conf {
     dataset: datasets_paths::SMOL_SONGS,
     queries: &[
@@ -53,34 +44,25 @@ const BASE_CONF: Conf = Conf {
 };
 
 fn bench_songs(c: &mut criterion::Criterion) {
-    let default_criterion: Vec<String> = milli::default_criteria()
-        .iter()
-        .map(|criteria| criteria.to_string())
-        .collect();
+    let default_criterion: Vec<String> =
+        milli::default_criteria().iter().map(|criteria| criteria.to_string()).collect();
     let default_criterion = default_criterion.iter().map(|s| s.as_str());
-    let asc_default: Vec<&str> = std::iter::once("asc(released-timestamp)")
-        .chain(default_criterion.clone())
-        .collect();
-    let desc_default: Vec<&str> = std::iter::once("desc(released-timestamp)")
-        .chain(default_criterion.clone())
-        .collect();
+    let asc_default: Vec<&str> =
+        std::iter::once("asc(released-timestamp)").chain(default_criterion.clone()).collect();
+    let desc_default: Vec<&str> =
+        std::iter::once("desc(released-timestamp)").chain(default_criterion.clone()).collect();
 
     let basic_with_quote: Vec<String> = BASE_CONF
         .queries
         .iter()
         .map(|s| {
-            s.trim()
-                .split(' ')
-                .map(|s| format!(r#""{}""#, s))
-                .collect::<Vec<String>>()
-                .join(" ")
+            s.trim().split(' ').map(|s| format!(r#""{}""#, s)).collect::<Vec<String>>().join(" ")
         })
         .collect();
-    let basic_with_quote: &[&str] = &basic_with_quote
-        .iter()
-        .map(|s| s.as_str())
-        .collect::<Vec<&str>>();
+    let basic_with_quote: &[&str] =
+        &basic_with_quote.iter().map(|s| s.as_str()).collect::<Vec<&str>>();
 
+    #[rustfmt::skip]
     let confs = &[
         /* first we bench each criterion alone */
         utils::Conf {
diff --git a/benchmarks/benches/utils.rs b/benchmarks/benches/utils.rs
index d5181849f..fd1df0a90 100644
--- a/benchmarks/benches/utils.rs
+++ b/benchmarks/benches/utils.rs
@@ -3,10 +3,8 @@ use std::path::Path;
 
 use criterion::BenchmarkId;
 use heed::EnvOpenOptions;
-use milli::{
-    update::{IndexDocumentsMethod, Settings, UpdateBuilder, UpdateFormat},
-    FilterCondition, Index,
-};
+use milli::update::{IndexDocumentsMethod, Settings, UpdateBuilder, UpdateFormat};
+use milli::{FilterCondition, Index};
 
 pub struct Conf<'a> {
     /// where we are going to create our database.mmdb directory
diff --git a/benchmarks/benches/wiki.rs b/benchmarks/benches/wiki.rs
index 11ffe87d5..3d8b6f1d4 100644
--- a/benchmarks/benches/wiki.rs
+++ b/benchmarks/benches/wiki.rs
@@ -6,16 +6,14 @@ use milli::update::Settings;
 use utils::Conf;
 
 fn base_conf(builder: &mut Settings) {
-    let displayed_fields = ["title", "body", "url"]
-        .iter()
-        .map(|s| s.to_string())
-        .collect();
+    let displayed_fields = ["title", "body", "url"].iter().map(|s| s.to_string()).collect();
     builder.set_displayed_fields(displayed_fields);
 
     let searchable_fields = ["title", "body"].iter().map(|s| s.to_string()).collect();
     builder.set_searchable_fields(searchable_fields);
 }
 
+#[rustfmt::skip]
 const BASE_CONF: Conf = Conf {
     dataset: datasets_paths::SMOL_WIKI_ARTICLES,
     queries: &[
@@ -37,18 +35,13 @@ fn bench_songs(c: &mut criterion::Criterion) {
         .queries
         .iter()
         .map(|s| {
-            s.trim()
-                .split(' ')
-                .map(|s| format!(r#""{}""#, s))
-                .collect::<Vec<String>>()
-                .join(" ")
+            s.trim().split(' ').map(|s| format!(r#""{}""#, s)).collect::<Vec<String>>().join(" ")
         })
         .collect();
-    let basic_with_quote: &[&str] = &basic_with_quote
-        .iter()
-        .map(|s| s.as_str())
-        .collect::<Vec<&str>>();
+    let basic_with_quote: &[&str] =
+        &basic_with_quote.iter().map(|s| s.as_str()).collect::<Vec<&str>>();
 
+    #[rustfmt::skip]
     let confs = &[
         /* first we bench each criterion alone */
         utils::Conf {
diff --git a/benchmarks/build.rs b/benchmarks/build.rs
index 58300bab9..b1edd5499 100644
--- a/benchmarks/build.rs
+++ b/benchmarks/build.rs
@@ -1,9 +1,7 @@
+use std::fs::File;
+use std::io::{Cursor, Read, Seek, Write};
 use std::path::{Path, PathBuf};
 use std::{env, fs};
-use std::{
-    fs::File,
-    io::{Cursor, Read, Seek, Write},
-};
 
 use bytes::Bytes;
 use convert_case::{Case, Casing};
@@ -45,7 +43,10 @@ fn main() -> anyhow::Result<()> {
         )?;
 
         if out_file.exists() {
-            eprintln!("The dataset {} already exists on the file system and will not be downloaded again", dataset);
+            eprintln!(
+                "The dataset {} already exists on the file system and will not be downloaded again",
+                dataset
+            );
             continue;
         }
         let url = format!("{}/{}.csv.gz", BASE_URL, dataset);
@@ -60,12 +61,8 @@ fn main() -> anyhow::Result<()> {
 }
 
 fn download_dataset<U: IntoUrl>(url: U) -> anyhow::Result<Cursor<Bytes>> {
-    let bytes = reqwest::blocking::Client::builder()
-        .timeout(None)
-        .build()?
-        .get(url)
-        .send()?
-        .bytes()?;
+    let bytes =
+        reqwest::blocking::Client::builder().timeout(None).build()?.get(url).send()?.bytes()?;
     Ok(Cursor::new(bytes))
 }
 
diff --git a/helpers/src/main.rs b/helpers/src/main.rs
index c916d0448..b325aef89 100644
--- a/helpers/src/main.rs
+++ b/helpers/src/main.rs
@@ -1,9 +1,8 @@
 use std::path::PathBuf;
 
 use byte_unit::Byte;
-use heed::{Env, EnvOpenOptions, CompactionOption};
+use heed::{CompactionOption, Env, EnvOpenOptions};
 use structopt::StructOpt;
-
 use Command::*;
 
 #[cfg(target_os = "linux")]
@@ -65,7 +64,7 @@ fn main() -> anyhow::Result<()> {
             use CompactionOption::*;
             let compaction = if enable_compaction { Enabled } else { Disabled };
             copy_main_database_to_stdout(env, compaction)
-        },
+        }
     }
 }
 
diff --git a/http-ui/src/main.rs b/http-ui/src/main.rs
index e23dddd4c..703861058 100644
--- a/http-ui/src/main.rs
+++ b/http-ui/src/main.rs
@@ -1,6 +1,5 @@
 mod update_store;
 
-use std::{io, mem};
 use std::collections::{BTreeMap, BTreeSet, HashMap, HashSet};
 use std::fmt::Display;
 use std::fs::{create_dir_all, File};
@@ -10,16 +9,19 @@ use std::path::PathBuf;
 use std::str::FromStr;
 use std::sync::Arc;
 use std::time::Instant;
+use std::{io, mem};
 
 use askama_warp::Template;
 use byte_unit::Byte;
 use either::Either;
 use flate2::read::GzDecoder;
-use futures::{FutureExt, StreamExt};
-use futures::stream;
+use futures::{stream, FutureExt, StreamExt};
 use grenad::CompressionType;
 use heed::EnvOpenOptions;
 use meilisearch_tokenizer::{Analyzer, AnalyzerConfig};
+use milli::update::UpdateIndexingStep::*;
+use milli::update::{IndexDocumentsMethod, Setting, UpdateBuilder, UpdateFormat};
+use milli::{obkv_to_json, FilterCondition, Index, MatchingWords, SearchResult};
 use once_cell::sync::OnceCell;
 use rayon::ThreadPool;
 use serde::{Deserialize, Serialize};
@@ -28,12 +30,9 @@ use structopt::StructOpt;
 use tokio::fs::File as TFile;
 use tokio::io::AsyncWriteExt;
 use tokio::sync::broadcast;
-use warp::{Filter, http::Response};
 use warp::filters::ws::Message;
-
-use milli::{FilterCondition, Index, MatchingWords, obkv_to_json, SearchResult};
-use milli::update::{IndexDocumentsMethod, Setting, UpdateBuilder, UpdateFormat};
-use milli::update::UpdateIndexingStep::*;
+use warp::http::Response;
+use warp::Filter;
 
 use self::update_store::UpdateStore;
 
@@ -149,25 +148,28 @@ impl<'a, A: AsRef<[u8]>> Highlighter<'a, A> {
                 for (word, token) in analyzed.reconstruct() {
                     if token.is_word() {
                         let to_highlight = matching_words.matching_bytes(token.text()).is_some();
-                        if to_highlight { string.push_str("<mark>") }
+                        if to_highlight {
+                            string.push_str("<mark>")
+                        }
                         string.push_str(word);
-                        if to_highlight { string.push_str("</mark>") }
+                        if to_highlight {
+                            string.push_str("</mark>")
+                        }
                     } else {
                         string.push_str(word);
                     }
                 }
                 Value::String(string)
             }
-            Value::Array(values) => {
-                Value::Array(values.into_iter()
-                    .map(|v| self.highlight_value(v, matching_words))
-                    .collect())
-            }
-            Value::Object(object) => {
-                Value::Object(object.into_iter()
+            Value::Array(values) => Value::Array(
+                values.into_iter().map(|v| self.highlight_value(v, matching_words)).collect(),
+            ),
+            Value::Object(object) => Value::Object(
+                object
+                    .into_iter()
                     .map(|(k, v)| (k, self.highlight_value(v, matching_words)))
-                    .collect())
-            }
+                    .collect(),
+            ),
         }
     }
 
@@ -236,12 +238,7 @@ enum UpdateMeta {
 #[derive(Debug, Clone, Serialize, Deserialize)]
 #[serde(tag = "type")]
 enum UpdateMetaProgress {
-    DocumentsAddition {
-        step: usize,
-        total_steps: usize,
-        current: usize,
-        total: Option<usize>,
-    },
+    DocumentsAddition { step: usize, total_steps: usize, current: usize, total: Option<usize> },
 }
 
 #[derive(Debug, Clone, Serialize, Deserialize, PartialEq)]
@@ -342,157 +339,185 @@ async fn main() -> anyhow::Result<()> {
             update_builder.max_memory(indexer_opt_cloned.max_memory.get_bytes() as usize);
             update_builder.linked_hash_map_size(indexer_opt_cloned.linked_hash_map_size);
             update_builder.chunk_compression_type(indexer_opt_cloned.chunk_compression_type);
-            update_builder.chunk_fusing_shrink_size(indexer_opt_cloned.chunk_fusing_shrink_size.get_bytes());
+            update_builder
+                .chunk_fusing_shrink_size(indexer_opt_cloned.chunk_fusing_shrink_size.get_bytes());
 
             let before_update = Instant::now();
             // we extract the update type and execute the update itself.
-            let result: anyhow::Result<()> = match meta {
-                UpdateMeta::DocumentsAddition { method, format, encoding } => {
-                    // We must use the write transaction of the update here.
-                    let mut wtxn = index_cloned.write_txn()?;
-                    let mut builder = update_builder.index_documents(&mut wtxn, &index_cloned);
+            let result: anyhow::Result<()> =
+                match meta {
+                    UpdateMeta::DocumentsAddition { method, format, encoding } => {
+                        // We must use the write transaction of the update here.
+                        let mut wtxn = index_cloned.write_txn()?;
+                        let mut builder = update_builder.index_documents(&mut wtxn, &index_cloned);
 
-                    match format.as_str() {
-                        "csv" => builder.update_format(UpdateFormat::Csv),
-                        "json" => builder.update_format(UpdateFormat::Json),
-                        "json-stream" => builder.update_format(UpdateFormat::JsonStream),
-                        otherwise => panic!("invalid update format {:?}", otherwise),
-                    };
-
-                    match method.as_str() {
-                        "replace" => builder.index_documents_method(IndexDocumentsMethod::ReplaceDocuments),
-                        "update" => builder.index_documents_method(IndexDocumentsMethod::UpdateDocuments),
-                        otherwise => panic!("invalid indexing method {:?}", otherwise),
-                    };
-
-                    let reader = match encoding.as_deref() {
-                        Some("gzip") => Box::new(GzDecoder::new(content)),
-                        None => Box::new(content) as Box<dyn io::Read>,
-                        otherwise => panic!("invalid encoding format {:?}", otherwise),
-                    };
-
-                    let result = builder.execute(reader, |indexing_step, update_id| {
-                        let (current, total) = match indexing_step {
-                            TransformFromUserIntoGenericFormat { documents_seen } => (documents_seen, None),
-                            ComputeIdsAndMergeDocuments { documents_seen, total_documents } => (documents_seen, Some(total_documents)),
-                            IndexDocuments { documents_seen, total_documents } => (documents_seen, Some(total_documents)),
-                            MergeDataIntoFinalDatabase { databases_seen, total_databases } => (databases_seen, Some(total_databases)),
+                        match format.as_str() {
+                            "csv" => builder.update_format(UpdateFormat::Csv),
+                            "json" => builder.update_format(UpdateFormat::Json),
+                            "json-stream" => builder.update_format(UpdateFormat::JsonStream),
+                            otherwise => panic!("invalid update format {:?}", otherwise),
                         };
-                        let _ = update_status_sender_cloned.send(UpdateStatus::Progressing {
-                            update_id,
-                            meta: UpdateMetaProgress::DocumentsAddition {
-                                step: indexing_step.step(),
-                                total_steps: indexing_step.number_of_steps(),
-                                current,
-                                total,
-                            },
-                        });
-                    });
 
-                    match result {
-                        Ok(_) => wtxn.commit().map_err(Into::into),
-                        Err(e) => Err(e.into()),
-                    }
-                }
-                UpdateMeta::ClearDocuments => {
-                    // We must use the write transaction of the update here.
-                    let mut wtxn = index_cloned.write_txn()?;
-                    let builder = update_builder.clear_documents(&mut wtxn, &index_cloned);
-
-                    match builder.execute() {
-                        Ok(_count) => wtxn.commit().map_err(Into::into),
-                        Err(e) => Err(e.into()),
-                    }
-                }
-                UpdateMeta::Settings(settings) => {
-                    // We must use the write transaction of the update here.
-                    let mut wtxn = index_cloned.write_txn()?;
-                    let mut builder = update_builder.settings(&mut wtxn, &index_cloned);
-
-                    // We transpose the settings JSON struct into a real setting update.
-                    match settings.searchable_attributes {
-                        Setting::Set(searchable_attributes) => builder.set_searchable_fields(searchable_attributes),
-                        Setting::Reset => builder.reset_searchable_fields(),
-                        Setting::NotSet => ()
-                    }
-
-                    // We transpose the settings JSON struct into a real setting update.
-                    match settings.displayed_attributes {
-                        Setting::Set(displayed_attributes) => builder.set_displayed_fields(displayed_attributes),
-                        Setting::Reset => builder.reset_displayed_fields(),
-                        Setting::NotSet => ()
-                    }
-
-                    // We transpose the settings JSON struct into a real setting update.
-                    match settings.filterable_attributes {
-                        Setting::Set(filterable_attributes) => builder.set_filterable_fields(filterable_attributes),
-                        Setting::Reset => builder.reset_filterable_fields(),
-                        Setting::NotSet => ()
-                    }
-
-                    // We transpose the settings JSON struct into a real setting update.
-                    match settings.criteria {
-                        Setting::Set(criteria) => builder.set_criteria(criteria),
-                        Setting::Reset => builder.reset_criteria(),
-                        Setting::NotSet => ()
-                    }
-
-                    // We transpose the settings JSON struct into a real setting update.
-                    match settings.stop_words {
-                        Setting::Set(stop_words) => builder.set_stop_words(stop_words),
-                        Setting::Reset => builder.reset_stop_words(),
-                        Setting::NotSet => ()
-                    }
-
-                    // We transpose the settings JSON struct into a real setting update.
-                    match settings.synonyms {
-                        Setting::Set(synonyms) => builder.set_synonyms(synonyms),
-                        Setting::Reset => builder.reset_synonyms(),
-                        Setting::NotSet => ()
-                    }
-
-                    let result = builder.execute(|indexing_step, update_id| {
-                        let (current, total) = match indexing_step {
-                            TransformFromUserIntoGenericFormat { documents_seen } => (documents_seen, None),
-                            ComputeIdsAndMergeDocuments { documents_seen, total_documents } => (documents_seen, Some(total_documents)),
-                            IndexDocuments { documents_seen, total_documents } => (documents_seen, Some(total_documents)),
-                            MergeDataIntoFinalDatabase { databases_seen, total_databases } => (databases_seen, Some(total_databases)),
+                        match method.as_str() {
+                            "replace" => builder
+                                .index_documents_method(IndexDocumentsMethod::ReplaceDocuments),
+                            "update" => builder
+                                .index_documents_method(IndexDocumentsMethod::UpdateDocuments),
+                            otherwise => panic!("invalid indexing method {:?}", otherwise),
                         };
-                        let _ = update_status_sender_cloned.send(UpdateStatus::Progressing {
-                            update_id,
-                            meta: UpdateMetaProgress::DocumentsAddition {
-                                step: indexing_step.step(),
-                                total_steps: indexing_step.number_of_steps(),
-                                current,
-                                total,
-                            },
-                        });
-                    });
 
-                    match result {
-                        Ok(_count) => wtxn.commit().map_err(Into::into),
-                        Err(e) => Err(e.into()),
+                        let reader = match encoding.as_deref() {
+                            Some("gzip") => Box::new(GzDecoder::new(content)),
+                            None => Box::new(content) as Box<dyn io::Read>,
+                            otherwise => panic!("invalid encoding format {:?}", otherwise),
+                        };
+
+                        let result = builder.execute(reader, |indexing_step, update_id| {
+                            let (current, total) = match indexing_step {
+                                TransformFromUserIntoGenericFormat { documents_seen } => {
+                                    (documents_seen, None)
+                                }
+                                ComputeIdsAndMergeDocuments { documents_seen, total_documents } => {
+                                    (documents_seen, Some(total_documents))
+                                }
+                                IndexDocuments { documents_seen, total_documents } => {
+                                    (documents_seen, Some(total_documents))
+                                }
+                                MergeDataIntoFinalDatabase { databases_seen, total_databases } => {
+                                    (databases_seen, Some(total_databases))
+                                }
+                            };
+                            let _ = update_status_sender_cloned.send(UpdateStatus::Progressing {
+                                update_id,
+                                meta: UpdateMetaProgress::DocumentsAddition {
+                                    step: indexing_step.step(),
+                                    total_steps: indexing_step.number_of_steps(),
+                                    current,
+                                    total,
+                                },
+                            });
+                        });
+
+                        match result {
+                            Ok(_) => wtxn.commit().map_err(Into::into),
+                            Err(e) => Err(e.into()),
+                        }
                     }
-                }
-                UpdateMeta::Facets(levels) => {
-                    // We must use the write transaction of the update here.
-                    let mut wtxn = index_cloned.write_txn()?;
-                    let mut builder = update_builder.facets(&mut wtxn, &index_cloned);
-                    if let Some(value) = levels.level_group_size {
-                        builder.level_group_size(value);
+                    UpdateMeta::ClearDocuments => {
+                        // We must use the write transaction of the update here.
+                        let mut wtxn = index_cloned.write_txn()?;
+                        let builder = update_builder.clear_documents(&mut wtxn, &index_cloned);
+
+                        match builder.execute() {
+                            Ok(_count) => wtxn.commit().map_err(Into::into),
+                            Err(e) => Err(e.into()),
+                        }
                     }
-                    if let Some(value) = levels.min_level_size {
-                        builder.min_level_size(value);
+                    UpdateMeta::Settings(settings) => {
+                        // We must use the write transaction of the update here.
+                        let mut wtxn = index_cloned.write_txn()?;
+                        let mut builder = update_builder.settings(&mut wtxn, &index_cloned);
+
+                        // We transpose the settings JSON struct into a real setting update.
+                        match settings.searchable_attributes {
+                            Setting::Set(searchable_attributes) => {
+                                builder.set_searchable_fields(searchable_attributes)
+                            }
+                            Setting::Reset => builder.reset_searchable_fields(),
+                            Setting::NotSet => (),
+                        }
+
+                        // We transpose the settings JSON struct into a real setting update.
+                        match settings.displayed_attributes {
+                            Setting::Set(displayed_attributes) => {
+                                builder.set_displayed_fields(displayed_attributes)
+                            }
+                            Setting::Reset => builder.reset_displayed_fields(),
+                            Setting::NotSet => (),
+                        }
+
+                        // We transpose the settings JSON struct into a real setting update.
+                        match settings.filterable_attributes {
+                            Setting::Set(filterable_attributes) => {
+                                builder.set_filterable_fields(filterable_attributes)
+                            }
+                            Setting::Reset => builder.reset_filterable_fields(),
+                            Setting::NotSet => (),
+                        }
+
+                        // We transpose the settings JSON struct into a real setting update.
+                        match settings.criteria {
+                            Setting::Set(criteria) => builder.set_criteria(criteria),
+                            Setting::Reset => builder.reset_criteria(),
+                            Setting::NotSet => (),
+                        }
+
+                        // We transpose the settings JSON struct into a real setting update.
+                        match settings.stop_words {
+                            Setting::Set(stop_words) => builder.set_stop_words(stop_words),
+                            Setting::Reset => builder.reset_stop_words(),
+                            Setting::NotSet => (),
+                        }
+
+                        // We transpose the settings JSON struct into a real setting update.
+                        match settings.synonyms {
+                            Setting::Set(synonyms) => builder.set_synonyms(synonyms),
+                            Setting::Reset => builder.reset_synonyms(),
+                            Setting::NotSet => (),
+                        }
+
+                        let result = builder.execute(|indexing_step, update_id| {
+                            let (current, total) = match indexing_step {
+                                TransformFromUserIntoGenericFormat { documents_seen } => {
+                                    (documents_seen, None)
+                                }
+                                ComputeIdsAndMergeDocuments { documents_seen, total_documents } => {
+                                    (documents_seen, Some(total_documents))
+                                }
+                                IndexDocuments { documents_seen, total_documents } => {
+                                    (documents_seen, Some(total_documents))
+                                }
+                                MergeDataIntoFinalDatabase { databases_seen, total_databases } => {
+                                    (databases_seen, Some(total_databases))
+                                }
+                            };
+                            let _ = update_status_sender_cloned.send(UpdateStatus::Progressing {
+                                update_id,
+                                meta: UpdateMetaProgress::DocumentsAddition {
+                                    step: indexing_step.step(),
+                                    total_steps: indexing_step.number_of_steps(),
+                                    current,
+                                    total,
+                                },
+                            });
+                        });
+
+                        match result {
+                            Ok(_count) => wtxn.commit().map_err(Into::into),
+                            Err(e) => Err(e.into()),
+                        }
                     }
-                    match builder.execute() {
-                        Ok(()) => wtxn.commit().map_err(Into::into),
-                        Err(e) => Err(e.into()),
+                    UpdateMeta::Facets(levels) => {
+                        // We must use the write transaction of the update here.
+                        let mut wtxn = index_cloned.write_txn()?;
+                        let mut builder = update_builder.facets(&mut wtxn, &index_cloned);
+                        if let Some(value) = levels.level_group_size {
+                            builder.level_group_size(value);
+                        }
+                        if let Some(value) = levels.min_level_size {
+                            builder.min_level_size(value);
+                        }
+                        match builder.execute() {
+                            Ok(()) => wtxn.commit().map_err(Into::into),
+                            Err(e) => Err(e.into()),
+                        }
                     }
-                }
-            };
+                };
 
             let meta = match result {
-                Ok(()) => format!("valid update content processed in {:.02?}", before_update.elapsed()),
+                Ok(()) => {
+                    format!("valid update content processed in {:.02?}", before_update.elapsed())
+                }
                 Err(e) => format!("error while processing update content: {:?}", e),
             };
 
@@ -500,7 +525,8 @@ async fn main() -> anyhow::Result<()> {
             let _ = update_status_sender_cloned.send(processed);
 
             Ok(meta)
-        })?;
+        },
+    )?;
 
     // The database name will not change.
     let db_name = opt.database.file_stem().and_then(|s| s.to_str()).unwrap_or("").to_string();
@@ -512,15 +538,11 @@ async fn main() -> anyhow::Result<()> {
     let db_name_cloned = db_name.clone();
     let lmdb_path_cloned = lmdb_path.clone();
     let index_cloned = index.clone();
-    let dash_html_route = warp::filters::method::get()
-        .and(warp::filters::path::end())
-        .map(move || {
+    let dash_html_route =
+        warp::filters::method::get().and(warp::filters::path::end()).map(move || {
             // We retrieve the database size.
-            let db_size = File::open(lmdb_path_cloned.clone())
-                .unwrap()
-                .metadata()
-                .unwrap()
-                .len() as usize;
+            let db_size =
+                File::open(lmdb_path_cloned.clone()).unwrap().metadata().unwrap().len() as usize;
 
             // And the number of documents in the database.
             let rtxn = index_cloned.read_txn().unwrap();
@@ -537,111 +559,105 @@ async fn main() -> anyhow::Result<()> {
         .and(warp::path!("updates"))
         .map(move |header: String| {
             let update_store = update_store_cloned.clone();
-            let mut updates = update_store.iter_metas(|processed, aborted, pending| {
-                let mut updates = Vec::<UpdateStatus<_, UpdateMetaProgress, _>>::new();
-                for result in processed {
-                    let (uid, meta) = result?;
-                    updates.push(UpdateStatus::Processed { update_id: uid.get(), meta });
-                }
-                for result in aborted {
-                    let (uid, meta) = result?;
-                    updates.push(UpdateStatus::Aborted { update_id: uid.get(), meta });
-                }
-                for result in pending {
-                    let (uid, meta) = result?;
-                    updates.push(UpdateStatus::Pending { update_id: uid.get(), meta });
-                }
-                Ok(updates)
-            }).unwrap();
+            let mut updates = update_store
+                .iter_metas(|processed, aborted, pending| {
+                    let mut updates = Vec::<UpdateStatus<_, UpdateMetaProgress, _>>::new();
+                    for result in processed {
+                        let (uid, meta) = result?;
+                        updates.push(UpdateStatus::Processed { update_id: uid.get(), meta });
+                    }
+                    for result in aborted {
+                        let (uid, meta) = result?;
+                        updates.push(UpdateStatus::Aborted { update_id: uid.get(), meta });
+                    }
+                    for result in pending {
+                        let (uid, meta) = result?;
+                        updates.push(UpdateStatus::Pending { update_id: uid.get(), meta });
+                    }
+                    Ok(updates)
+                })
+                .unwrap();
 
             updates.sort_unstable_by(|s1, s2| s1.update_id().cmp(&s2.update_id()).reverse());
 
             if header.contains("text/html") {
                 // We retrieve the database size.
-                let db_size = File::open(lmdb_path_cloned.clone())
-                    .unwrap()
-                    .metadata()
-                    .unwrap()
-                    .len() as usize;
+                let db_size =
+                    File::open(lmdb_path_cloned.clone()).unwrap().metadata().unwrap().len()
+                        as usize;
 
                 // And the number of documents in the database.
                 let rtxn = index_cloned.read_txn().unwrap();
                 let docs_count = index_cloned.clone().number_of_documents(&rtxn).unwrap() as usize;
 
-                let template = UpdatesTemplate {
-                    db_name: db_name.clone(),
-                    db_size,
-                    docs_count,
-                    updates,
-                };
+                let template =
+                    UpdatesTemplate { db_name: db_name.clone(), db_size, docs_count, updates };
                 Box::new(template) as Box<dyn warp::Reply>
             } else {
                 Box::new(warp::reply::json(&updates))
             }
         });
 
-    let dash_bulma_route = warp::filters::method::get()
-        .and(warp::path!("bulma.min.css"))
-        .map(|| Response::builder()
-            .header("content-type", "text/css; charset=utf-8")
-            .body(include_str!("../public/bulma.min.css"))
-        );
+    let dash_bulma_route =
+        warp::filters::method::get().and(warp::path!("bulma.min.css")).map(|| {
+            Response::builder()
+                .header("content-type", "text/css; charset=utf-8")
+                .body(include_str!("../public/bulma.min.css"))
+        });
 
-    let dash_bulma_dark_route = warp::filters::method::get()
-        .and(warp::path!("bulma-prefers-dark.min.css"))
-        .map(|| Response::builder()
-            .header("content-type", "text/css; charset=utf-8")
-            .body(include_str!("../public/bulma-prefers-dark.min.css"))
-        );
+    let dash_bulma_dark_route =
+        warp::filters::method::get().and(warp::path!("bulma-prefers-dark.min.css")).map(|| {
+            Response::builder()
+                .header("content-type", "text/css; charset=utf-8")
+                .body(include_str!("../public/bulma-prefers-dark.min.css"))
+        });
 
-    let dash_style_route = warp::filters::method::get()
-        .and(warp::path!("style.css"))
-        .map(|| Response::builder()
+    let dash_style_route = warp::filters::method::get().and(warp::path!("style.css")).map(|| {
+        Response::builder()
             .header("content-type", "text/css; charset=utf-8")
             .body(include_str!("../public/style.css"))
-        );
+    });
 
-    let dash_jquery_route = warp::filters::method::get()
-        .and(warp::path!("jquery-3.4.1.min.js"))
-        .map(|| Response::builder()
-            .header("content-type", "application/javascript; charset=utf-8")
-            .body(include_str!("../public/jquery-3.4.1.min.js"))
-        );
+    let dash_jquery_route =
+        warp::filters::method::get().and(warp::path!("jquery-3.4.1.min.js")).map(|| {
+            Response::builder()
+                .header("content-type", "application/javascript; charset=utf-8")
+                .body(include_str!("../public/jquery-3.4.1.min.js"))
+        });
 
-    let dash_filesize_route = warp::filters::method::get()
-        .and(warp::path!("filesize.min.js"))
-        .map(|| Response::builder()
-            .header("content-type", "application/javascript; charset=utf-8")
-            .body(include_str!("../public/filesize.min.js"))
-        );
+    let dash_filesize_route =
+        warp::filters::method::get().and(warp::path!("filesize.min.js")).map(|| {
+            Response::builder()
+                .header("content-type", "application/javascript; charset=utf-8")
+                .body(include_str!("../public/filesize.min.js"))
+        });
 
-    let dash_script_route = warp::filters::method::get()
-        .and(warp::path!("script.js"))
-        .map(|| Response::builder()
+    let dash_script_route = warp::filters::method::get().and(warp::path!("script.js")).map(|| {
+        Response::builder()
             .header("content-type", "application/javascript; charset=utf-8")
             .body(include_str!("../public/script.js"))
-        );
+    });
 
-    let updates_script_route = warp::filters::method::get()
-        .and(warp::path!("updates-script.js"))
-        .map(|| Response::builder()
-            .header("content-type", "application/javascript; charset=utf-8")
-            .body(include_str!("../public/updates-script.js"))
-        );
+    let updates_script_route =
+        warp::filters::method::get().and(warp::path!("updates-script.js")).map(|| {
+            Response::builder()
+                .header("content-type", "application/javascript; charset=utf-8")
+                .body(include_str!("../public/updates-script.js"))
+        });
 
-    let dash_logo_white_route = warp::filters::method::get()
-        .and(warp::path!("logo-white.svg"))
-        .map(|| Response::builder()
-            .header("content-type", "image/svg+xml")
-            .body(include_str!("../public/logo-white.svg"))
-        );
+    let dash_logo_white_route =
+        warp::filters::method::get().and(warp::path!("logo-white.svg")).map(|| {
+            Response::builder()
+                .header("content-type", "image/svg+xml")
+                .body(include_str!("../public/logo-white.svg"))
+        });
 
-    let dash_logo_black_route = warp::filters::method::get()
-        .and(warp::path!("logo-black.svg"))
-        .map(|| Response::builder()
-            .header("content-type", "image/svg+xml")
-            .body(include_str!("../public/logo-black.svg"))
-        );
+    let dash_logo_black_route =
+        warp::filters::method::get().and(warp::path!("logo-black.svg")).map(|| {
+            Response::builder()
+                .header("content-type", "image/svg+xml")
+                .body(include_str!("../public/logo-black.svg"))
+        });
 
     #[derive(Debug, Deserialize)]
     #[serde(untagged)]
@@ -719,7 +735,8 @@ async fn main() -> anyhow::Result<()> {
                 search.filter(condition);
             }
 
-            let SearchResult { matching_words, candidates, documents_ids } = search.execute().unwrap();
+            let SearchResult { matching_words, candidates, documents_ids } =
+                search.execute().unwrap();
 
             let number_of_candidates = candidates.len();
             let facets = if query.facet_distribution == Some(true) {
@@ -745,17 +762,18 @@ async fn main() -> anyhow::Result<()> {
             for (_id, obkv) in index.documents(&rtxn, documents_ids).unwrap() {
                 let mut object = obkv_to_json(&displayed_fields, &fields_ids_map, obkv).unwrap();
                 if !disable_highlighting {
-                    highlighter.highlight_record(&mut object, &matching_words, &attributes_to_highlight);
+                    highlighter.highlight_record(
+                        &mut object,
+                        &matching_words,
+                        &attributes_to_highlight,
+                    );
                 }
 
                 documents.push(object);
             }
 
-            let answer = Answer {
-                documents,
-                number_of_candidates,
-                facets: facets.unwrap_or_default(),
-            };
+            let answer =
+                Answer { documents, number_of_candidates, facets: facets.unwrap_or_default() };
 
             Response::builder()
                 .header("Content-Type", "application/json")
@@ -764,9 +782,8 @@ async fn main() -> anyhow::Result<()> {
         });
 
     let index_cloned = index.clone();
-    let document_route = warp::filters::method::get()
-        .and(warp::path!("document" / String))
-        .map(move |id: String| {
+    let document_route = warp::filters::method::get().and(warp::path!("document" / String)).map(
+        move |id: String| {
             let index = index_cloned.clone();
             let rtxn = index.read_txn().unwrap();
 
@@ -780,30 +797,31 @@ async fn main() -> anyhow::Result<()> {
             match external_documents_ids.get(&id) {
                 Some(document_id) => {
                     let document_id = document_id as u32;
-                    let (_, obkv) = index.documents(&rtxn, Some(document_id)).unwrap().pop().unwrap();
+                    let (_, obkv) =
+                        index.documents(&rtxn, Some(document_id)).unwrap().pop().unwrap();
                     let document = obkv_to_json(&displayed_fields, &fields_ids_map, obkv).unwrap();
 
                     Response::builder()
                         .header("Content-Type", "application/json")
                         .body(serde_json::to_string(&document).unwrap())
                 }
-                None => {
-                    Response::builder()
-                        .status(404)
-                        .body(format!("Document with id {:?} not found.", id))
-                }
+                None => Response::builder()
+                    .status(404)
+                    .body(format!("Document with id {:?} not found.", id)),
             }
-        });
+        },
+    );
 
     async fn buf_stream(
         update_store: Arc<UpdateStore<UpdateMeta, String>>,
-        update_status_sender: broadcast::Sender<UpdateStatus<UpdateMeta, UpdateMetaProgress, String>>,
+        update_status_sender: broadcast::Sender<
+            UpdateStatus<UpdateMeta, UpdateMetaProgress, String>,
+        >,
         update_method: Option<String>,
         update_format: UpdateFormat,
         encoding: Option<String>,
-        mut stream: impl futures::Stream<Item=Result<impl bytes::Buf, warp::Error>> + Unpin,
-    ) -> Result<impl warp::Reply, warp::Rejection>
-    {
+        mut stream: impl futures::Stream<Item = Result<impl bytes::Buf, warp::Error>> + Unpin,
+    ) -> Result<impl warp::Reply, warp::Rejection> {
         let file = tokio::task::block_in_place(tempfile::tempfile).unwrap();
         let mut file = TFile::from_std(file);
 
@@ -869,9 +887,8 @@ async fn main() -> anyhow::Result<()> {
 
     let update_store_cloned = update_store.clone();
     let update_status_sender_cloned = update_status_sender.clone();
-    let clearing_route = warp::filters::method::post()
-        .and(warp::path!("clear-documents"))
-        .map(move || {
+    let clearing_route =
+        warp::filters::method::post().and(warp::path!("clear-documents")).map(move || {
             let meta = UpdateMeta::ClearDocuments;
             let update_id = update_store_cloned.register_update(&meta, &[]).unwrap();
             let _ = update_status_sender_cloned.send(UpdateStatus::Pending { update_id, meta });
@@ -919,9 +936,8 @@ async fn main() -> anyhow::Result<()> {
 
     let update_store_cloned = update_store.clone();
     let update_status_sender_cloned = update_status_sender.clone();
-    let abort_pending_updates_route = warp::filters::method::delete()
-        .and(warp::path!("updates"))
-        .map(move || {
+    let abort_pending_updates_route =
+        warp::filters::method::delete().and(warp::path!("updates")).map(move || {
             let updates = update_store_cloned.abort_pendings().unwrap();
             for (update_id, meta) in updates {
                 let _ = update_status_sender_cloned.send(UpdateStatus::Aborted { update_id, meta });
@@ -930,25 +946,22 @@ async fn main() -> anyhow::Result<()> {
             warp::reply()
         });
 
-    let update_ws_route = warp::ws()
-        .and(warp::path!("updates" / "ws"))
-        .map(move |ws: warp::ws::Ws| {
+    let update_ws_route =
+        warp::ws().and(warp::path!("updates" / "ws")).map(move |ws: warp::ws::Ws| {
             // And then our closure will be called when it completes...
             let update_status_receiver = update_status_sender.subscribe();
             ws.on_upgrade(|websocket| {
                 // Just echo all updates messages...
                 update_status_receiver
                     .into_stream()
-                    .flat_map(|result| {
-                        match result {
-                            Ok(status) => {
-                                let msg = serde_json::to_string(&status).unwrap();
-                                stream::iter(Some(Ok(Message::text(msg))))
-                            }
-                            Err(e) => {
-                                eprintln!("channel error: {:?}", e);
-                                stream::iter(None)
-                            }
+                    .flat_map(|result| match result {
+                        Ok(status) => {
+                            let msg = serde_json::to_string(&status).unwrap();
+                            stream::iter(Some(Ok(Message::text(msg))))
+                        }
+                        Err(e) => {
+                            eprintln!("channel error: {:?}", e);
+                            stream::iter(None)
                         }
                     })
                     .forward(websocket)
@@ -988,10 +1001,9 @@ async fn main() -> anyhow::Result<()> {
 
 #[cfg(test)]
 mod tests {
-    use maplit::{btreeset,hashmap, hashset};
-    use serde_test::{assert_tokens, Token};
-
+    use maplit::{btreeset, hashmap, hashset};
     use milli::update::Setting;
+    use serde_test::{assert_tokens, Token};
 
     use crate::Settings;
 
@@ -1000,50 +1012,53 @@ mod tests {
         let settings = Settings {
             displayed_attributes: Setting::Set(vec!["name".to_string()]),
             searchable_attributes: Setting::Set(vec!["age".to_string()]),
-            filterable_attributes: Setting::Set(hashset!{ "age".to_string() }),
+            filterable_attributes: Setting::Set(hashset! { "age".to_string() }),
             criteria: Setting::Set(vec!["asc(age)".to_string()]),
             stop_words: Setting::Set(btreeset! { "and".to_string() }),
-            synonyms: Setting::Set(hashmap!{ "alex".to_string() => vec!["alexey".to_string()] })
+            synonyms: Setting::Set(hashmap! { "alex".to_string() => vec!["alexey".to_string()] }),
         };
 
-        assert_tokens(&settings, &[
-            Token::Struct { name: "Settings", len: 6 },
-            Token::Str("displayedAttributes"),
-            Token::Some,
-            Token::Seq { len: Some(1) },
-            Token::Str("name"),
-            Token::SeqEnd,
-            Token::Str("searchableAttributes"),
-            Token::Some,
-            Token::Seq { len: Some(1) },
-            Token::Str("age"),
-            Token::SeqEnd,
-            Token::Str("facetedAttributes"),
-            Token::Some,
-            Token::Map { len: Some(1) },
-            Token::Str("age"),
-            Token::Str("integer"),
-            Token::MapEnd,
-            Token::Str("criteria"),
-            Token::Some,
-            Token::Seq { len: Some(1) },
-            Token::Str("asc(age)"),
-            Token::SeqEnd,
-            Token::Str("stopWords"),
-            Token::Some,
-            Token::Seq { len: Some(1) },
-            Token::Str("and"),
-            Token::SeqEnd,
-            Token::Str("synonyms"),
-            Token::Some,
-            Token::Map { len: Some(1) },
-            Token::Str("alex"),
-            Token::Seq {len: Some(1) },
-            Token::Str("alexey"),
-            Token::SeqEnd,
-            Token::MapEnd,
-            Token::StructEnd,
-        ]);
+        assert_tokens(
+            &settings,
+            &[
+                Token::Struct { name: "Settings", len: 6 },
+                Token::Str("displayedAttributes"),
+                Token::Some,
+                Token::Seq { len: Some(1) },
+                Token::Str("name"),
+                Token::SeqEnd,
+                Token::Str("searchableAttributes"),
+                Token::Some,
+                Token::Seq { len: Some(1) },
+                Token::Str("age"),
+                Token::SeqEnd,
+                Token::Str("facetedAttributes"),
+                Token::Some,
+                Token::Map { len: Some(1) },
+                Token::Str("age"),
+                Token::Str("integer"),
+                Token::MapEnd,
+                Token::Str("criteria"),
+                Token::Some,
+                Token::Seq { len: Some(1) },
+                Token::Str("asc(age)"),
+                Token::SeqEnd,
+                Token::Str("stopWords"),
+                Token::Some,
+                Token::Seq { len: Some(1) },
+                Token::Str("and"),
+                Token::SeqEnd,
+                Token::Str("synonyms"),
+                Token::Some,
+                Token::Map { len: Some(1) },
+                Token::Str("alex"),
+                Token::Seq { len: Some(1) },
+                Token::Str("alexey"),
+                Token::SeqEnd,
+                Token::MapEnd,
+                Token::StructEnd,
+            ],
+        );
     }
 
     #[test]
@@ -1057,22 +1072,25 @@ mod tests {
             synonyms: Setting::Reset,
         };
 
-        assert_tokens(&settings, &[
-            Token::Struct { name: "Settings", len: 6 },
-            Token::Str("displayedAttributes"),
-            Token::None,
-            Token::Str("searchableAttributes"),
-            Token::None,
-            Token::Str("facetedAttributes"),
-            Token::None,
-            Token::Str("criteria"),
-            Token::None,
-            Token::Str("stopWords"),
-            Token::None,
-            Token::Str("synonyms"),
-            Token::None,
-            Token::StructEnd,
-        ]);
+        assert_tokens(
+            &settings,
+            &[
+                Token::Struct { name: "Settings", len: 6 },
+                Token::Str("displayedAttributes"),
+                Token::None,
+                Token::Str("searchableAttributes"),
+                Token::None,
+                Token::Str("facetedAttributes"),
+                Token::None,
+                Token::Str("criteria"),
+                Token::None,
+                Token::Str("stopWords"),
+                Token::None,
+                Token::Str("synonyms"),
+                Token::None,
+                Token::StructEnd,
+            ],
+        );
     }
 
     #[test]
@@ -1086,9 +1104,6 @@ mod tests {
             synonyms: Setting::NotSet,
         };
 
-        assert_tokens(&settings, &[
-            Token::Struct { name: "Settings", len: 0 },
-            Token::StructEnd,
-        ]);
+        assert_tokens(&settings, &[Token::Struct { name: "Settings", len: 0 }, Token::StructEnd]);
     }
 }
diff --git a/http-ui/src/update_store.rs b/http-ui/src/update_store.rs
index 122ee6031..b77057fda 100644
--- a/http-ui/src/update_store.rs
+++ b/http-ui/src/update_store.rs
@@ -4,9 +4,9 @@ use std::path::Path;
 use std::sync::Arc;
 
 use crossbeam_channel::Sender;
-use heed::types::{OwnedType, DecodeIgnore, SerdeJson, ByteSlice};
-use heed::{EnvOpenOptions, Env, Database};
-use serde::{Serialize, Deserialize};
+use heed::types::{ByteSlice, DecodeIgnore, OwnedType, SerdeJson};
+use heed::{Database, Env, EnvOpenOptions};
+use serde::{Deserialize, Serialize};
 
 pub type BEU64 = heed::zerocopy::U64<heed::byteorder::BE>;
 
@@ -25,7 +25,9 @@ pub trait UpdateHandler<M, N> {
 }
 
 impl<M, N, F> UpdateHandler<M, N> for F
-where F: FnMut(u64, M, &[u8]) -> heed::Result<N> + Send + 'static {
+where
+    F: FnMut(u64, M, &[u8]) -> heed::Result<N> + Send + 'static,
+{
     fn handle_update(&mut self, update_id: u64, meta: M, content: &[u8]) -> heed::Result<N> {
         self(update_id, meta, content)
     }
@@ -82,26 +84,17 @@ impl<M: 'static, N: 'static> UpdateStore<M, N> {
 
     /// Returns the new biggest id to use to store the new update.
     fn new_update_id(&self, txn: &heed::RoTxn) -> heed::Result<u64> {
-        let last_pending = self.pending_meta
-            .remap_data_type::<DecodeIgnore>()
-            .last(txn)?
-            .map(|(k, _)| k.get());
+        let last_pending =
+            self.pending_meta.remap_data_type::<DecodeIgnore>().last(txn)?.map(|(k, _)| k.get());
 
-        let last_processed = self.processed_meta
-            .remap_data_type::<DecodeIgnore>()
-            .last(txn)?
-            .map(|(k, _)| k.get());
+        let last_processed =
+            self.processed_meta.remap_data_type::<DecodeIgnore>().last(txn)?.map(|(k, _)| k.get());
 
-        let last_aborted = self.aborted_meta
-            .remap_data_type::<DecodeIgnore>()
-            .last(txn)?
-            .map(|(k, _)| k.get());
+        let last_aborted =
+            self.aborted_meta.remap_data_type::<DecodeIgnore>().last(txn)?.map(|(k, _)| k.get());
 
-        let last_update_id = [last_pending, last_processed, last_aborted]
-            .iter()
-            .copied()
-            .flatten()
-            .max();
+        let last_update_id =
+            [last_pending, last_processed, last_aborted].iter().copied().flatten().max();
 
         match last_update_id {
             Some(last_id) => Ok(last_id + 1),
@@ -112,7 +105,8 @@ impl<M: 'static, N: 'static> UpdateStore<M, N> {
     /// Registers the update content in the pending store and the meta
     /// into the pending-meta store. Returns the new unique update id.
     pub fn register_update(&self, meta: &M, content: &[u8]) -> heed::Result<u64>
-    where M: Serialize,
+    where
+        M: Serialize,
     {
         let mut wtxn = self.env.write_txn()?;
 
@@ -152,9 +146,8 @@ impl<M: 'static, N: 'static> UpdateStore<M, N> {
         // a reader while processing it, not a writer.
         match first_meta {
             Some((first_id, first_meta)) => {
-                let first_content = self.pending
-                    .get(&rtxn, &first_id)?
-                    .expect("associated update content");
+                let first_content =
+                    self.pending.get(&rtxn, &first_id)?.expect("associated update content");
 
                 // Process the pending update using the provided user function.
                 let new_meta = handler.handle_update(first_id.get(), first_meta, first_content)?;
@@ -170,15 +163,16 @@ impl<M: 'static, N: 'static> UpdateStore<M, N> {
                 wtxn.commit()?;
 
                 Ok(Some((first_id.get(), new_meta)))
-            },
-            None => Ok(None)
+            }
+            None => Ok(None),
         }
     }
 
     /// The id and metadata of the update that is currently being processed,
     /// `None` if no update is being processed.
     pub fn processing_update(&self) -> heed::Result<Option<(u64, M)>>
-    where M: for<'a> Deserialize<'a>,
+    where
+        M: for<'a> Deserialize<'a>,
     {
         let rtxn = self.env.read_txn()?;
         match self.pending_meta.first(&rtxn)? {
@@ -242,7 +236,8 @@ impl<M: 'static, N: 'static> UpdateStore<M, N> {
     /// that as already been processed or which doesn't actually exist, will
     /// return `None`.
     pub fn abort_update(&self, update_id: u64) -> heed::Result<Option<M>>
-    where M: Serialize + for<'a> Deserialize<'a>,
+    where
+        M: Serialize + for<'a> Deserialize<'a>,
     {
         let mut wtxn = self.env.write_txn()?;
         let key = BEU64::new(update_id);
@@ -269,7 +264,8 @@ impl<M: 'static, N: 'static> UpdateStore<M, N> {
     /// Aborts all the pending updates, and not the one being currently processed.
     /// Returns the update metas and ids that were successfully aborted.
     pub fn abort_pendings(&self) -> heed::Result<Vec<(u64, M)>>
-    where M: Serialize + for<'a> Deserialize<'a>,
+    where
+        M: Serialize + for<'a> Deserialize<'a>,
     {
         let mut wtxn = self.env.write_txn()?;
         let mut aborted_updates = Vec::new();
@@ -303,17 +299,19 @@ pub enum UpdateStatusMeta<M, N> {
 
 #[cfg(test)]
 mod tests {
-    use super::*;
     use std::thread;
     use std::time::{Duration, Instant};
 
+    use super::*;
+
     #[test]
     fn simple() {
         let dir = tempfile::tempdir().unwrap();
         let options = EnvOpenOptions::new();
-        let update_store = UpdateStore::open(options, dir, |_id, meta: String, _content:&_| {
+        let update_store = UpdateStore::open(options, dir, |_id, meta: String, _content: &_| {
             Ok(meta + " processed")
-        }).unwrap();
+        })
+        .unwrap();
 
         let meta = String::from("kiki");
         let update_id = update_store.register_update(&meta, &[]).unwrap();
@@ -329,10 +327,11 @@ mod tests {
     fn long_running_update() {
         let dir = tempfile::tempdir().unwrap();
         let options = EnvOpenOptions::new();
-        let update_store = UpdateStore::open(options, dir, |_id, meta: String, _content:&_| {
+        let update_store = UpdateStore::open(options, dir, |_id, meta: String, _content: &_| {
             thread::sleep(Duration::from_millis(400));
             Ok(meta + " processed")
-        }).unwrap();
+        })
+        .unwrap();
 
         let before_register = Instant::now();
 
diff --git a/infos/src/main.rs b/infos/src/main.rs
index b0c304de0..151e8c664 100644
--- a/infos/src/main.rs
+++ b/infos/src/main.rs
@@ -1,16 +1,14 @@
 use std::fmt::Write as _;
 use std::path::PathBuf;
-use std::{str, io, fmt};
+use std::{fmt, io, str};
 
 use anyhow::Context;
 use byte_unit::Byte;
 use heed::EnvOpenOptions;
-use structopt::StructOpt;
-
 use milli::facet::FacetType;
 use milli::index::db_name::*;
 use milli::{Index, TreeLevel};
-
+use structopt::StructOpt;
 use Command::*;
 
 #[cfg(target_os = "linux")]
@@ -257,53 +255,55 @@ fn main() -> anyhow::Result<()> {
         WordsDocids { full_display, words } => words_docids(&index, &rtxn, !full_display, words),
         WordsPrefixesDocids { full_display, prefixes } => {
             words_prefixes_docids(&index, &rtxn, !full_display, prefixes)
-        },
+        }
         FacetNumbersDocids { full_display, field_name } => {
             facet_values_docids(&index, &rtxn, !full_display, FacetType::Number, field_name)
-        },
+        }
         FacetStringsDocids { full_display, field_name } => {
             facet_values_docids(&index, &rtxn, !full_display, FacetType::String, field_name)
-        },
+        }
         WordsLevelPositionsDocids { full_display, words } => {
             words_level_positions_docids(&index, &rtxn, !full_display, words)
-        },
+        }
         WordPrefixesLevelPositionsDocids { full_display, prefixes } => {
             word_prefixes_level_positions_docids(&index, &rtxn, !full_display, prefixes)
-        },
+        }
         FieldIdWordCountDocids { full_display, field_name } => {
             field_id_word_count_docids(&index, &rtxn, !full_display, field_name)
-        },
+        }
         DocidsWordsPositions { full_display, internal_documents_ids } => {
             docids_words_positions(&index, &rtxn, !full_display, internal_documents_ids)
-        },
+        }
         FacetNumberStats { field_name } => facet_number_stats(&index, &rtxn, field_name),
         AverageNumberOfWordsByDoc => average_number_of_words_by_doc(&index, &rtxn),
-        AverageNumberOfPositionsByWord => {
-            average_number_of_positions_by_word(&index, &rtxn)
-        },
+        AverageNumberOfPositionsByWord => average_number_of_positions_by_word(&index, &rtxn),
         SizeOfDatabase { databases } => size_of_databases(&index, &rtxn, databases),
         DatabaseStats { database } => database_stats(&index, &rtxn, &database),
         WordPairProximitiesDocids { full_display, word1, word2 } => {
             word_pair_proximities_docids(&index, &rtxn, !full_display, word1, word2)
-        },
+        }
         ExportWordsFst => export_words_fst(&index, &rtxn),
         ExportWordsPrefixFst => export_words_prefix_fst(&index, &rtxn),
         ExportDocuments { internal_documents_ids } => {
             export_documents(&index, &rtxn, internal_documents_ids)
-        },
+        }
     }
 }
 
 fn most_common_words(index: &Index, rtxn: &heed::RoTxn, limit: usize) -> anyhow::Result<()> {
-    use std::collections::BinaryHeap;
     use std::cmp::Reverse;
+    use std::collections::BinaryHeap;
 
     let mut heap = BinaryHeap::with_capacity(limit + 1);
     for result in index.word_docids.iter(rtxn)? {
-        if limit == 0 { break }
+        if limit == 0 {
+            break;
+        }
         let (word, docids) = result?;
         heap.push((Reverse(docids.len()), word));
-        if heap.len() > limit { heap.pop(); }
+        if heap.len() > limit {
+            heap.pop();
+        }
     }
 
     let stdout = io::stdout();
@@ -323,7 +323,7 @@ fn facet_values_iter<'txn, KC: 'txn, DC: 'txn>(
     rtxn: &'txn heed::RoTxn,
     db: heed::Database<KC, DC>,
     field_id: u8,
-) -> heed::Result<Box<dyn Iterator<Item=heed::Result<(KC::DItem, DC::DItem)>> + 'txn>>
+) -> heed::Result<Box<dyn Iterator<Item = heed::Result<(KC::DItem, DC::DItem)>> + 'txn>>
 where
     KC: heed::BytesDecode<'txn>,
     DC: heed::BytesDecode<'txn>,
@@ -347,7 +347,8 @@ fn facet_number_value_to_string<T: fmt::Debug>(level: u8, left: T, right: T) ->
 fn biggest_value_sizes(index: &Index, rtxn: &heed::RoTxn, limit: usize) -> anyhow::Result<()> {
     use std::cmp::Reverse;
     use std::collections::BinaryHeap;
-    use heed::types::{Str, ByteSlice};
+
+    use heed::types::{ByteSlice, Str};
 
     let Index {
         env: _env,
@@ -387,71 +388,93 @@ fn biggest_value_sizes(index: &Index, rtxn: &heed::RoTxn, limit: usize) -> anyho
         let words_fst = index.words_fst(rtxn)?;
         let length = words_fst.as_fst().as_bytes().len();
         heap.push(Reverse((length, format!("words-fst"), main_name)));
-        if heap.len() > limit { heap.pop(); }
+        if heap.len() > limit {
+            heap.pop();
+        }
 
         // Fetch the word prefix FST
         let words_prefixes_fst = index.words_prefixes_fst(rtxn)?;
         let length = words_prefixes_fst.as_fst().as_bytes().len();
         heap.push(Reverse((length, format!("words-prefixes-fst"), main_name)));
-        if heap.len() > limit { heap.pop(); }
+        if heap.len() > limit {
+            heap.pop();
+        }
 
         if let Some(documents_ids) = main.get::<_, Str, ByteSlice>(rtxn, "documents-ids")? {
             heap.push(Reverse((documents_ids.len(), format!("documents-ids"), main_name)));
-            if heap.len() > limit { heap.pop(); }
+            if heap.len() > limit {
+                heap.pop();
+            }
         }
 
         for result in word_docids.remap_data_type::<ByteSlice>().iter(rtxn)? {
             let (word, value) = result?;
             heap.push(Reverse((value.len(), word.to_string(), word_docids_name)));
-            if heap.len() > limit { heap.pop(); }
+            if heap.len() > limit {
+                heap.pop();
+            }
         }
 
         for result in word_prefix_docids.remap_data_type::<ByteSlice>().iter(rtxn)? {
             let (word, value) = result?;
             heap.push(Reverse((value.len(), word.to_string(), word_prefix_docids_name)));
-            if heap.len() > limit { heap.pop(); }
+            if heap.len() > limit {
+                heap.pop();
+            }
         }
 
         for result in docid_word_positions.remap_data_type::<ByteSlice>().iter(rtxn)? {
             let ((docid, word), value) = result?;
             let key = format!("{} {}", docid, word);
             heap.push(Reverse((value.len(), key, docid_word_positions_name)));
-            if heap.len() > limit { heap.pop(); }
+            if heap.len() > limit {
+                heap.pop();
+            }
         }
 
         for result in word_pair_proximity_docids.remap_data_type::<ByteSlice>().iter(rtxn)? {
             let ((word1, word2, prox), value) = result?;
             let key = format!("{} {} {}", word1, word2, prox);
             heap.push(Reverse((value.len(), key, word_pair_proximity_docids_name)));
-            if heap.len() > limit { heap.pop(); }
+            if heap.len() > limit {
+                heap.pop();
+            }
         }
 
         for result in word_prefix_pair_proximity_docids.remap_data_type::<ByteSlice>().iter(rtxn)? {
             let ((word, prefix, prox), value) = result?;
             let key = format!("{} {} {}", word, prefix, prox);
             heap.push(Reverse((value.len(), key, word_prefix_pair_proximity_docids_name)));
-            if heap.len() > limit { heap.pop(); }
+            if heap.len() > limit {
+                heap.pop();
+            }
         }
 
         for result in word_level_position_docids.remap_data_type::<ByteSlice>().iter(rtxn)? {
             let ((word, level, left, right), value) = result?;
             let key = format!("{} {} {:?}", word, level, left..=right);
             heap.push(Reverse((value.len(), key, word_level_position_docids_name)));
-            if heap.len() > limit { heap.pop(); }
+            if heap.len() > limit {
+                heap.pop();
+            }
         }
 
         for result in word_prefix_level_position_docids.remap_data_type::<ByteSlice>().iter(rtxn)? {
             let ((word, level, left, right), value) = result?;
             let key = format!("{} {} {:?}", word, level, left..=right);
             heap.push(Reverse((value.len(), key, word_prefix_level_position_docids_name)));
-            if heap.len() > limit { heap.pop(); }
+            if heap.len() > limit {
+                heap.pop();
+            }
         }
 
         for result in field_id_word_count_docids.remap_data_type::<ByteSlice>().iter(rtxn)? {
             let ((field_id, word_count), docids) = result?;
             let key = format!("{} {}", field_id, word_count);
             heap.push(Reverse((docids.len(), key, field_id_word_count_docids_name)));
-            if heap.len() > limit { heap.pop(); }
+            if heap.len() > limit {
+                heap.pop();
+            }
         }
 
         let faceted_fields = index.faceted_fields_ids(rtxn)?;
@@ -468,7 +491,9 @@ fn biggest_value_sizes(index: &Index, rtxn: &heed::RoTxn, limit: usize) -> anyho
                 write!(&mut output, " (level {})", level)?;
                 let key = format!("{} {}", facet_name, output);
                 heap.push(Reverse((value.len(), key, facet_id_f64_docids_name)));
-                if heap.len() > limit { heap.pop(); }
+                if heap.len() > limit {
+                    heap.pop();
+                }
             }
 
             // List the facet strings of this facet id.
@@ -477,14 +502,18 @@ fn biggest_value_sizes(index: &Index, rtxn: &heed::RoTxn, limit: usize) -> anyho
                 let ((_fid, fvalue), value) = result?;
                 let key = format!("{} {}", facet_name, fvalue);
                 heap.push(Reverse((value.len(), key, facet_id_string_docids_name)));
-                if heap.len() > limit { heap.pop(); }
+                if heap.len() > limit {
+                    heap.pop();
+                }
             }
         }
 
         for result in documents.remap_data_type::<ByteSlice>().iter(rtxn)? {
             let (id, value) = result?;
             heap.push(Reverse((value.len(), id.to_string(), documents_name)));
-            if heap.len() > limit { heap.pop(); }
+            if heap.len() > limit {
+                heap.pop();
+            }
         }
     }
 
@@ -499,7 +528,12 @@ fn biggest_value_sizes(index: &Index, rtxn: &heed::RoTxn, limit: usize) -> anyho
     Ok(wtr.flush()?)
 }
 
-fn words_docids(index: &Index, rtxn: &heed::RoTxn, debug: bool, words: Vec<String>) -> anyhow::Result<()> {
+fn words_docids(
+    index: &Index,
+    rtxn: &heed::RoTxn,
+    debug: bool,
+    words: Vec<String>,
+) -> anyhow::Result<()> {
     let stdout = io::stdout();
     let mut wtr = csv::Writer::from_writer(stdout.lock());
     wtr.write_record(&["word", "documents_ids"])?;
@@ -523,8 +557,7 @@ fn words_prefixes_docids(
     rtxn: &heed::RoTxn,
     debug: bool,
     prefixes: Vec<String>,
-) -> anyhow::Result<()>
-{
+) -> anyhow::Result<()> {
     let stdout = io::stdout();
     let mut wtr = csv::Writer::from_writer(stdout.lock());
     wtr.write_record(&["prefix", "documents_ids"])?;
@@ -561,12 +594,12 @@ fn facet_values_docids(
     debug: bool,
     facet_type: FacetType,
     field_name: String,
-) -> anyhow::Result<()>
-{
+) -> anyhow::Result<()> {
     let fields_ids_map = index.fields_ids_map(&rtxn)?;
     let faceted_fields = index.faceted_fields_ids(&rtxn)?;
 
-    let field_id = fields_ids_map.id(&field_name)
+    let field_id = fields_ids_map
+        .id(&field_name)
         .with_context(|| format!("field {} not found", field_name))?;
 
     if !faceted_fields.contains(&field_id) {
@@ -590,7 +623,7 @@ fn facet_values_docids(
                 };
                 wtr.write_record(&[value, level.to_string(), count.to_string(), docids])?;
             }
-        },
+        }
         FacetType::String => {
             wtr.write_record(&["facet_string", "documents_count", "documents_ids"])?;
             for result in facet_values_iter(rtxn, index.facet_id_string_docids, field_id)? {
@@ -614,8 +647,7 @@ fn words_level_positions_docids(
     rtxn: &heed::RoTxn,
     debug: bool,
     words: Vec<String>,
-) -> anyhow::Result<()>
-{
+) -> anyhow::Result<()> {
     let stdout = io::stdout();
     let mut wtr = csv::Writer::from_writer(stdout.lock());
     wtr.write_record(&["word", "level", "positions", "documents_count", "documents_ids"])?;
@@ -653,8 +685,7 @@ fn word_prefixes_level_positions_docids(
     rtxn: &heed::RoTxn,
     debug: bool,
     prefixes: Vec<String>,
-) -> anyhow::Result<()>
-{
+) -> anyhow::Result<()> {
     let stdout = io::stdout();
     let mut wtr = csv::Writer::from_writer(stdout.lock());
     wtr.write_record(&["prefix", "level", "positions", "documents_count", "documents_ids"])?;
@@ -691,21 +722,20 @@ fn field_id_word_count_docids(
     index: &Index,
     rtxn: &heed::RoTxn,
     debug: bool,
-    field_name: String
-) -> anyhow::Result<()>
-{
+    field_name: String,
+) -> anyhow::Result<()> {
     let stdout = io::stdout();
     let mut wtr = csv::Writer::from_writer(stdout.lock());
     wtr.write_record(&["field_name", "word_count", "docids"])?;
 
-    let field_id = index.fields_ids_map(rtxn)?
+    let field_id = index
+        .fields_ids_map(rtxn)?
         .id(&field_name)
         .with_context(|| format!("unknown field name: {}", &field_name))?;
 
     let left = (field_id, 0);
     let right = (field_id, u8::max_value());
-    let iter = index.field_id_word_count_docids
-        .range(rtxn, &(left..=right))?;
+    let iter = index.field_id_word_count_docids.range(rtxn, &(left..=right))?;
 
     for result in iter {
         let ((_, word_count), docids) = result?;
@@ -725,8 +755,7 @@ fn docids_words_positions(
     rtxn: &heed::RoTxn,
     debug: bool,
     internal_ids: Vec<u32>,
-) -> anyhow::Result<()>
-{
+) -> anyhow::Result<()> {
     let stdout = io::stdout();
     let mut wtr = csv::Writer::from_writer(stdout.lock());
     wtr.write_record(&["document_id", "word", "positions"])?;
@@ -734,9 +763,10 @@ fn docids_words_positions(
     let iter: Box<dyn Iterator<Item = _>> = if internal_ids.is_empty() {
         Box::new(index.docid_word_positions.iter(rtxn)?)
     } else {
-        let vec: heed::Result<Vec<_>> = internal_ids.into_iter().map(|id| {
-            index.docid_word_positions.prefix_iter(rtxn, &(id, ""))
-        }).collect();
+        let vec: heed::Result<Vec<_>> = internal_ids
+            .into_iter()
+            .map(|id| index.docid_word_positions.prefix_iter(rtxn, &(id, "")))
+            .collect();
         Box::new(vec?.into_iter().flatten())
     };
 
@@ -757,7 +787,8 @@ fn facet_number_stats(index: &Index, rtxn: &heed::RoTxn, field_name: String) ->
     let fields_ids_map = index.fields_ids_map(&rtxn)?;
     let faceted_fields = index.faceted_fields_ids(&rtxn)?;
 
-    let field_id = fields_ids_map.id(&field_name)
+    let field_id = fields_ids_map
+        .id(&field_name)
         .with_context(|| format!("field {} not found", field_name))?;
 
     if !faceted_fields.contains(&field_id) {
@@ -808,9 +839,14 @@ fn export_words_prefix_fst(index: &Index, rtxn: &heed::RoTxn) -> anyhow::Result<
     Ok(())
 }
 
-fn export_documents(index: &Index, rtxn: &heed::RoTxn, internal_ids: Vec<u32>) -> anyhow::Result<()> {
+fn export_documents(
+    index: &Index,
+    rtxn: &heed::RoTxn,
+    internal_ids: Vec<u32>,
+) -> anyhow::Result<()> {
     use std::io::{BufWriter, Write as _};
-    use milli::{BEU32, obkv_to_json};
+
+    use milli::{obkv_to_json, BEU32};
 
     let stdout = io::stdout();
     let mut out = BufWriter::new(stdout);
@@ -819,13 +855,13 @@ fn export_documents(index: &Index, rtxn: &heed::RoTxn, internal_ids: Vec<u32>) -
     let displayed_fields: Vec<_> = fields_ids_map.iter().map(|(id, _name)| id).collect();
 
     let iter: Box<dyn Iterator<Item = _>> = if internal_ids.is_empty() {
-        Box::new(index.documents.iter(rtxn)?.map(|result| {
-            result.map(|(_id, obkv)| obkv)
-        }))
+        Box::new(index.documents.iter(rtxn)?.map(|result| result.map(|(_id, obkv)| obkv)))
     } else {
-        Box::new(internal_ids.into_iter().flat_map(|id| {
-            index.documents.get(rtxn, &BEU32::new(id)).transpose()
-        }))
+        Box::new(
+            internal_ids
+                .into_iter()
+                .flat_map(|id| index.documents.get(rtxn, &BEU32::new(id)).transpose()),
+        )
     };
 
     for result in iter {
@@ -842,26 +878,27 @@ fn export_documents(index: &Index, rtxn: &heed::RoTxn, internal_ids: Vec<u32>) -
 
 fn average_number_of_words_by_doc(index: &Index, rtxn: &heed::RoTxn) -> anyhow::Result<()> {
     use heed::types::DecodeIgnore;
-    use milli::{DocumentId, BEU32StrCodec};
+    use milli::{BEU32StrCodec, DocumentId};
 
     let mut words_counts = Vec::new();
     let mut count = 0;
     let mut prev = None as Option<(DocumentId, u32)>;
 
-    let iter = index.docid_word_positions.as_polymorph().iter::<_, BEU32StrCodec, DecodeIgnore>(rtxn)?;
+    let iter =
+        index.docid_word_positions.as_polymorph().iter::<_, BEU32StrCodec, DecodeIgnore>(rtxn)?;
     for result in iter {
         let ((docid, _word), ()) = result?;
 
         match prev.as_mut() {
             Some((prev_docid, prev_count)) if docid == *prev_docid => {
                 *prev_count += 1;
-            },
+            }
             Some((prev_docid, prev_count)) => {
                 words_counts.push(*prev_count);
                 *prev_docid = docid;
                 *prev_count = 0;
                 count += 1;
-            },
+            }
             None => prev = Some((docid, 1)),
         }
     }
@@ -970,16 +1007,15 @@ fn size_of_databases(index: &Index, rtxn: &heed::RoTxn, names: Vec<String>) -> a
 
 fn database_stats(index: &Index, rtxn: &heed::RoTxn, name: &str) -> anyhow::Result<()> {
     use heed::types::ByteSlice;
-    use heed::{Error, BytesDecode};
-    use roaring::RoaringBitmap;
+    use heed::{BytesDecode, Error};
     use milli::{BoRoaringBitmapCodec, CboRoaringBitmapCodec, RoaringBitmapCodec};
+    use roaring::RoaringBitmap;
 
     fn compute_stats<'a, DC: BytesDecode<'a, DItem = RoaringBitmap>>(
         db: heed::PolyDatabase,
         rtxn: &'a heed::RoTxn,
         name: &str,
-    ) -> anyhow::Result<()>
-    {
+    ) -> anyhow::Result<()> {
         let mut key_size = 0u64;
         let mut val_size = 0u64;
         let mut values_length = Vec::new();
@@ -1028,27 +1064,27 @@ fn database_stats(index: &Index, rtxn: &heed::RoTxn, name: &str) -> anyhow::Resu
         WORD_DOCIDS => {
             let db = index.word_docids.as_polymorph();
             compute_stats::<RoaringBitmapCodec>(*db, rtxn, name)
-        },
+        }
         WORD_PREFIX_DOCIDS => {
             let db = index.word_prefix_docids.as_polymorph();
             compute_stats::<RoaringBitmapCodec>(*db, rtxn, name)
-        },
+        }
         DOCID_WORD_POSITIONS => {
             let db = index.docid_word_positions.as_polymorph();
             compute_stats::<BoRoaringBitmapCodec>(*db, rtxn, name)
-        },
+        }
         WORD_PAIR_PROXIMITY_DOCIDS => {
             let db = index.word_pair_proximity_docids.as_polymorph();
             compute_stats::<CboRoaringBitmapCodec>(*db, rtxn, name)
-        },
+        }
         WORD_PREFIX_PAIR_PROXIMITY_DOCIDS => {
             let db = index.word_prefix_pair_proximity_docids.as_polymorph();
             compute_stats::<CboRoaringBitmapCodec>(*db, rtxn, name)
-        },
+        }
         FIELD_ID_WORD_COUNT_DOCIDS => {
             let db = index.field_id_word_count_docids.as_polymorph();
             compute_stats::<CboRoaringBitmapCodec>(*db, rtxn, name)
-        },
+        }
         unknown => anyhow::bail!("unknown database {:?}", unknown),
     }
 }
@@ -1059,8 +1095,7 @@ fn word_pair_proximities_docids(
     debug: bool,
     word1: String,
     word2: String,
-) -> anyhow::Result<()>
-{
+) -> anyhow::Result<()> {
     use heed::types::ByteSlice;
     use milli::RoaringBitmapCodec;
 
@@ -1081,7 +1116,9 @@ fn word_pair_proximities_docids(
 
         // Skip keys that are longer than the requested one,
         // a longer key means that the second word is a prefix of the request word.
-        if key.len() != prefix.len() + 1 { continue; }
+        if key.len() != prefix.len() + 1 {
+            continue;
+        }
 
         let proximity = key.last().unwrap();
         let docids = if debug {
diff --git a/milli/src/criterion.rs b/milli/src/criterion.rs
index 931cf8588..cc1fca01f 100644
--- a/milli/src/criterion.rs
+++ b/milli/src/criterion.rs
@@ -1,15 +1,14 @@
 use std::fmt;
 use std::str::FromStr;
 
-use regex::Regex;
-use serde::{Serialize, Deserialize};
 use once_cell::sync::Lazy;
+use regex::Regex;
+use serde::{Deserialize, Serialize};
 
 use crate::error::{Error, UserError};
 
-static ASC_DESC_REGEX: Lazy<Regex> = Lazy::new(|| {
-    Regex::new(r#"(asc|desc)\(([\w_-]+)\)"#).unwrap()
-});
+static ASC_DESC_REGEX: Lazy<Regex> =
+    Lazy::new(|| Regex::new(r#"(asc|desc)\(([\w_-]+)\)"#).unwrap());
 
 #[derive(Debug, Serialize, Deserialize, Clone, PartialEq, Eq)]
 pub enum Criterion {
@@ -52,17 +51,21 @@ impl FromStr for Criterion {
             "attribute" => Ok(Criterion::Attribute),
             "exactness" => Ok(Criterion::Exactness),
             text => {
-                let caps = ASC_DESC_REGEX.captures(text).ok_or_else(|| {
-                    UserError::InvalidCriterionName { name: text.to_string() }
-                })?;
+                let caps = ASC_DESC_REGEX
+                    .captures(text)
+                    .ok_or_else(|| UserError::InvalidCriterionName { name: text.to_string() })?;
                 let order = caps.get(1).unwrap().as_str();
                 let field_name = caps.get(2).unwrap().as_str();
                 match order {
                     "asc" => Ok(Criterion::Asc(field_name.to_string())),
                     "desc" => Ok(Criterion::Desc(field_name.to_string())),
-                    text => return Err(UserError::InvalidCriterionName { name: text.to_string() }.into()),
+                    text => {
+                        return Err(
+                            UserError::InvalidCriterionName { name: text.to_string() }.into()
+                        )
+                    }
                 }
-            },
+            }
         }
     }
 }
@@ -82,13 +85,13 @@ impl fmt::Display for Criterion {
         use Criterion::*;
 
         match self {
-            Words           => f.write_str("words"),
-            Typo            => f.write_str("typo"),
-            Proximity       => f.write_str("proximity"),
-            Attribute       => f.write_str("attribute"),
-            Exactness       => f.write_str("exactness"),
-            Asc(attr)       => write!(f, "asc({})", attr),
-            Desc(attr)      => write!(f, "desc({})", attr),
+            Words => f.write_str("words"),
+            Typo => f.write_str("typo"),
+            Proximity => f.write_str("proximity"),
+            Attribute => f.write_str("attribute"),
+            Exactness => f.write_str("exactness"),
+            Asc(attr) => write!(f, "asc({})", attr),
+            Desc(attr) => write!(f, "desc({})", attr),
         }
     }
 }
diff --git a/milli/src/error.rs b/milli/src/error.rs
index 78a1b1c59..31012c690 100644
--- a/milli/src/error.rs
+++ b/milli/src/error.rs
@@ -2,7 +2,7 @@ use std::convert::Infallible;
 use std::error::Error as StdError;
 use std::{fmt, io, str};
 
-use heed::{MdbError, Error as HeedError};
+use heed::{Error as HeedError, MdbError};
 use rayon::ThreadPoolBuildError;
 use serde_json::{Map, Value};
 
@@ -80,14 +80,17 @@ impl From<fst::Error> for Error {
     }
 }
 
-impl<E> From<grenad::Error<E>> for Error where Error: From<E> {
+impl<E> From<grenad::Error<E>> for Error
+where
+    Error: From<E>,
+{
     fn from(error: grenad::Error<E>) -> Error {
         match error {
             grenad::Error::Io(error) => Error::IoError(error),
             grenad::Error::Merge(error) => Error::from(error),
             grenad::Error::InvalidCompressionType => {
                 Error::InternalError(InternalError::GrenadInvalidCompressionType)
-            },
+            }
         }
     }
 }
@@ -171,15 +174,15 @@ impl fmt::Display for InternalError {
         match self {
             Self::DatabaseMissingEntry { db_name, key } => {
                 write!(f, "missing {} in the {} database", key.unwrap_or("key"), db_name)
-            },
+            }
             Self::FieldIdMapMissingEntry(error) => error.fmt(f),
             Self::Fst(error) => error.fmt(f),
             Self::GrenadInvalidCompressionType => {
                 f.write_str("invalid compression type have been specified to grenad")
-            },
+            }
             Self::IndexingMergingKeys { process } => {
                 write!(f, "invalid merge while processing {}", process)
-            },
+            }
             Self::Serialization(error) => error.fmt(f),
             Self::InvalidDatabaseTyping => HeedError::InvalidDatabaseTyping.fmt(f),
             Self::RayonThreadPool(error) => error.fmt(f),
@@ -204,12 +207,12 @@ impl fmt::Display for UserError {
             Self::InvalidDocumentId { document_id } => {
                 let json = serde_json::to_string(document_id).unwrap();
                 write!(f, "document identifier is invalid {}", json)
-            },
+            }
             Self::InvalidFilterAttribute(error) => error.fmt(f),
             Self::MissingDocumentId { document } => {
                 let json = serde_json::to_string(document).unwrap();
                 write!(f, "document doesn't have an identifier {}", json)
-            },
+            }
             Self::MissingPrimaryKey => f.write_str("missing primary key"),
             Self::MaxDatabaseSizeReached => f.write_str("maximum database size reached"),
             // TODO where can we find it instead of writing the text ourselves?
@@ -217,14 +220,14 @@ impl fmt::Display for UserError {
             Self::InvalidStoreFile => f.write_str("store file is not a valid database file"),
             Self::PrimaryKeyCannotBeChanged => {
                 f.write_str("primary key cannot be changed if the database contains documents")
-            },
+            }
             Self::PrimaryKeyCannotBeReset => {
                 f.write_str("primary key cannot be reset if the database contains documents")
-            },
+            }
             Self::SerdeJson(error) => error.fmt(f),
             Self::UnknownInternalDocumentId { document_id } => {
                 write!(f, "an unknown internal document id have been used ({})", document_id)
-            },
+            }
         }
     }
 }
@@ -236,10 +239,10 @@ impl fmt::Display for FieldIdMapMissingEntry {
         match self {
             Self::FieldId { field_id, process } => {
                 write!(f, "unknown field id {} coming from the {} process", field_id, process)
-            },
+            }
             Self::FieldName { field_name, process } => {
                 write!(f, "unknown field name {} coming from the {} process", field_name, process)
-            },
+            }
         }
     }
 }
@@ -251,11 +254,11 @@ impl fmt::Display for SerializationError {
         match self {
             Self::Decoding { db_name: Some(name) } => {
                 write!(f, "decoding from the {} database failed", name)
-            },
+            }
             Self::Decoding { db_name: None } => f.write_str("decoding failed"),
             Self::Encoding { db_name: Some(name) } => {
                 write!(f, "encoding into the {} database failed", name)
-            },
+            }
             Self::Encoding { db_name: None } => f.write_str("encoding failed"),
             Self::InvalidNumberSerialization => f.write_str("number is not a valid finite number"),
         }
diff --git a/milli/src/external_documents_ids.rs b/milli/src/external_documents_ids.rs
index ee2a6c7bb..3dec638da 100644
--- a/milli/src/external_documents_ids.rs
+++ b/milli/src/external_documents_ids.rs
@@ -1,6 +1,7 @@
 use std::borrow::Cow;
 use std::convert::TryInto;
-use fst::{Streamer, IntoStreamer};
+
+use fst::{IntoStreamer, Streamer};
 
 pub struct ExternalDocumentsIds<'a> {
     pub(crate) hard: fst::Map<Cow<'a, [u8]>>,
@@ -8,7 +9,10 @@ pub struct ExternalDocumentsIds<'a> {
 }
 
 impl<'a> ExternalDocumentsIds<'a> {
-    pub fn new(hard: fst::Map<Cow<'a, [u8]>>, soft: fst::Map<Cow<'a, [u8]>>) -> ExternalDocumentsIds<'a> {
+    pub fn new(
+        hard: fst::Map<Cow<'a, [u8]>>,
+        soft: fst::Map<Cow<'a, [u8]>>,
+    ) -> ExternalDocumentsIds<'a> {
         ExternalDocumentsIds { hard, soft }
     }
 
@@ -29,7 +33,7 @@ impl<'a> ExternalDocumentsIds<'a> {
         match self.soft.get(external_id).or_else(|| self.hard.get(external_id)) {
             // u64 MAX means deleted in the soft fst map
             Some(id) if id != u64::MAX => Some(id.try_into().unwrap()),
-            _otherwise => None
+            _otherwise => None,
         }
     }
 
diff --git a/milli/src/facet/facet_type.rs b/milli/src/facet/facet_type.rs
index 09f29bc00..51dd448e2 100644
--- a/milli/src/facet/facet_type.rs
+++ b/milli/src/facet/facet_type.rs
@@ -2,10 +2,9 @@ use std::error::Error;
 use std::fmt;
 use std::str::FromStr;
 
-use serde::{Serialize, Deserialize};
+use serde::{Deserialize, Serialize};
 
-#[derive(Debug, Copy, Clone, PartialOrd, Ord, PartialEq, Eq, Hash)]
-#[derive(Serialize, Deserialize)]
+#[derive(Debug, Copy, Clone, PartialOrd, Ord, PartialEq, Eq, Hash, Serialize, Deserialize)]
 pub enum FacetType {
     String,
     Number,
@@ -43,4 +42,4 @@ impl fmt::Display for InvalidFacetType {
     }
 }
 
-impl Error for InvalidFacetType { }
+impl Error for InvalidFacetType {}
diff --git a/milli/src/facet/facet_value.rs b/milli/src/facet/facet_value.rs
index 99455fa27..eb7fb3c5e 100644
--- a/milli/src/facet/facet_value.rs
+++ b/milli/src/facet/facet_value.rs
@@ -50,7 +50,7 @@ impl Serialize for FacetValue {
             FacetValue::Number(number) => {
                 let string = number.to_string();
                 serializer.serialize_str(&string)
-            },
+            }
         }
     }
 }
diff --git a/milli/src/facet/value_encoding.rs b/milli/src/facet/value_encoding.rs
index 7259243e5..31c00bd2d 100644
--- a/milli/src/facet/value_encoding.rs
+++ b/milli/src/facet/value_encoding.rs
@@ -28,6 +28,7 @@ fn xor_all_bits(mut x: [u8; 8]) -> [u8; 8] {
 #[cfg(test)]
 mod tests {
     use std::cmp::Ordering::Less;
+
     use super::*;
 
     fn is_sorted<T: Ord>(x: &[T]) -> bool {
@@ -39,8 +40,8 @@ mod tests {
         let a = -13_f64;
         let b = -10.0;
         let c = -0.0;
-        let d =  1.0;
-        let e =  43.0;
+        let d = 1.0;
+        let e = 43.0;
 
         let vec: Vec<_> = [a, b, c, d, e].iter().cloned().map(f64_into_bytes).collect();
         assert!(is_sorted(&vec), "{:?}", vec);
diff --git a/milli/src/fields_ids_map.rs b/milli/src/fields_ids_map.rs
index 76ff2d281..b0a084c3c 100644
--- a/milli/src/fields_ids_map.rs
+++ b/milli/src/fields_ids_map.rs
@@ -1,5 +1,7 @@
 use std::collections::BTreeMap;
-use serde::{Serialize, Deserialize};
+
+use serde::{Deserialize, Serialize};
+
 use crate::FieldId;
 
 #[derive(Debug, Clone, Serialize, Deserialize)]
@@ -11,11 +13,7 @@ pub struct FieldsIdsMap {
 
 impl FieldsIdsMap {
     pub fn new() -> FieldsIdsMap {
-        FieldsIdsMap {
-            names_ids: BTreeMap::new(),
-            ids_names: BTreeMap::new(),
-            next_id: Some(0),
-        }
+        FieldsIdsMap { names_ids: BTreeMap::new(), ids_names: BTreeMap::new(), next_id: Some(0) }
     }
 
     /// Returns the number of fields ids in the map.
@@ -62,17 +60,17 @@ impl FieldsIdsMap {
     }
 
     /// Iterate over the ids and names in the ids order.
-    pub fn iter(&self) -> impl Iterator<Item=(FieldId, &str)> {
+    pub fn iter(&self) -> impl Iterator<Item = (FieldId, &str)> {
         self.ids_names.iter().map(|(id, name)| (*id, name.as_str()))
     }
 
     /// Iterate over the ids in the order of the ids.
-    pub fn ids<'a>(&'a self) -> impl Iterator<Item=FieldId> + 'a {
+    pub fn ids<'a>(&'a self) -> impl Iterator<Item = FieldId> + 'a {
         self.ids_names.keys().copied()
     }
 
     /// Iterate over the names in the order of the ids.
-    pub fn names(&self) -> impl Iterator<Item=&str> {
+    pub fn names(&self) -> impl Iterator<Item = &str> {
         self.ids_names.values().map(AsRef::as_ref)
     }
 }
diff --git a/milli/src/heed_codec/facet/facet_level_value_f64_codec.rs b/milli/src/heed_codec/facet/facet_level_value_f64_codec.rs
index a4642f961..b23dcb269 100644
--- a/milli/src/heed_codec/facet/facet_level_value_f64_codec.rs
+++ b/milli/src/heed_codec/facet/facet_level_value_f64_codec.rs
@@ -71,7 +71,8 @@ impl heed::BytesEncode<'_> for FacetLevelValueF64Codec {
 
 #[cfg(test)]
 mod tests {
-    use heed::{BytesEncode, BytesDecode};
+    use heed::{BytesDecode, BytesEncode};
+
     use super::*;
 
     #[test]
diff --git a/milli/src/heed_codec/facet/field_doc_id_facet_f64_codec.rs b/milli/src/heed_codec/facet/field_doc_id_facet_f64_codec.rs
index e9b5abeb8..b3c0fa381 100644
--- a/milli/src/heed_codec/facet/field_doc_id_facet_f64_codec.rs
+++ b/milli/src/heed_codec/facet/field_doc_id_facet_f64_codec.rs
@@ -1,8 +1,8 @@
 use std::borrow::Cow;
 use std::convert::TryInto;
 
-use crate::{FieldId, DocumentId};
 use crate::facet::value_encoding::f64_into_bytes;
+use crate::{DocumentId, FieldId};
 
 pub struct FieldDocIdFacetF64Codec;
 
diff --git a/milli/src/heed_codec/facet/field_doc_id_facet_string_codec.rs b/milli/src/heed_codec/facet/field_doc_id_facet_string_codec.rs
index b002346e9..fd3f1143d 100644
--- a/milli/src/heed_codec/facet/field_doc_id_facet_string_codec.rs
+++ b/milli/src/heed_codec/facet/field_doc_id_facet_string_codec.rs
@@ -2,12 +2,17 @@ use std::borrow::Cow;
 use std::convert::TryInto;
 use std::str;
 
-use crate::{FieldId, DocumentId};
+use crate::{DocumentId, FieldId};
 
 pub struct FieldDocIdFacetStringCodec;
 
 impl FieldDocIdFacetStringCodec {
-    pub fn serialize_into(field_id: FieldId, document_id: DocumentId, value: &str, out: &mut Vec<u8>) {
+    pub fn serialize_into(
+        field_id: FieldId,
+        document_id: DocumentId,
+        value: &str,
+        out: &mut Vec<u8>,
+    ) {
         out.reserve(1 + 4 + value.len());
         out.push(field_id);
         out.extend_from_slice(&document_id.to_be_bytes());
diff --git a/milli/src/heed_codec/field_id_word_count_codec.rs b/milli/src/heed_codec/field_id_word_count_codec.rs
index 5796e5020..64f0e1db6 100644
--- a/milli/src/heed_codec/field_id_word_count_codec.rs
+++ b/milli/src/heed_codec/field_id_word_count_codec.rs
@@ -1,4 +1,5 @@
-use std::{borrow::Cow, convert::TryInto};
+use std::borrow::Cow;
+use std::convert::TryInto;
 
 use crate::FieldId;
 
diff --git a/milli/src/heed_codec/mod.rs b/milli/src/heed_codec/mod.rs
index 65a06573e..7bd7dff2d 100644
--- a/milli/src/heed_codec/mod.rs
+++ b/milli/src/heed_codec/mod.rs
@@ -1,16 +1,18 @@
 mod beu32_str_codec;
+pub mod facet;
+mod field_id_word_count_codec;
 mod obkv_codec;
 mod roaring_bitmap;
 mod roaring_bitmap_length;
 mod str_level_position_codec;
 mod str_str_u8_codec;
-mod field_id_word_count_codec;
-pub mod facet;
 
 pub use self::beu32_str_codec::BEU32StrCodec;
+pub use self::field_id_word_count_codec::FieldIdWordCountCodec;
 pub use self::obkv_codec::ObkvCodec;
 pub use self::roaring_bitmap::{BoRoaringBitmapCodec, CboRoaringBitmapCodec, RoaringBitmapCodec};
-pub use self::roaring_bitmap_length::{BoRoaringBitmapLenCodec, CboRoaringBitmapLenCodec, RoaringBitmapLenCodec};
+pub use self::roaring_bitmap_length::{
+    BoRoaringBitmapLenCodec, CboRoaringBitmapLenCodec, RoaringBitmapLenCodec,
+};
 pub use self::str_level_position_codec::StrLevelPositionCodec;
 pub use self::str_str_u8_codec::StrStrU8Codec;
-pub use self::field_id_word_count_codec::FieldIdWordCountCodec;
diff --git a/milli/src/heed_codec/obkv_codec.rs b/milli/src/heed_codec/obkv_codec.rs
index 94a230e05..b7414b693 100644
--- a/milli/src/heed_codec/obkv_codec.rs
+++ b/milli/src/heed_codec/obkv_codec.rs
@@ -1,4 +1,5 @@
 use std::borrow::Cow;
+
 use obkv::{KvReader, KvWriter};
 
 pub struct ObkvCodec;
diff --git a/milli/src/heed_codec/roaring_bitmap/cbo_roaring_bitmap_codec.rs b/milli/src/heed_codec/roaring_bitmap/cbo_roaring_bitmap_codec.rs
index 325effa73..53f64d648 100644
--- a/milli/src/heed_codec/roaring_bitmap/cbo_roaring_bitmap_codec.rs
+++ b/milli/src/heed_codec/roaring_bitmap/cbo_roaring_bitmap_codec.rs
@@ -75,7 +75,9 @@ impl heed::BytesEncode<'_> for CboRoaringBitmapCodec {
 #[cfg(test)]
 mod tests {
     use std::iter::FromIterator;
-    use heed::{BytesEncode, BytesDecode};
+
+    use heed::{BytesDecode, BytesEncode};
+
     use super::*;
 
     #[test]
diff --git a/milli/src/heed_codec/roaring_bitmap/roaring_bitmap_codec.rs b/milli/src/heed_codec/roaring_bitmap/roaring_bitmap_codec.rs
index 755296704..8fae9b8fd 100644
--- a/milli/src/heed_codec/roaring_bitmap/roaring_bitmap_codec.rs
+++ b/milli/src/heed_codec/roaring_bitmap/roaring_bitmap_codec.rs
@@ -1,4 +1,5 @@
 use std::borrow::Cow;
+
 use roaring::RoaringBitmap;
 
 pub struct RoaringBitmapCodec;
diff --git a/milli/src/heed_codec/roaring_bitmap_length/roaring_bitmap_len_codec.rs b/milli/src/heed_codec/roaring_bitmap_length/roaring_bitmap_len_codec.rs
index 042b5cf6b..4d266e413 100644
--- a/milli/src/heed_codec/roaring_bitmap_length/roaring_bitmap_len_codec.rs
+++ b/milli/src/heed_codec/roaring_bitmap_length/roaring_bitmap_len_codec.rs
@@ -1,7 +1,7 @@
-use std::io::{self, Read, BufRead};
+use std::io::{self, BufRead, Read};
 use std::mem;
 
-use byteorder::{ReadBytesExt, LittleEndian};
+use byteorder::{LittleEndian, ReadBytesExt};
 
 const SERIAL_COOKIE_NO_RUNCONTAINER: u32 = 12346;
 const SERIAL_COOKIE: u16 = 12347;
@@ -16,20 +16,14 @@ impl RoaringBitmapLenCodec {
             if cookie == SERIAL_COOKIE_NO_RUNCONTAINER {
                 (bytes.read_u32::<LittleEndian>()? as usize, true)
             } else if (cookie as u16) == SERIAL_COOKIE {
-                return Err(io::Error::new(
-                    io::ErrorKind::Other,
-                    "run containers are unsupported",
-                ));
+                return Err(io::Error::new(io::ErrorKind::Other, "run containers are unsupported"));
             } else {
                 return Err(io::Error::new(io::ErrorKind::Other, "unknown cookie value"));
             }
         };
 
         if size > u16::max_value() as usize + 1 {
-            return Err(io::Error::new(
-                io::ErrorKind::Other,
-                "size is greater than supported",
-            ));
+            return Err(io::Error::new(io::ErrorKind::Other, "size is greater than supported"));
         }
 
         let mut description_bytes = vec![0u8; size * 4];
@@ -67,12 +61,12 @@ impl heed::BytesDecode<'_> for RoaringBitmapLenCodec {
 
 #[cfg(test)]
 mod tests {
-    use super::*;
-
-    use crate::heed_codec::RoaringBitmapCodec;
     use heed::BytesEncode;
     use roaring::RoaringBitmap;
 
+    use super::*;
+    use crate::heed_codec::RoaringBitmapCodec;
+
     #[test]
     fn deserialize_roaring_bitmap_length() {
         let bitmap: RoaringBitmap = (0..500).chain(800..800_000).chain(920_056..930_032).collect();
diff --git a/milli/src/heed_codec/str_level_position_codec.rs b/milli/src/heed_codec/str_level_position_codec.rs
index 810e91940..5be45bbeb 100644
--- a/milli/src/heed_codec/str_level_position_codec.rs
+++ b/milli/src/heed_codec/str_level_position_codec.rs
@@ -13,7 +13,9 @@ impl<'a> heed::BytesDecode<'a> for StrLevelPositionCodec {
     fn bytes_decode(bytes: &'a [u8]) -> Option<Self::DItem> {
         let footer_len = size_of::<u8>() + size_of::<u32>() * 2;
 
-        if bytes.len() < footer_len { return None }
+        if bytes.len() < footer_len {
+            return None;
+        }
 
         let (word, bytes) = bytes.split_at(bytes.len() - footer_len);
         let word = str::from_utf8(word).ok()?;
diff --git a/milli/src/index.rs b/milli/src/index.rs
index bf4b3e023..a6c09f3d3 100644
--- a/milli/src/index.rs
+++ b/milli/src/index.rs
@@ -3,23 +3,22 @@ use std::collections::{HashMap, HashSet};
 use std::path::Path;
 
 use chrono::{DateTime, Utc};
-use heed::{Database, PolyDatabase, RoTxn, RwTxn};
 use heed::types::*;
+use heed::{Database, PolyDatabase, RoTxn, RwTxn};
 use roaring::RoaringBitmap;
 
-use crate::error::{UserError, FieldIdMapMissingEntry, InternalError};
-use crate::{Criterion, default_criteria, FacetDistribution, FieldsDistribution, Search};
-use crate::{BEU32, DocumentId, ExternalDocumentsIds, FieldId, Result};
-use crate::{
-    BEU32StrCodec, BoRoaringBitmapCodec, CboRoaringBitmapCodec,
-    ObkvCodec, RoaringBitmapCodec, RoaringBitmapLenCodec, StrLevelPositionCodec, StrStrU8Codec,
-    FieldIdWordCountCodec,
-};
-use crate::heed_codec::facet::{
-    FieldDocIdFacetF64Codec, FieldDocIdFacetStringCodec,
-    FacetValueStringCodec, FacetLevelValueF64Codec,
-};
+use crate::error::{FieldIdMapMissingEntry, InternalError, UserError};
 use crate::fields_ids_map::FieldsIdsMap;
+use crate::heed_codec::facet::{
+    FacetLevelValueF64Codec, FacetValueStringCodec, FieldDocIdFacetF64Codec,
+    FieldDocIdFacetStringCodec,
+};
+use crate::{
+    default_criteria, BEU32StrCodec, BoRoaringBitmapCodec, CboRoaringBitmapCodec, Criterion,
+    DocumentId, ExternalDocumentsIds, FacetDistribution, FieldId, FieldIdWordCountCodec,
+    FieldsDistribution, ObkvCodec, Result, RoaringBitmapCodec, RoaringBitmapLenCodec, Search,
+    StrLevelPositionCodec, StrStrU8Codec, BEU32,
+};
 
 pub mod main_key {
     pub const CRITERIA_KEY: &str = "criteria";
@@ -114,14 +113,17 @@ impl Index {
         let word_prefix_docids = env.create_database(Some(WORD_PREFIX_DOCIDS))?;
         let docid_word_positions = env.create_database(Some(DOCID_WORD_POSITIONS))?;
         let word_pair_proximity_docids = env.create_database(Some(WORD_PAIR_PROXIMITY_DOCIDS))?;
-        let word_prefix_pair_proximity_docids = env.create_database(Some(WORD_PREFIX_PAIR_PROXIMITY_DOCIDS))?;
+        let word_prefix_pair_proximity_docids =
+            env.create_database(Some(WORD_PREFIX_PAIR_PROXIMITY_DOCIDS))?;
         let word_level_position_docids = env.create_database(Some(WORD_LEVEL_POSITION_DOCIDS))?;
         let field_id_word_count_docids = env.create_database(Some(FIELD_ID_WORD_COUNT_DOCIDS))?;
-        let word_prefix_level_position_docids = env.create_database(Some(WORD_PREFIX_LEVEL_POSITION_DOCIDS))?;
+        let word_prefix_level_position_docids =
+            env.create_database(Some(WORD_PREFIX_LEVEL_POSITION_DOCIDS))?;
         let facet_id_f64_docids = env.create_database(Some(FACET_ID_F64_DOCIDS))?;
         let facet_id_string_docids = env.create_database(Some(FACET_ID_STRING_DOCIDS))?;
         let field_id_docid_facet_f64s = env.create_database(Some(FIELD_ID_DOCID_FACET_F64S))?;
-        let field_id_docid_facet_strings = env.create_database(Some(FIELD_ID_DOCID_FACET_STRINGS))?;
+        let field_id_docid_facet_strings =
+            env.create_database(Some(FIELD_ID_DOCID_FACET_STRINGS))?;
         let documents = env.create_database(Some(DOCUMENTS))?;
 
         Index::initialize_creation_dates(&env, main)?;
@@ -184,18 +186,26 @@ impl Index {
     /* documents ids */
 
     /// Writes the documents ids that corresponds to the user-ids-documents-ids FST.
-    pub(crate) fn put_documents_ids(&self, wtxn: &mut RwTxn, docids: &RoaringBitmap) -> heed::Result<()> {
+    pub(crate) fn put_documents_ids(
+        &self,
+        wtxn: &mut RwTxn,
+        docids: &RoaringBitmap,
+    ) -> heed::Result<()> {
         self.main.put::<_, Str, RoaringBitmapCodec>(wtxn, main_key::DOCUMENTS_IDS_KEY, docids)
     }
 
     /// Returns the internal documents ids.
     pub fn documents_ids(&self, rtxn: &RoTxn) -> heed::Result<RoaringBitmap> {
-        Ok(self.main.get::<_, Str, RoaringBitmapCodec>(rtxn, main_key::DOCUMENTS_IDS_KEY)?.unwrap_or_default())
+        Ok(self
+            .main
+            .get::<_, Str, RoaringBitmapCodec>(rtxn, main_key::DOCUMENTS_IDS_KEY)?
+            .unwrap_or_default())
     }
 
     /// Returns the number of documents indexed in the database.
     pub fn number_of_documents(&self, rtxn: &RoTxn) -> Result<u64> {
-        let count = self.main.get::<_, Str, RoaringBitmapLenCodec>(rtxn, main_key::DOCUMENTS_IDS_KEY)?;
+        let count =
+            self.main.get::<_, Str, RoaringBitmapLenCodec>(rtxn, main_key::DOCUMENTS_IDS_KEY)?;
         Ok(count.unwrap_or_default())
     }
 
@@ -224,21 +234,30 @@ impl Index {
         &self,
         wtxn: &mut RwTxn,
         external_documents_ids: &ExternalDocumentsIds<'a>,
-    ) -> heed::Result<()>
-    {
+    ) -> heed::Result<()> {
         let ExternalDocumentsIds { hard, soft } = external_documents_ids;
         let hard = hard.as_fst().as_bytes();
         let soft = soft.as_fst().as_bytes();
-        self.main.put::<_, Str, ByteSlice>(wtxn, main_key::HARD_EXTERNAL_DOCUMENTS_IDS_KEY, hard)?;
-        self.main.put::<_, Str, ByteSlice>(wtxn, main_key::SOFT_EXTERNAL_DOCUMENTS_IDS_KEY, soft)?;
+        self.main.put::<_, Str, ByteSlice>(
+            wtxn,
+            main_key::HARD_EXTERNAL_DOCUMENTS_IDS_KEY,
+            hard,
+        )?;
+        self.main.put::<_, Str, ByteSlice>(
+            wtxn,
+            main_key::SOFT_EXTERNAL_DOCUMENTS_IDS_KEY,
+            soft,
+        )?;
         Ok(())
     }
 
     /// Returns the external documents ids map which associate the external ids
     /// with the internal ids (i.e. `u32`).
     pub fn external_documents_ids<'t>(&self, rtxn: &'t RoTxn) -> Result<ExternalDocumentsIds<'t>> {
-        let hard = self.main.get::<_, Str, ByteSlice>(rtxn, main_key::HARD_EXTERNAL_DOCUMENTS_IDS_KEY)?;
-        let soft = self.main.get::<_, Str, ByteSlice>(rtxn, main_key::SOFT_EXTERNAL_DOCUMENTS_IDS_KEY)?;
+        let hard =
+            self.main.get::<_, Str, ByteSlice>(rtxn, main_key::HARD_EXTERNAL_DOCUMENTS_IDS_KEY)?;
+        let soft =
+            self.main.get::<_, Str, ByteSlice>(rtxn, main_key::SOFT_EXTERNAL_DOCUMENTS_IDS_KEY)?;
         let hard = match hard {
             Some(hard) => fst::Map::new(hard)?.map_data(Cow::Borrowed)?,
             None => fst::Map::default().map_data(Cow::Owned)?,
@@ -254,42 +273,62 @@ impl Index {
 
     /// Writes the fields ids map which associate the documents keys with an internal field id
     /// (i.e. `u8`), this field id is used to identify fields in the obkv documents.
-    pub(crate) fn put_fields_ids_map(&self, wtxn: &mut RwTxn, map: &FieldsIdsMap) -> heed::Result<()> {
+    pub(crate) fn put_fields_ids_map(
+        &self,
+        wtxn: &mut RwTxn,
+        map: &FieldsIdsMap,
+    ) -> heed::Result<()> {
         self.main.put::<_, Str, SerdeJson<FieldsIdsMap>>(wtxn, main_key::FIELDS_IDS_MAP_KEY, map)
     }
 
     /// Returns the fields ids map which associate the documents keys with an internal field id
     /// (i.e. `u8`), this field id is used to identify fields in the obkv documents.
     pub fn fields_ids_map(&self, rtxn: &RoTxn) -> heed::Result<FieldsIdsMap> {
-        Ok(self.main.get::<_, Str, SerdeJson<FieldsIdsMap>>(
-            rtxn,
-            main_key::FIELDS_IDS_MAP_KEY,
-        )?.unwrap_or_default())
+        Ok(self
+            .main
+            .get::<_, Str, SerdeJson<FieldsIdsMap>>(rtxn, main_key::FIELDS_IDS_MAP_KEY)?
+            .unwrap_or_default())
     }
 
     /* fields distribution */
 
     /// Writes the fields distribution which associates every field name with
     /// the number of times it occurs in the documents.
-    pub(crate) fn put_fields_distribution(&self, wtxn: &mut RwTxn, distribution: &FieldsDistribution) -> heed::Result<()> {
-        self.main.put::<_, Str, SerdeJson<FieldsDistribution>>(wtxn, main_key::FIELDS_DISTRIBUTION_KEY, distribution)
+    pub(crate) fn put_fields_distribution(
+        &self,
+        wtxn: &mut RwTxn,
+        distribution: &FieldsDistribution,
+    ) -> heed::Result<()> {
+        self.main.put::<_, Str, SerdeJson<FieldsDistribution>>(
+            wtxn,
+            main_key::FIELDS_DISTRIBUTION_KEY,
+            distribution,
+        )
     }
 
     /// Returns the fields distribution which associates every field name with
     /// the number of times it occurs in the documents.
     pub fn fields_distribution(&self, rtxn: &RoTxn) -> heed::Result<FieldsDistribution> {
-        Ok(self.main.get::<_, Str, SerdeJson<FieldsDistribution>>(
-            rtxn,
-            main_key::FIELDS_DISTRIBUTION_KEY,
-        )?.unwrap_or_default())
+        Ok(self
+            .main
+            .get::<_, Str, SerdeJson<FieldsDistribution>>(rtxn, main_key::FIELDS_DISTRIBUTION_KEY)?
+            .unwrap_or_default())
     }
 
     /* displayed fields */
 
     /// Writes the fields that must be displayed in the defined order.
     /// There must be not be any duplicate field id.
-    pub(crate) fn put_displayed_fields(&self, wtxn: &mut RwTxn, fields: &[&str]) -> heed::Result<()> {
-        self.main.put::<_, Str, SerdeBincode<&[&str]>>(wtxn, main_key::DISPLAYED_FIELDS_KEY, &fields)
+    pub(crate) fn put_displayed_fields(
+        &self,
+        wtxn: &mut RwTxn,
+        fields: &[&str],
+    ) -> heed::Result<()> {
+        self.main.put::<_, Str, SerdeBincode<&[&str]>>(
+            wtxn,
+            main_key::DISPLAYED_FIELDS_KEY,
+            &fields,
+        )
     }
 
     /// Deletes the displayed fields ids, this will make the engine to display
@@ -313,14 +352,17 @@ impl Index {
                 for name in fields.into_iter() {
                     match fields_ids_map.id(name) {
                         Some(field_id) => fields_ids.push(field_id),
-                        None => return Err(FieldIdMapMissingEntry::FieldName {
-                            field_name: name.to_string(),
-                            process: "Index::displayed_fields_ids",
-                        }.into()),
+                        None => {
+                            return Err(FieldIdMapMissingEntry::FieldName {
+                                field_name: name.to_string(),
+                                process: "Index::displayed_fields_ids",
+                            }
+                            .into())
+                        }
                     }
                 }
                 Ok(Some(fields_ids))
-            },
+            }
             None => Ok(None),
         }
     }
@@ -328,8 +370,16 @@ impl Index {
     /* searchable fields */
 
     /// Writes the searchable fields, when this list is specified, only these are indexed.
-    pub(crate) fn put_searchable_fields(&self, wtxn: &mut RwTxn, fields: &[&str]) -> heed::Result<()> {
-        self.main.put::<_, Str, SerdeBincode<&[&str]>>(wtxn, main_key::SEARCHABLE_FIELDS_KEY, &fields)
+    pub(crate) fn put_searchable_fields(
+        &self,
+        wtxn: &mut RwTxn,
+        fields: &[&str],
+    ) -> heed::Result<()> {
+        self.main.put::<_, Str, SerdeBincode<&[&str]>>(
+            wtxn,
+            main_key::SEARCHABLE_FIELDS_KEY,
+            &fields,
+        )
     }
 
     /// Deletes the searchable fields, when no fields are specified, all fields are indexed.
@@ -352,14 +402,17 @@ impl Index {
                 for name in fields {
                     match fields_ids_map.id(name) {
                         Some(field_id) => fields_ids.push(field_id),
-                        None => return Err(FieldIdMapMissingEntry::FieldName {
-                            field_name: name.to_string(),
-                            process: "Index::searchable_fields_ids",
-                        }.into()),
+                        None => {
+                            return Err(FieldIdMapMissingEntry::FieldName {
+                                field_name: name.to_string(),
+                                process: "Index::searchable_fields_ids",
+                            }
+                            .into())
+                        }
                     }
                 }
                 Ok(Some(fields_ids))
-            },
+            }
             None => Ok(None),
         }
     }
@@ -367,7 +420,11 @@ impl Index {
     /* filterable fields */
 
     /// Writes the filterable fields names in the database.
-    pub(crate) fn put_filterable_fields(&self, wtxn: &mut RwTxn, fields: &HashSet<String>) -> heed::Result<()> {
+    pub(crate) fn put_filterable_fields(
+        &self,
+        wtxn: &mut RwTxn,
+        fields: &HashSet<String>,
+    ) -> heed::Result<()> {
         self.main.put::<_, Str, SerdeJson<_>>(wtxn, main_key::FILTERABLE_FIELDS_KEY, fields)
     }
 
@@ -378,10 +435,10 @@ impl Index {
 
     /// Returns the filterable fields names.
     pub fn filterable_fields(&self, rtxn: &RoTxn) -> heed::Result<HashSet<String>> {
-        Ok(self.main.get::<_, Str, SerdeJson<_>>(
-            rtxn,
-            main_key::FILTERABLE_FIELDS_KEY,
-        )?.unwrap_or_default())
+        Ok(self
+            .main
+            .get::<_, Str, SerdeJson<_>>(rtxn, main_key::FILTERABLE_FIELDS_KEY)?
+            .unwrap_or_default())
     }
 
     /// Identical to `filterable_fields`, but returns ids instead.
@@ -394,11 +451,14 @@ impl Index {
             match fields_ids_map.id(&name) {
                 Some(field_id) => {
                     fields_ids.insert(field_id);
-                },
-                None => return Err(FieldIdMapMissingEntry::FieldName {
-                    field_name: name,
-                    process: "Index::filterable_fields_ids",
-                }.into()),
+                }
+                None => {
+                    return Err(FieldIdMapMissingEntry::FieldName {
+                        field_name: name,
+                        process: "Index::filterable_fields_ids",
+                    }
+                    .into())
+                }
             }
         }
 
@@ -413,9 +473,8 @@ impl Index {
     pub fn faceted_fields(&self, rtxn: &RoTxn) -> Result<HashSet<String>> {
         let filterable_fields = self.filterable_fields(rtxn)?;
         let distinct_field = self.distinct_field(rtxn)?;
-        let asc_desc_fields = self.criteria(rtxn)?
-            .into_iter()
-            .filter_map(|criterion| match criterion {
+        let asc_desc_fields =
+            self.criteria(rtxn)?.into_iter().filter_map(|criterion| match criterion {
                 Criterion::Asc(field) | Criterion::Desc(field) => Some(field),
                 _otherwise => None,
             });
@@ -439,11 +498,14 @@ impl Index {
             match fields_ids_map.id(&name) {
                 Some(field_id) => {
                     fields_ids.insert(field_id);
-                },
-                None => return Err(FieldIdMapMissingEntry::FieldName {
-                    field_name: name,
-                    process: "Index::faceted_fields_ids",
-                }.into()),
+                }
+                None => {
+                    return Err(FieldIdMapMissingEntry::FieldName {
+                        field_name: name,
+                        process: "Index::faceted_fields_ids",
+                    }
+                    .into())
+                }
             }
         }
 
@@ -458,8 +520,7 @@ impl Index {
         wtxn: &mut RwTxn,
         field_id: FieldId,
         docids: &RoaringBitmap,
-    ) -> heed::Result<()>
-    {
+    ) -> heed::Result<()> {
         let mut buffer = [0u8; main_key::STRING_FACETED_DOCUMENTS_IDS_PREFIX.len() + 1];
         buffer[..main_key::STRING_FACETED_DOCUMENTS_IDS_PREFIX.len()]
             .copy_from_slice(main_key::STRING_FACETED_DOCUMENTS_IDS_PREFIX.as_bytes());
@@ -472,8 +533,7 @@ impl Index {
         &self,
         rtxn: &RoTxn,
         field_id: FieldId,
-    ) -> heed::Result<RoaringBitmap>
-    {
+    ) -> heed::Result<RoaringBitmap> {
         let mut buffer = [0u8; main_key::STRING_FACETED_DOCUMENTS_IDS_PREFIX.len() + 1];
         buffer[..main_key::STRING_FACETED_DOCUMENTS_IDS_PREFIX.len()]
             .copy_from_slice(main_key::STRING_FACETED_DOCUMENTS_IDS_PREFIX.as_bytes());
@@ -490,8 +550,7 @@ impl Index {
         wtxn: &mut RwTxn,
         field_id: FieldId,
         docids: &RoaringBitmap,
-    ) -> heed::Result<()>
-    {
+    ) -> heed::Result<()> {
         let mut buffer = [0u8; main_key::NUMBER_FACETED_DOCUMENTS_IDS_PREFIX.len() + 1];
         buffer[..main_key::NUMBER_FACETED_DOCUMENTS_IDS_PREFIX.len()]
             .copy_from_slice(main_key::NUMBER_FACETED_DOCUMENTS_IDS_PREFIX.as_bytes());
@@ -504,8 +563,7 @@ impl Index {
         &self,
         rtxn: &RoTxn,
         field_id: FieldId,
-    ) -> heed::Result<RoaringBitmap>
-    {
+    ) -> heed::Result<RoaringBitmap> {
         let mut buffer = [0u8; main_key::NUMBER_FACETED_DOCUMENTS_IDS_PREFIX.len() + 1];
         buffer[..main_key::NUMBER_FACETED_DOCUMENTS_IDS_PREFIX.len()]
             .copy_from_slice(main_key::NUMBER_FACETED_DOCUMENTS_IDS_PREFIX.as_bytes());
@@ -518,7 +576,11 @@ impl Index {
 
     /* distinct field */
 
-    pub(crate) fn put_distinct_field(&self, wtxn: &mut RwTxn, distinct_field: &str) -> heed::Result<()> {
+    pub(crate) fn put_distinct_field(
+        &self,
+        wtxn: &mut RwTxn,
+        distinct_field: &str,
+    ) -> heed::Result<()> {
         self.main.put::<_, Str, Str>(wtxn, main_key::DISTINCT_FIELD_KEY, distinct_field)
     }
 
@@ -532,7 +594,11 @@ impl Index {
 
     /* criteria */
 
-    pub(crate) fn put_criteria(&self, wtxn: &mut RwTxn, criteria: &[Criterion]) -> heed::Result<()> {
+    pub(crate) fn put_criteria(
+        &self,
+        wtxn: &mut RwTxn,
+        criteria: &[Criterion],
+    ) -> heed::Result<()> {
         self.main.put::<_, Str, SerdeJson<&[Criterion]>>(wtxn, main_key::CRITERIA_KEY, &criteria)
     }
 
@@ -550,7 +616,11 @@ impl Index {
     /* words fst */
 
     /// Writes the FST which is the words dictionary of the engine.
-    pub(crate) fn put_words_fst<A: AsRef<[u8]>>(&self, wtxn: &mut RwTxn, fst: &fst::Set<A>) -> heed::Result<()> {
+    pub(crate) fn put_words_fst<A: AsRef<[u8]>>(
+        &self,
+        wtxn: &mut RwTxn,
+        fst: &fst::Set<A>,
+    ) -> heed::Result<()> {
         self.main.put::<_, Str, ByteSlice>(wtxn, main_key::WORDS_FST_KEY, fst.as_fst().as_bytes())
     }
 
@@ -564,7 +634,11 @@ impl Index {
 
     /* stop words */
 
-    pub(crate) fn put_stop_words<A: AsRef<[u8]>>(&self, wtxn: &mut RwTxn, fst: &fst::Set<A>) -> heed::Result<()> {
+    pub(crate) fn put_stop_words<A: AsRef<[u8]>>(
+        &self,
+        wtxn: &mut RwTxn,
+        fst: &fst::Set<A>,
+    ) -> heed::Result<()> {
         self.main.put::<_, Str, ByteSlice>(wtxn, main_key::STOP_WORDS_KEY, fst.as_fst().as_bytes())
     }
 
@@ -585,8 +659,7 @@ impl Index {
         &self,
         wtxn: &mut RwTxn,
         synonyms: &HashMap<Vec<String>, Vec<Vec<String>>>,
-    ) -> heed::Result<()>
-    {
+    ) -> heed::Result<()> {
         self.main.put::<_, Str, SerdeBincode<_>>(wtxn, main_key::SYNONYMS_KEY, synonyms)
     }
 
@@ -595,15 +668,17 @@ impl Index {
     }
 
     pub fn synonyms(&self, rtxn: &RoTxn) -> heed::Result<HashMap<Vec<String>, Vec<Vec<String>>>> {
-        Ok(self.main.get::<_, Str, SerdeBincode<_>>(rtxn, main_key::SYNONYMS_KEY)?.unwrap_or_default())
+        Ok(self
+            .main
+            .get::<_, Str, SerdeBincode<_>>(rtxn, main_key::SYNONYMS_KEY)?
+            .unwrap_or_default())
     }
 
     pub fn words_synonyms<S: AsRef<str>>(
         &self,
         rtxn: &RoTxn,
         words: &[S],
-    ) -> heed::Result<Option<Vec<Vec<String>>>>
-    {
+    ) -> heed::Result<Option<Vec<Vec<String>>>> {
         let words: Vec<_> = words.iter().map(|s| s.as_ref().to_owned()).collect();
         Ok(self.synonyms(rtxn)?.remove(&words))
     }
@@ -611,8 +686,16 @@ impl Index {
     /* words prefixes fst */
 
     /// Writes the FST which is the words prefixes dictionnary of the engine.
-    pub(crate) fn put_words_prefixes_fst<A: AsRef<[u8]>>(&self, wtxn: &mut RwTxn, fst: &fst::Set<A>) -> heed::Result<()> {
-        self.main.put::<_, Str, ByteSlice>(wtxn, main_key::WORDS_PREFIXES_FST_KEY, fst.as_fst().as_bytes())
+    pub(crate) fn put_words_prefixes_fst<A: AsRef<[u8]>>(
+        &self,
+        wtxn: &mut RwTxn,
+        fst: &fst::Set<A>,
+    ) -> heed::Result<()> {
+        self.main.put::<_, Str, ByteSlice>(
+            wtxn,
+            main_key::WORDS_PREFIXES_FST_KEY,
+            fst.as_fst().as_bytes(),
+        )
     }
 
     /// Returns the FST which is the words prefixes dictionnary of the engine.
@@ -637,13 +720,14 @@ impl Index {
     pub fn documents<'t>(
         &self,
         rtxn: &'t RoTxn,
-        ids: impl IntoIterator<Item=DocumentId>,
-    ) -> Result<Vec<(DocumentId, obkv::KvReader<'t>)>>
-    {
+        ids: impl IntoIterator<Item = DocumentId>,
+    ) -> Result<Vec<(DocumentId, obkv::KvReader<'t>)>> {
         let mut documents = Vec::new();
 
         for id in ids {
-            let kv = self.documents.get(rtxn, &BEU32::new(id))?
+            let kv = self
+                .documents
+                .get(rtxn, &BEU32::new(id))?
                 .ok_or_else(|| UserError::UnknownInternalDocumentId { document_id: id })?;
             documents.push((id, kv));
         }
@@ -673,7 +757,8 @@ impl Index {
 
     /// Returns the index creation time.
     pub fn created_at(&self, rtxn: &RoTxn) -> Result<DateTime<Utc>> {
-        Ok(self.main
+        Ok(self
+            .main
             .get::<_, Str, SerdeJson<DateTime<Utc>>>(rtxn, main_key::CREATED_AT_KEY)?
             .ok_or(InternalError::DatabaseMissingEntry {
                 db_name: db_name::MAIN,
@@ -683,7 +768,8 @@ impl Index {
 
     /// Returns the index last updated time.
     pub fn updated_at(&self, rtxn: &RoTxn) -> Result<DateTime<Utc>> {
-        Ok(self.main
+        Ok(self
+            .main
             .get::<_, Str, SerdeJson<DateTime<Utc>>>(rtxn, main_key::UPDATED_AT_KEY)?
             .ok_or(InternalError::DatabaseMissingEntry {
                 db_name: db_name::MAIN,
@@ -691,7 +777,11 @@ impl Index {
             })?)
     }
 
-    pub(crate) fn set_updated_at(&self, wtxn: &mut RwTxn, time: &DateTime<Utc>) -> heed::Result<()> {
+    pub(crate) fn set_updated_at(
+        &self,
+        wtxn: &mut RwTxn,
+        time: &DateTime<Utc>,
+    ) -> heed::Result<()> {
         self.main.put::<_, Str, SerdeJson<DateTime<Utc>>>(wtxn, main_key::UPDATED_AT_KEY, &time)
     }
 }
@@ -704,8 +794,8 @@ pub(crate) mod tests {
     use maplit::hashmap;
     use tempfile::TempDir;
 
-    use crate::Index;
     use crate::update::{IndexDocuments, UpdateFormat};
+    use crate::Index;
 
     pub(crate) struct TempIndex {
         inner: Index,
@@ -728,10 +818,7 @@ pub(crate) mod tests {
             options.map_size(100 * 4096);
             let _tempdir = TempDir::new_in(".").unwrap();
             let inner = Index::new(options, _tempdir.path()).unwrap();
-            Self {
-                inner,
-                _tempdir
-            }
+            Self { inner, _tempdir }
         }
     }
 
@@ -756,10 +843,13 @@ pub(crate) mod tests {
         let rtxn = index.read_txn().unwrap();
 
         let fields_distribution = index.fields_distribution(&rtxn).unwrap();
-        assert_eq!(fields_distribution, hashmap! {
-            "id".to_string() => 2,
-            "name".to_string() => 2,
-            "age".to_string() => 1,
-        });
+        assert_eq!(
+            fields_distribution,
+            hashmap! {
+                "id".to_string() => 2,
+                "name".to_string() => 2,
+                "age".to_string() => 1,
+            }
+        );
     }
 }
diff --git a/milli/src/lib.rs b/milli/src/lib.rs
index f37244114..201035a8a 100644
--- a/milli/src/lib.rs
+++ b/milli/src/lib.rs
@@ -1,14 +1,15 @@
-#[macro_use] extern crate pest_derive;
+#[macro_use]
+extern crate pest_derive;
 
 mod criterion;
 mod error;
 mod external_documents_ids;
-mod fields_ids_map;
-mod search;
 pub mod facet;
+mod fields_ids_map;
 pub mod heed_codec;
 pub mod index;
 pub mod proximity;
+mod search;
 pub mod tree_level;
 pub mod update;
 
@@ -20,15 +21,17 @@ use std::result::Result as StdResult;
 use fxhash::{FxHasher32, FxHasher64};
 use serde_json::{Map, Value};
 
-pub use self::criterion::{Criterion, default_criteria};
+pub use self::criterion::{default_criteria, Criterion};
 pub use self::error::Error;
 pub use self::external_documents_ids::ExternalDocumentsIds;
 pub use self::fields_ids_map::FieldsIdsMap;
-pub use self::heed_codec::{BEU32StrCodec, StrStrU8Codec, StrLevelPositionCodec, ObkvCodec, FieldIdWordCountCodec};
-pub use self::heed_codec::{RoaringBitmapCodec, BoRoaringBitmapCodec, CboRoaringBitmapCodec};
-pub use self::heed_codec::{RoaringBitmapLenCodec, BoRoaringBitmapLenCodec, CboRoaringBitmapLenCodec};
+pub use self::heed_codec::{
+    BEU32StrCodec, BoRoaringBitmapCodec, BoRoaringBitmapLenCodec, CboRoaringBitmapCodec,
+    CboRoaringBitmapLenCodec, FieldIdWordCountCodec, ObkvCodec, RoaringBitmapCodec,
+    RoaringBitmapLenCodec, StrLevelPositionCodec, StrStrU8Codec,
+};
 pub use self::index::Index;
-pub use self::search::{Search, FacetDistribution, FilterCondition, SearchResult, MatchingWords};
+pub use self::search::{FacetDistribution, FilterCondition, MatchingWords, Search, SearchResult};
 pub use self::tree_level::TreeLevel;
 
 pub type Result<T> = std::result::Result<T, error::Error>;
@@ -54,9 +57,9 @@ pub fn obkv_to_json(
     displayed_fields: &[FieldId],
     fields_ids_map: &FieldsIdsMap,
     obkv: obkv::KvReader,
-) -> Result<Map<String, Value>>
-{
-    displayed_fields.iter()
+) -> Result<Map<String, Value>> {
+    displayed_fields
+        .iter()
         .copied()
         .flat_map(|id| obkv.get(id).map(|value| (id, value)))
         .map(|(id, value)| {
@@ -72,7 +75,6 @@ pub fn obkv_to_json(
 
 /// Transform a JSON value into a string that can be indexed.
 pub fn json_to_string(value: &Value) -> Option<String> {
-
     fn inner(value: &Value, output: &mut String) -> bool {
         use std::fmt::Write;
         match value {
@@ -90,7 +92,7 @@ pub fn json_to_string(value: &Value) -> Option<String> {
                 }
                 // check that at least one value was written
                 count != 0
-            },
+            }
             Value::Object(object) => {
                 let mut buffer = String::new();
                 let mut count = 0;
@@ -107,7 +109,7 @@ pub fn json_to_string(value: &Value) -> Option<String> {
                 }
                 // check that at least one value was written
                 count != 0
-            },
+            }
         }
     }
 
@@ -121,9 +123,10 @@ pub fn json_to_string(value: &Value) -> Option<String> {
 
 #[cfg(test)]
 mod tests {
-    use super::*;
     use serde_json::json;
 
+    use super::*;
+
     #[test]
     fn json_to_string_object() {
         let value = json!({
diff --git a/milli/src/proximity.rs b/milli/src/proximity.rs
index 0186eb3d0..db98426a5 100644
--- a/milli/src/proximity.rs
+++ b/milli/src/proximity.rs
@@ -1,4 +1,5 @@
 use std::cmp;
+
 use crate::{Attribute, Position};
 
 const ONE_ATTRIBUTE: u32 = 1000;
@@ -15,8 +16,11 @@ pub fn index_proximity(lhs: u32, rhs: u32) -> u32 {
 pub fn positions_proximity(lhs: Position, rhs: Position) -> u32 {
     let (lhs_attr, lhs_index) = extract_position(lhs);
     let (rhs_attr, rhs_index) = extract_position(rhs);
-    if lhs_attr != rhs_attr { MAX_DISTANCE }
-    else { index_proximity(lhs_index, rhs_index) }
+    if lhs_attr != rhs_attr {
+        MAX_DISTANCE
+    } else {
+        index_proximity(lhs_index, rhs_index)
+    }
 }
 
 pub fn extract_position(position: Position) -> (Attribute, Position) {
diff --git a/milli/src/search/criteria/asc_desc.rs b/milli/src/search/criteria/asc_desc.rs
index 95f77fd78..ccee2c393 100644
--- a/milli/src/search/criteria/asc_desc.rs
+++ b/milli/src/search/criteria/asc_desc.rs
@@ -5,12 +5,12 @@ use log::debug;
 use ordered_float::OrderedFloat;
 use roaring::RoaringBitmap;
 
+use super::{Criterion, CriterionParameters, CriterionResult};
 use crate::error::FieldIdMapMissingEntry;
 use crate::search::criteria::{resolve_query_tree, CriteriaBuilder};
 use crate::search::facet::FacetIter;
 use crate::search::query_tree::Operation;
 use crate::{FieldId, Index, Result};
-use super::{Criterion, CriterionParameters, CriterionResult};
 
 /// Threshold on the number of candidates that will make
 /// the system to choose between one algorithm or another.
@@ -57,9 +57,8 @@ impl<'t> AscDesc<'t> {
         ascending: bool,
     ) -> Result<Self> {
         let fields_ids_map = index.fields_ids_map(rtxn)?;
-        let field_id = fields_ids_map
-            .id(&field_name)
-            .ok_or_else(|| FieldIdMapMissingEntry::FieldName {
+        let field_id =
+            fields_ids_map.id(&field_name).ok_or_else(|| FieldIdMapMissingEntry::FieldName {
                 field_name: field_name.clone(),
                 process: "AscDesc::new",
             })?;
@@ -101,44 +100,47 @@ impl<'t> Criterion for AscDesc<'t> {
                         filtered_candidates: None,
                         bucket_candidates: Some(take(&mut self.bucket_candidates)),
                     }));
-                },
-                None => {
-                    match self.parent.next(params)? {
-                        Some(CriterionResult { query_tree, candidates, filtered_candidates, bucket_candidates }) => {
-                            self.query_tree = query_tree;
-                            let mut candidates = match (&self.query_tree, candidates) {
-                                (_, Some(candidates)) => candidates,
-                                (Some(qt), None) => {
-                                    let context = CriteriaBuilder::new(&self.rtxn, &self.index)?;
-                                    resolve_query_tree(&context, qt, params.wdcache)?
-                                },
-                                (None, None) => self.index.documents_ids(self.rtxn)?,
-                            };
-
-                            if let Some(filtered_candidates) = filtered_candidates {
-                                candidates &= filtered_candidates;
+                }
+                None => match self.parent.next(params)? {
+                    Some(CriterionResult {
+                        query_tree,
+                        candidates,
+                        filtered_candidates,
+                        bucket_candidates,
+                    }) => {
+                        self.query_tree = query_tree;
+                        let mut candidates = match (&self.query_tree, candidates) {
+                            (_, Some(candidates)) => candidates,
+                            (Some(qt), None) => {
+                                let context = CriteriaBuilder::new(&self.rtxn, &self.index)?;
+                                resolve_query_tree(&context, qt, params.wdcache)?
                             }
+                            (None, None) => self.index.documents_ids(self.rtxn)?,
+                        };
 
-                            match bucket_candidates {
-                                Some(bucket_candidates) => self.bucket_candidates |= bucket_candidates,
-                                None => self.bucket_candidates |= &candidates,
-                            }
+                        if let Some(filtered_candidates) = filtered_candidates {
+                            candidates &= filtered_candidates;
+                        }
 
-                            if candidates.is_empty() {
-                                continue;
-                            }
+                        match bucket_candidates {
+                            Some(bucket_candidates) => self.bucket_candidates |= bucket_candidates,
+                            None => self.bucket_candidates |= &candidates,
+                        }
 
-                            self.allowed_candidates = &candidates - params.excluded_candidates;
-                            self.candidates = facet_ordered(
-                                self.index,
-                                self.rtxn,
-                                self.field_id,
-                                self.ascending,
-                                candidates & &self.faceted_candidates,
-                            )?;
-                        },
-                        None => return Ok(None),
+                        if candidates.is_empty() {
+                            continue;
+                        }
+
+                        self.allowed_candidates = &candidates - params.excluded_candidates;
+                        self.candidates = facet_ordered(
+                            self.index,
+                            self.rtxn,
+                            self.field_id,
+                            self.ascending,
+                            candidates & &self.faceted_candidates,
+                        )?;
                     }
+                    None => return Ok(None),
                 },
                 Some(mut candidates) => {
                     candidates -= params.excluded_candidates;
@@ -170,11 +172,8 @@ fn facet_ordered<'t>(
         let iter = iterative_facet_ordered_iter(index, rtxn, field_id, ascending, candidates)?;
         Ok(Box::new(iter.map(Ok)) as Box<dyn Iterator<Item = _>>)
     } else {
-        let facet_fn = if ascending {
-            FacetIter::new_reducing
-        } else {
-            FacetIter::new_reverse_reducing
-        };
+        let facet_fn =
+            if ascending { FacetIter::new_reducing } else { FacetIter::new_reverse_reducing };
         let iter = facet_fn(rtxn, index, field_id, candidates)?;
         Ok(Box::new(iter.map(|res| res.map(|(_, docids)| docids))))
     }
@@ -194,9 +193,7 @@ fn iterative_facet_ordered_iter<'t>(
     for docid in candidates.iter() {
         let left = (field_id, docid, f64::MIN);
         let right = (field_id, docid, f64::MAX);
-        let mut iter = index
-            .field_id_docid_facet_f64s
-            .range(rtxn, &(left..=right))?;
+        let mut iter = index.field_id_docid_facet_f64s.range(rtxn, &(left..=right))?;
         let entry = if ascending { iter.next() } else { iter.last() };
         if let Some(((_, _, value), ())) = entry.transpose()? {
             docids_values.push((docid, OrderedFloat(value)));
diff --git a/milli/src/search/criteria/attribute.rs b/milli/src/search/criteria/attribute.rs
index f191defe1..6e0bb40d5 100644
--- a/milli/src/search/criteria/attribute.rs
+++ b/milli/src/search/criteria/attribute.rs
@@ -1,15 +1,16 @@
-use std::{borrow::Cow, cmp::{self, Ordering}, collections::BinaryHeap};
-use std::collections::{BTreeMap, HashMap, btree_map};
+use std::borrow::Cow;
+use std::cmp::{self, Ordering};
 use std::collections::binary_heap::PeekMut;
+use std::collections::{btree_map, BTreeMap, BinaryHeap, HashMap};
 use std::mem::take;
 
 use roaring::RoaringBitmap;
 
-use crate::{TreeLevel, Result, search::build_dfa};
+use super::{resolve_query_tree, Context, Criterion, CriterionParameters, CriterionResult};
 use crate::search::criteria::Query;
 use crate::search::query_tree::{Operation, QueryKind};
-use crate::search::{word_derivations, WordDerivationsCache};
-use super::{Criterion, CriterionParameters, CriterionResult, Context, resolve_query_tree};
+use crate::search::{build_dfa, word_derivations, WordDerivationsCache};
+use crate::{Result, TreeLevel};
 
 /// To be able to divide integers by the number of words in the query
 /// we want to find a multiplier that allow us to divide by any number between 1 and 10.
@@ -63,15 +64,19 @@ impl<'t> Criterion for Attribute<'t> {
                         filtered_candidates: None,
                         bucket_candidates: Some(take(&mut self.bucket_candidates)),
                     }));
-                },
+                }
                 Some((query_tree, flattened_query_tree, mut allowed_candidates)) => {
                     let found_candidates = if allowed_candidates.len() < CANDIDATES_THRESHOLD {
                         let current_buckets = match self.current_buckets.as_mut() {
                             Some(current_buckets) => current_buckets,
                             None => {
-                                let new_buckets = linear_compute_candidates(self.ctx, &flattened_query_tree, &allowed_candidates)?;
+                                let new_buckets = linear_compute_candidates(
+                                    self.ctx,
+                                    &flattened_query_tree,
+                                    &allowed_candidates,
+                                )?;
                                 self.current_buckets.get_or_insert(new_buckets.into_iter())
-                            },
+                            }
                         };
 
                         match current_buckets.next() {
@@ -83,10 +88,15 @@ impl<'t> Criterion for Attribute<'t> {
                                     filtered_candidates: None,
                                     bucket_candidates: Some(take(&mut self.bucket_candidates)),
                                 }));
-                            },
+                            }
                         }
                     } else {
-                        match set_compute_candidates(self.ctx, &flattened_query_tree, &allowed_candidates, params.wdcache)? {
+                        match set_compute_candidates(
+                            self.ctx,
+                            &flattened_query_tree,
+                            &allowed_candidates,
+                            params.wdcache,
+                        )? {
                             Some(candidates) => candidates,
                             None => {
                                 return Ok(Some(CriterionResult {
@@ -95,13 +105,14 @@ impl<'t> Criterion for Attribute<'t> {
                                     filtered_candidates: None,
                                     bucket_candidates: Some(take(&mut self.bucket_candidates)),
                                 }));
-                            },
+                            }
                         }
                     };
 
                     allowed_candidates -= &found_candidates;
 
-                    self.state = Some((query_tree.clone(), flattened_query_tree, allowed_candidates));
+                    self.state =
+                        Some((query_tree.clone(), flattened_query_tree, allowed_candidates));
 
                     return Ok(Some(CriterionResult {
                         query_tree: Some(query_tree),
@@ -109,39 +120,50 @@ impl<'t> Criterion for Attribute<'t> {
                         filtered_candidates: None,
                         bucket_candidates: Some(take(&mut self.bucket_candidates)),
                     }));
-                },
-                None => {
-                    match self.parent.next(params)? {
-                        Some(CriterionResult { query_tree: Some(query_tree), candidates, filtered_candidates, bucket_candidates }) => {
-                            let mut candidates = match candidates {
-                                Some(candidates) => candidates,
-                                None => resolve_query_tree(self.ctx, &query_tree, params.wdcache)? - params.excluded_candidates,
-                            };
-
-                            if let Some(filtered_candidates) = filtered_candidates {
-                                candidates &= filtered_candidates;
+                }
+                None => match self.parent.next(params)? {
+                    Some(CriterionResult {
+                        query_tree: Some(query_tree),
+                        candidates,
+                        filtered_candidates,
+                        bucket_candidates,
+                    }) => {
+                        let mut candidates = match candidates {
+                            Some(candidates) => candidates,
+                            None => {
+                                resolve_query_tree(self.ctx, &query_tree, params.wdcache)?
+                                    - params.excluded_candidates
                             }
+                        };
 
-                            let flattened_query_tree = flatten_query_tree(&query_tree);
+                        if let Some(filtered_candidates) = filtered_candidates {
+                            candidates &= filtered_candidates;
+                        }
 
-                            match bucket_candidates {
-                                Some(bucket_candidates) => self.bucket_candidates |= bucket_candidates,
-                                None => self.bucket_candidates |= &candidates,
-                            }
+                        let flattened_query_tree = flatten_query_tree(&query_tree);
 
-                            self.state = Some((query_tree, flattened_query_tree, candidates));
-                            self.current_buckets = None;
-                        },
-                        Some(CriterionResult { query_tree: None, candidates, filtered_candidates, bucket_candidates }) => {
-                            return Ok(Some(CriterionResult {
-                                query_tree: None,
-                                candidates,
-                                filtered_candidates,
-                                bucket_candidates,
-                            }));
-                        },
-                        None => return Ok(None),
+                        match bucket_candidates {
+                            Some(bucket_candidates) => self.bucket_candidates |= bucket_candidates,
+                            None => self.bucket_candidates |= &candidates,
+                        }
+
+                        self.state = Some((query_tree, flattened_query_tree, candidates));
+                        self.current_buckets = None;
                     }
+                    Some(CriterionResult {
+                        query_tree: None,
+                        candidates,
+                        filtered_candidates,
+                        bucket_candidates,
+                    }) => {
+                        return Ok(Some(CriterionResult {
+                            query_tree: None,
+                            candidates,
+                            filtered_candidates,
+                            bucket_candidates,
+                        }));
+                    }
+                    None => return Ok(None),
                 },
             }
         }
@@ -152,7 +174,9 @@ impl<'t> Criterion for Attribute<'t> {
 /// it will begin at the first non-empty interval and will return every interval without
 /// jumping over empty intervals.
 struct WordLevelIterator<'t, 'q> {
-    inner: Box<dyn Iterator<Item =heed::Result<((&'t str, TreeLevel, u32, u32), RoaringBitmap)>> + 't>,
+    inner: Box<
+        dyn Iterator<Item = heed::Result<((&'t str, TreeLevel, u32, u32), RoaringBitmap)>> + 't,
+    >,
     level: TreeLevel,
     interval_size: u32,
     word: Cow<'q, str>,
@@ -162,49 +186,80 @@ struct WordLevelIterator<'t, 'q> {
 }
 
 impl<'t, 'q> WordLevelIterator<'t, 'q> {
-    fn new(ctx: &'t dyn Context<'t>, word: Cow<'q, str>, in_prefix_cache: bool) -> heed::Result<Option<Self>> {
+    fn new(
+        ctx: &'t dyn Context<'t>,
+        word: Cow<'q, str>,
+        in_prefix_cache: bool,
+    ) -> heed::Result<Option<Self>> {
         match ctx.word_position_last_level(&word, in_prefix_cache)? {
-            Some(level) =>  {
+            Some(level) => {
                 let interval_size = LEVEL_EXPONENTIATION_BASE.pow(Into::<u8>::into(level) as u32);
-                let inner = ctx.word_position_iterator(&word, level, in_prefix_cache, None, None)?;
-                Ok(Some(Self { inner, level, interval_size, word, in_prefix_cache, inner_next: None, current_interval: None }))
-            },
+                let inner =
+                    ctx.word_position_iterator(&word, level, in_prefix_cache, None, None)?;
+                Ok(Some(Self {
+                    inner,
+                    level,
+                    interval_size,
+                    word,
+                    in_prefix_cache,
+                    inner_next: None,
+                    current_interval: None,
+                }))
+            }
             None => Ok(None),
         }
     }
 
-    fn dig(&self, ctx: &'t dyn Context<'t>, level: &TreeLevel, left_interval: Option<u32>) -> heed::Result<Self> {
+    fn dig(
+        &self,
+        ctx: &'t dyn Context<'t>,
+        level: &TreeLevel,
+        left_interval: Option<u32>,
+    ) -> heed::Result<Self> {
         let level = *level.min(&self.level);
         let interval_size = LEVEL_EXPONENTIATION_BASE.pow(Into::<u8>::into(level) as u32);
         let word = self.word.clone();
         let in_prefix_cache = self.in_prefix_cache;
-        let inner = ctx.word_position_iterator(&word, level, in_prefix_cache, left_interval, None)?;
+        let inner =
+            ctx.word_position_iterator(&word, level, in_prefix_cache, left_interval, None)?;
 
-        Ok(Self {inner, level, interval_size, word, in_prefix_cache, inner_next: None, current_interval: None})
+        Ok(Self {
+            inner,
+            level,
+            interval_size,
+            word,
+            in_prefix_cache,
+            inner_next: None,
+            current_interval: None,
+        })
     }
 
     fn next(&mut self) -> heed::Result<Option<(u32, u32, RoaringBitmap)>> {
-        fn is_next_interval(last_right: u32, next_left: u32) -> bool { last_right + 1 == next_left }
+        fn is_next_interval(last_right: u32, next_left: u32) -> bool {
+            last_right + 1 == next_left
+        }
 
         let inner_next = match self.inner_next.take() {
             Some(inner_next) => Some(inner_next),
-            None => self.inner.next().transpose()?.map(|((_, _, left, right), docids)| (left, right, docids)),
+            None => self
+                .inner
+                .next()
+                .transpose()?
+                .map(|((_, _, left, right), docids)| (left, right, docids)),
         };
 
         match inner_next {
-            Some((left, right, docids)) => {
-                match self.current_interval {
-                    Some((last_left, last_right)) if !is_next_interval(last_right, left) => {
-                        let blank_left = last_left + self.interval_size;
-                        let blank_right = last_right + self.interval_size;
-                        self.current_interval = Some((blank_left, blank_right));
-                        self.inner_next = Some((left, right, docids));
-                        Ok(Some((blank_left, blank_right, RoaringBitmap::new())))
-                    },
-                    _ => {
-                        self.current_interval = Some((left, right));
-                        Ok(Some((left, right, docids)))
-                    }
+            Some((left, right, docids)) => match self.current_interval {
+                Some((last_left, last_right)) if !is_next_interval(last_right, left) => {
+                    let blank_left = last_left + self.interval_size;
+                    let blank_right = last_right + self.interval_size;
+                    self.current_interval = Some((blank_left, blank_right));
+                    self.inner_next = Some((left, right, docids));
+                    Ok(Some((blank_left, blank_right, RoaringBitmap::new())))
+                }
+                _ => {
+                    self.current_interval = Some((left, right));
+                    Ok(Some((left, right, docids)))
                 }
             },
             None => Ok(None),
@@ -228,30 +283,37 @@ impl<'t, 'q> QueryLevelIterator<'t, 'q> {
         ctx: &'t dyn Context<'t>,
         queries: &'q [Query],
         wdcache: &mut WordDerivationsCache,
-    ) -> Result<Option<Self>>
-    {
+    ) -> Result<Option<Self>> {
         let mut inner = Vec::with_capacity(queries.len());
         for query in queries {
             match &query.kind {
                 QueryKind::Exact { word, .. } => {
                     if !query.prefix || ctx.in_prefix_cache(&word) {
                         let word = Cow::Borrowed(query.kind.word());
-                        if let Some(word_level_iterator) = WordLevelIterator::new(ctx, word, query.prefix)? {
+                        if let Some(word_level_iterator) =
+                            WordLevelIterator::new(ctx, word, query.prefix)?
+                        {
                             inner.push(word_level_iterator);
                         }
                     } else {
-                        for (word, _) in word_derivations(&word, true, 0, ctx.words_fst(), wdcache)? {
+                        for (word, _) in word_derivations(&word, true, 0, ctx.words_fst(), wdcache)?
+                        {
                             let word = Cow::Owned(word.to_owned());
-                            if let Some(word_level_iterator) = WordLevelIterator::new(ctx, word, false)? {
+                            if let Some(word_level_iterator) =
+                                WordLevelIterator::new(ctx, word, false)?
+                            {
                                 inner.push(word_level_iterator);
                             }
                         }
                     }
-                },
+                }
                 QueryKind::Tolerant { typo, word } => {
-                    for (word, _) in word_derivations(&word, query.prefix, *typo, ctx.words_fst(), wdcache)? {
+                    for (word, _) in
+                        word_derivations(&word, query.prefix, *typo, ctx.words_fst(), wdcache)?
+                    {
                         let word = Cow::Owned(word.to_owned());
-                        if let Some(word_level_iterator) = WordLevelIterator::new(ctx, word, false)? {
+                        if let Some(word_level_iterator) = WordLevelIterator::new(ctx, word, false)?
+                        {
                             inner.push(word_level_iterator);
                         }
                     }
@@ -284,17 +346,28 @@ impl<'t, 'q> QueryLevelIterator<'t, 'q> {
             Some(parent) => {
                 let parent = parent.dig(ctx)?;
                 (parent.level.min(self.level), Some(Box::new(parent)))
-            },
+            }
             None => (self.level.saturating_sub(1), None),
         };
 
-        let left_interval = self.accumulator.get(self.interval_to_skip).map(|opt| opt.as_ref().map(|(left, _, _)| *left)).flatten();
+        let left_interval = self
+            .accumulator
+            .get(self.interval_to_skip)
+            .map(|opt| opt.as_ref().map(|(left, _, _)| *left))
+            .flatten();
         let mut inner = Vec::with_capacity(self.inner.len());
         for word_level_iterator in self.inner.iter() {
             inner.push(word_level_iterator.dig(ctx, &level, left_interval)?);
         }
 
-        Ok(Self {parent, inner, level, accumulator: vec![], parent_accumulator: vec![], interval_to_skip: 0})
+        Ok(Self {
+            parent,
+            inner,
+            level,
+            accumulator: vec![],
+            parent_accumulator: vec![],
+            interval_to_skip: 0,
+        })
     }
 
     fn inner_next(&mut self, level: TreeLevel) -> heed::Result<Option<(u32, u32, RoaringBitmap)>> {
@@ -305,12 +378,12 @@ impl<'t, 'q> QueryLevelIterator<'t, 'q> {
             let wli_u8_level = Into::<u8>::into(wli.level);
             let accumulated_count = LEVEL_EXPONENTIATION_BASE.pow((u8_level - wli_u8_level) as u32);
             for _ in 0..accumulated_count {
-                if let Some((next_left, _, next_docids)) =  wli.next()? {
-                    accumulated = match accumulated.take(){
+                if let Some((next_left, _, next_docids)) = wli.next()? {
+                    accumulated = match accumulated.take() {
                         Some((acc_left, acc_right, mut acc_docids)) => {
                             acc_docids |= next_docids;
                             Some((acc_left, acc_right, acc_docids))
-                        },
+                        }
                         None => Some((next_left, next_left + interval_size, next_docids)),
                     };
                 }
@@ -322,7 +395,11 @@ impl<'t, 'q> QueryLevelIterator<'t, 'q> {
 
     /// return the next meta-interval created from inner WordLevelIterators,
     /// and from eventual chainned QueryLevelIterator.
-    fn next(&mut self, allowed_candidates: &RoaringBitmap, tree_level: TreeLevel) -> heed::Result<Option<(u32, u32, RoaringBitmap)>> {
+    fn next(
+        &mut self,
+        allowed_candidates: &RoaringBitmap,
+        tree_level: TreeLevel,
+    ) -> heed::Result<Option<(u32, u32, RoaringBitmap)>> {
         let parent_result = match self.parent.as_mut() {
             Some(parent) => Some(parent.next(allowed_candidates, tree_level)?),
             None => None,
@@ -335,22 +412,30 @@ impl<'t, 'q> QueryLevelIterator<'t, 'q> {
                     &self.parent_accumulator,
                     &self.accumulator,
                     self.interval_to_skip,
-                    allowed_candidates
+                    allowed_candidates,
                 );
                 self.accumulator.push(inner_next);
                 self.parent_accumulator.push(parent_next);
                 let mut merged_interval: Option<(u32, u32, RoaringBitmap)> = None;
 
-                for current in self.accumulator.iter().rev().zip(self.parent_accumulator.iter()).skip(self.interval_to_skip) {
+                for current in self
+                    .accumulator
+                    .iter()
+                    .rev()
+                    .zip(self.parent_accumulator.iter())
+                    .skip(self.interval_to_skip)
+                {
                     if let (Some((left_a, right_a, a)), Some((left_b, right_b, b))) = current {
                         match merged_interval.as_mut() {
                             Some((_, _, merged_docids)) => *merged_docids |= a & b,
-                            None => merged_interval = Some((left_a + left_b, right_a + right_b, a & b)),
+                            None => {
+                                merged_interval = Some((left_a + left_b, right_a + right_b, a & b))
+                            }
                         }
                     }
                 }
                 Ok(merged_interval)
-            },
+            }
             None => {
                 let level = self.level;
                 match self.inner_next(level)? {
@@ -358,12 +443,11 @@ impl<'t, 'q> QueryLevelIterator<'t, 'q> {
                         self.accumulator = vec![Some((left, right, RoaringBitmap::new()))];
                         candidates &= allowed_candidates;
                         Ok(Some((left, right, candidates)))
-
-                    },
+                    }
                     None => {
                         self.accumulator = vec![None];
                         Ok(None)
-                    },
+                    }
                 }
             }
         }
@@ -379,16 +463,18 @@ fn interval_to_skip(
     already_skiped: usize,
     allowed_candidates: &RoaringBitmap,
 ) -> usize {
-    parent_accumulator.iter()
+    parent_accumulator
+        .iter()
         .zip(current_accumulator.iter())
         .skip(already_skiped)
         .take_while(|(parent, current)| {
             let skip_parent = parent.as_ref().map_or(true, |(_, _, docids)| docids.is_empty());
-            let skip_current = current.as_ref().map_or(true, |(_, _, docids)| docids.is_disjoint(allowed_candidates));
+            let skip_current = current
+                .as_ref()
+                .map_or(true, |(_, _, docids)| docids.is_disjoint(allowed_candidates));
             skip_parent && skip_current
         })
         .count()
-
 }
 
 /// A Branch is represent a possible alternative of the original query and is build with the Query Tree,
@@ -410,7 +496,7 @@ impl<'t, 'q> Branch<'t, 'q> {
                 self.last_result = last_result;
                 self.tree_level = tree_level;
                 Ok(true)
-            },
+            }
             None => Ok(false),
         }
     }
@@ -429,7 +515,7 @@ impl<'t, 'q> Branch<'t, 'q> {
         let interval_size = LEVEL_EXPONENTIATION_BASE.pow(u8_level as u32);
         let (left, right, _) = self.last_result;
 
-        self.last_result = (left + interval_size,  right + interval_size, RoaringBitmap::new());
+        self.last_result = (left + interval_size, right + interval_size, RoaringBitmap::new());
     }
 
     /// return the score of the current inner interval.
@@ -477,31 +563,31 @@ fn initialize_query_level_iterators<'t, 'q>(
     allowed_candidates: &RoaringBitmap,
     wdcache: &mut WordDerivationsCache,
 ) -> Result<BinaryHeap<Branch<'t, 'q>>> {
-
     let mut positions = BinaryHeap::with_capacity(branches.len());
     for branch in branches {
         let mut branch_positions = Vec::with_capacity(branch.len());
-        for queries in  branch {
+        for queries in branch {
             match QueryLevelIterator::new(ctx, queries, wdcache)? {
                 Some(qli) => branch_positions.push(qli),
                 None => {
                     // the branch seems to be invalid, so we skip it.
                     branch_positions.clear();
                     break;
-                },
+                }
             }
         }
         // QueryLevelIterator need to be sorted by level and folded in descending order.
         branch_positions.sort_unstable_by_key(|qli| qli.level);
-        let folded_query_level_iterators = branch_positions
-            .into_iter()
-            .fold(None, |fold: Option<QueryLevelIterator>, mut qli| match fold {
-                Some(fold) => {
-                    qli.parent(fold);
-                    Some(qli)
-                },
-                None => Some(qli),
-        });
+        let folded_query_level_iterators =
+            branch_positions.into_iter().fold(None, |fold: Option<QueryLevelIterator>, mut qli| {
+                match fold {
+                    Some(fold) => {
+                        qli.parent(fold);
+                        Some(qli)
+                    }
+                    None => Some(qli),
+                }
+            });
 
         if let Some(mut folded_query_level_iterators) = folded_query_level_iterators {
             let tree_level = folded_query_level_iterators.level;
@@ -526,9 +612,9 @@ fn set_compute_candidates<'t>(
     branches: &FlattenedQueryTree,
     allowed_candidates: &RoaringBitmap,
     wdcache: &mut WordDerivationsCache,
-) -> Result<Option<RoaringBitmap>>
-{
-    let mut branches_heap = initialize_query_level_iterators(ctx, branches, allowed_candidates, wdcache)?;
+) -> Result<Option<RoaringBitmap>> {
+    let mut branches_heap =
+        initialize_query_level_iterators(ctx, branches, allowed_candidates, wdcache)?;
     let lowest_level = TreeLevel::min_value();
     let mut final_candidates: Option<(u32, RoaringBitmap)> = None;
     let mut allowed_candidates = allowed_candidates.clone();
@@ -539,15 +625,18 @@ fn set_compute_candidates<'t>(
         // if current is worst than best we break to return
         // candidates that correspond to the best rank
         if let Some((best_rank, _)) = final_candidates {
-            if branch_rank > best_rank { break }
+            if branch_rank > best_rank {
+                break;
+            }
         }
         let _left = branch.last_result.0;
         let candidates = take(&mut branch.last_result.2);
         if candidates.is_empty() {
             // we don't have candidates, get next interval.
-            if !branch.next(&allowed_candidates)? { PeekMut::pop(branch); }
-        }
-        else if is_lowest_level {
+            if !branch.next(&allowed_candidates)? {
+                PeekMut::pop(branch);
+            }
+        } else if is_lowest_level {
             // we have candidates, but we can't dig deeper.
             allowed_candidates -= &candidates;
             final_candidates = match final_candidates.take() {
@@ -556,19 +645,20 @@ fn set_compute_candidates<'t>(
                     best_candidates |= candidates;
                     branch.lazy_next();
                     Some((best_rank, best_candidates))
-                },
+                }
                 // we take current candidates as best candidates
                 None => {
                     branch.lazy_next();
                     Some((branch_rank, candidates))
-                },
+                }
             };
         } else {
             // we have candidates, lets dig deeper in levels.
             branch.dig(ctx)?;
-            if !branch.next(&allowed_candidates)? { PeekMut::pop(branch); }
+            if !branch.next(&allowed_candidates)? {
+                PeekMut::pop(branch);
+            }
         }
-
     }
 
     Ok(final_candidates.map(|(_rank, candidates)| candidates))
@@ -578,9 +668,11 @@ fn linear_compute_candidates(
     ctx: &dyn Context,
     branches: &FlattenedQueryTree,
     allowed_candidates: &RoaringBitmap,
-) -> Result<BTreeMap<u64, RoaringBitmap>>
-{
-    fn compute_candidate_rank(branches: &FlattenedQueryTree, words_positions: HashMap<String, RoaringBitmap>) -> u64 {
+) -> Result<BTreeMap<u64, RoaringBitmap>> {
+    fn compute_candidate_rank(
+        branches: &FlattenedQueryTree,
+        words_positions: HashMap<String, RoaringBitmap>,
+    ) -> u64 {
         let mut min_rank = u64::max_value();
         for branch in branches {
             let branch_len = branch.len();
@@ -593,17 +685,20 @@ fn linear_compute_candidates(
                         QueryKind::Exact { word, .. } => {
                             if *prefix {
                                 word_derivations(word, true, 0, &words_positions)
-                                    .flat_map(|positions| positions.iter().next()).min()
+                                    .flat_map(|positions| positions.iter().next())
+                                    .min()
                             } else {
-                                words_positions.get(word)
+                                words_positions
+                                    .get(word)
                                     .map(|positions| positions.iter().next())
                                     .flatten()
                             }
-                        },
+                        }
                         QueryKind::Tolerant { typo, word } => {
                             word_derivations(word, *prefix, *typo, &words_positions)
-                                .flat_map(|positions| positions.iter().next()).min()
-                        },
+                                .flat_map(|positions| positions.iter().next())
+                                .min()
+                        }
                     };
 
                     match (position, current_position) {
@@ -627,9 +722,11 @@ fn linear_compute_candidates(
                 branch_rank.sort_unstable();
                 // because several words in same query can't match all a the position 0,
                 // we substract the word index to the position.
-                let branch_rank: u64 = branch_rank.into_iter().enumerate().map(|(i, r)| r - i as u64).sum();
+                let branch_rank: u64 =
+                    branch_rank.into_iter().enumerate().map(|(i, r)| r - i as u64).sum();
                 // here we do the means of the words of the branch
-                min_rank = min_rank.min(branch_rank * LCM_10_FIRST_NUMBERS as u64 / branch_len as u64);
+                min_rank =
+                    min_rank.min(branch_rank * LCM_10_FIRST_NUMBERS as u64 / branch_len as u64);
             }
         }
 
@@ -641,8 +738,7 @@ fn linear_compute_candidates(
         is_prefix: bool,
         max_typo: u8,
         words_positions: &'a HashMap<String, RoaringBitmap>,
-    ) -> impl Iterator<Item = &'a RoaringBitmap>
-    {
+    ) -> impl Iterator<Item = &'a RoaringBitmap> {
         let dfa = build_dfa(word, max_typo, is_prefix);
         words_positions.iter().filter_map(move |(document_word, positions)| {
             use levenshtein_automata::Distance;
@@ -680,25 +776,26 @@ fn flatten_query_tree(query_tree: &Operation) -> FlattenedQueryTree {
                     }
                 }
                 out
-            },
+            }
             None => recurse(head),
         }
     }
 
     fn recurse(op: &Operation) -> FlattenedQueryTree {
         match op {
-            And(ops) => {
-                ops.split_first().map_or_else(Vec::new, |(h, t)| and_recurse(h, t))
-            },
-            Or(_, ops) => if ops.iter().all(|op| op.query().is_some()) {
-                vec![vec![ops.iter().flat_map(|op| op.query()).cloned().collect()]]
-            } else {
-                ops.iter().map(recurse).flatten().collect()
-            },
+            And(ops) => ops.split_first().map_or_else(Vec::new, |(h, t)| and_recurse(h, t)),
+            Or(_, ops) => {
+                if ops.iter().all(|op| op.query().is_some()) {
+                    vec![vec![ops.iter().flat_map(|op| op.query()).cloned().collect()]]
+                } else {
+                    ops.iter().map(recurse).flatten().collect()
+                }
+            }
             Phrase(words) => {
-                let queries = words.iter().map(|word| {
-                    vec![Query {prefix: false, kind: QueryKind::exact(word.clone())}]
-                }).collect();
+                let queries = words
+                    .iter()
+                    .map(|word| vec![Query { prefix: false, kind: QueryKind::exact(word.clone()) }])
+                    .collect();
                 vec![queries]
             }
             Operation::Query(query) => vec![vec![vec![query.clone()]]],
@@ -712,28 +809,43 @@ fn flatten_query_tree(query_tree: &Operation) -> FlattenedQueryTree {
 mod tests {
     use big_s::S;
 
-    use crate::search::criteria::QueryKind;
     use super::*;
+    use crate::search::criteria::QueryKind;
 
     #[test]
     fn simple_flatten_query_tree() {
-        let query_tree = Operation::Or(false, vec![
-            Operation::Query(Query { prefix: false, kind: QueryKind::exact(S("manythefish")) }),
-            Operation::And(vec![
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact(S("manythe")) }),
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact(S("fish")) }),
-            ]),
-            Operation::And(vec![
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact(S("many")) }),
-                Operation::Or(false, vec![
-                    Operation::Query(Query { prefix: false, kind: QueryKind::exact(S("thefish")) }),
-                    Operation::And(vec![
-                        Operation::Query(Query { prefix: false, kind: QueryKind::exact(S("the")) }),
-                        Operation::Query(Query { prefix: false, kind: QueryKind::exact(S("fish")) }),
-                    ]),
+        let query_tree = Operation::Or(
+            false,
+            vec![
+                Operation::Query(Query { prefix: false, kind: QueryKind::exact(S("manythefish")) }),
+                Operation::And(vec![
+                    Operation::Query(Query { prefix: false, kind: QueryKind::exact(S("manythe")) }),
+                    Operation::Query(Query { prefix: false, kind: QueryKind::exact(S("fish")) }),
                 ]),
-            ]),
-        ]);
+                Operation::And(vec![
+                    Operation::Query(Query { prefix: false, kind: QueryKind::exact(S("many")) }),
+                    Operation::Or(
+                        false,
+                        vec![
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::exact(S("thefish")),
+                            }),
+                            Operation::And(vec![
+                                Operation::Query(Query {
+                                    prefix: false,
+                                    kind: QueryKind::exact(S("the")),
+                                }),
+                                Operation::Query(Query {
+                                    prefix: false,
+                                    kind: QueryKind::exact(S("fish")),
+                                }),
+                            ]),
+                        ],
+                    ),
+                ]),
+            ],
+        );
 
         let expected = vec![
             vec![vec![Query { prefix: false, kind: QueryKind::exact(S("manythefish")) }]],
diff --git a/milli/src/search/criteria/exactness.rs b/milli/src/search/criteria/exactness.rs
index eb44b7b8e..1e4d4e7a2 100644
--- a/milli/src/search/criteria/exactness.rs
+++ b/milli/src/search/criteria/exactness.rs
@@ -2,19 +2,15 @@ use std::convert::TryFrom;
 use std::mem::take;
 use std::ops::BitOr;
 
+use itertools::Itertools;
 use log::debug;
 use roaring::RoaringBitmap;
-use itertools::Itertools;
 
-use crate::search::query_tree::{Operation, PrimitiveQueryPart};
 use crate::search::criteria::{
-    Context,
-    Criterion,
-    CriterionParameters,
-    CriterionResult,
-    resolve_query_tree,
+    resolve_query_tree, Context, Criterion, CriterionParameters, CriterionResult,
 };
-use crate::{TreeLevel, Result};
+use crate::search::query_tree::{Operation, PrimitiveQueryPart};
+use crate::{Result, TreeLevel};
 
 pub struct Exactness<'t> {
     ctx: &'t dyn Context<'t>,
@@ -26,7 +22,11 @@ pub struct Exactness<'t> {
 }
 
 impl<'t> Exactness<'t> {
-    pub fn new(ctx: &'t dyn Context<'t>, parent: Box<dyn Criterion + 't>, primitive_query: &[PrimitiveQueryPart]) -> heed::Result<Self> {
+    pub fn new(
+        ctx: &'t dyn Context<'t>,
+        parent: Box<dyn Criterion + 't>,
+        primitive_query: &[PrimitiveQueryPart],
+    ) -> heed::Result<Self> {
         let mut query: Vec<_> = Vec::with_capacity(primitive_query.len());
         for part in primitive_query {
             query.push(ExactQueryPart::from_primitive_query_part(ctx, part)?);
@@ -59,7 +59,7 @@ impl<'t> Criterion for Exactness<'t> {
                     // reset state
                     self.state = None;
                     self.query_tree = None;
-                },
+                }
                 Some(state) => {
                     let (candidates, state) = resolve_state(self.ctx, take(state), &self.query)?;
                     self.state = state;
@@ -70,40 +70,51 @@ impl<'t> Criterion for Exactness<'t> {
                         filtered_candidates: None,
                         bucket_candidates: Some(take(&mut self.bucket_candidates)),
                     }));
-                },
-                None => {
-                    match self.parent.next(params)? {
-                        Some(CriterionResult { query_tree: Some(query_tree), candidates, filtered_candidates, bucket_candidates }) => {
-                            let mut candidates = match candidates {
-                                Some(candidates) => candidates,
-                                None => resolve_query_tree(self.ctx, &query_tree, params.wdcache)? - params.excluded_candidates,
-                            };
-
-                            if let Some(filtered_candidates) = filtered_candidates {
-                                candidates &= filtered_candidates;
+                }
+                None => match self.parent.next(params)? {
+                    Some(CriterionResult {
+                        query_tree: Some(query_tree),
+                        candidates,
+                        filtered_candidates,
+                        bucket_candidates,
+                    }) => {
+                        let mut candidates = match candidates {
+                            Some(candidates) => candidates,
+                            None => {
+                                resolve_query_tree(self.ctx, &query_tree, params.wdcache)?
+                                    - params.excluded_candidates
                             }
+                        };
 
-                            match bucket_candidates {
-                                Some(bucket_candidates) => self.bucket_candidates |= bucket_candidates,
-                                None => self.bucket_candidates |= &candidates,
-                            }
+                        if let Some(filtered_candidates) = filtered_candidates {
+                            candidates &= filtered_candidates;
+                        }
 
-                            self.state = Some(State::new(candidates));
-                            self.query_tree = Some(query_tree);
-                        },
-                        Some(CriterionResult { query_tree: None, candidates, filtered_candidates, bucket_candidates }) => {
-                            return Ok(Some(CriterionResult {
-                                query_tree: None,
-                                candidates,
-                                filtered_candidates,
-                                bucket_candidates,
-                            }));
-                        },
-                        None => return Ok(None),
+                        match bucket_candidates {
+                            Some(bucket_candidates) => self.bucket_candidates |= bucket_candidates,
+                            None => self.bucket_candidates |= &candidates,
+                        }
+
+                        self.state = Some(State::new(candidates));
+                        self.query_tree = Some(query_tree);
                     }
+                    Some(CriterionResult {
+                        query_tree: None,
+                        candidates,
+                        filtered_candidates,
+                        bucket_candidates,
+                    }) => {
+                        return Ok(Some(CriterionResult {
+                            query_tree: None,
+                            candidates,
+                            filtered_candidates,
+                            bucket_candidates,
+                        }));
+                    }
+                    None => return Ok(None),
                 },
             }
-         }
+        }
     }
 }
 
@@ -125,9 +136,9 @@ impl State {
 
     fn difference_with(&mut self, lhs: &RoaringBitmap) {
         match self {
-            Self::ExactAttribute(candidates) |
-            Self::AttributeStartsWith(candidates) |
-            Self::ExactWords(candidates) => *candidates -= lhs,
+            Self::ExactAttribute(candidates)
+            | Self::AttributeStartsWith(candidates)
+            | Self::ExactWords(candidates) => *candidates -= lhs,
             Self::Remainings(candidates_array) => {
                 candidates_array.iter_mut().for_each(|candidates| *candidates -= lhs);
                 candidates_array.retain(|candidates| !candidates.is_empty());
@@ -137,9 +148,9 @@ impl State {
 
     fn is_empty(&self) -> bool {
         match self {
-            Self::ExactAttribute(candidates) |
-            Self::AttributeStartsWith(candidates) |
-            Self::ExactWords(candidates) => candidates.is_empty(),
+            Self::ExactAttribute(candidates)
+            | Self::AttributeStartsWith(candidates)
+            | Self::ExactWords(candidates) => candidates.is_empty(),
             Self::Remainings(candidates_array) => {
                 candidates_array.iter().all(RoaringBitmap::is_empty)
             }
@@ -158,8 +169,7 @@ fn resolve_state(
     ctx: &dyn Context,
     state: State,
     query: &[ExactQueryPart],
-) -> Result<(RoaringBitmap, Option<State>)>
-{
+) -> Result<(RoaringBitmap, Option<State>)> {
     use State::*;
     match state {
         ExactAttribute(mut allowed_candidates) => {
@@ -167,8 +177,11 @@ fn resolve_state(
             if let Ok(query_len) = u8::try_from(query.len()) {
                 let attributes_ids = ctx.searchable_fields_ids()?;
                 for id in attributes_ids {
-                    if let Some(attribute_allowed_docids) = ctx.field_id_word_count_docids(id, query_len)? {
-                        let mut attribute_candidates_array = attribute_start_with_docids(ctx, id as u32, query)?;
+                    if let Some(attribute_allowed_docids) =
+                        ctx.field_id_word_count_docids(id, query_len)?
+                    {
+                        let mut attribute_candidates_array =
+                            attribute_start_with_docids(ctx, id as u32, query)?;
                         attribute_candidates_array.push(attribute_allowed_docids);
                         candidates |= intersection_of(attribute_candidates_array.iter().collect());
                     }
@@ -181,12 +194,13 @@ fn resolve_state(
             }
 
             Ok((candidates, Some(AttributeStartsWith(allowed_candidates))))
-        },
+        }
         AttributeStartsWith(mut allowed_candidates) => {
             let mut candidates = RoaringBitmap::new();
             let attributes_ids = ctx.searchable_fields_ids()?;
             for id in attributes_ids {
-                let attribute_candidates_array = attribute_start_with_docids(ctx, id as u32, query)?;
+                let attribute_candidates_array =
+                    attribute_start_with_docids(ctx, id as u32, query)?;
                 candidates |= intersection_of(attribute_candidates_array.iter().collect());
             }
 
@@ -195,7 +209,7 @@ fn resolve_state(
             // remove current candidates from allowed candidates
             allowed_candidates -= &candidates;
             Ok((candidates, Some(ExactWords(allowed_candidates))))
-        },
+        }
         ExactWords(mut allowed_candidates) => {
             let number_of_part = query.len();
             let mut parts_candidates_array = Vec::with_capacity(number_of_part);
@@ -210,7 +224,7 @@ fn resolve_state(
                                 candidates |= synonym_candidates;
                             }
                         }
-                    },
+                    }
                     // compute intersection on pair of words with a proximity of 0.
                     Phrase(phrase) => {
                         let mut bitmaps = Vec::with_capacity(phrase.len().saturating_sub(1));
@@ -220,8 +234,8 @@ fn resolve_state(
                                     Some(docids) => bitmaps.push(docids),
                                     None => {
                                         bitmaps.clear();
-                                        break
-                                    },
+                                        break;
+                                    }
                                 }
                             }
                         }
@@ -247,7 +261,7 @@ fn resolve_state(
                     // intersect each word candidates in combinations
                     .map(intersection_of)
                     // union combinations of `c_count` exact words
-                    .fold(RoaringBitmap::new(),  RoaringBitmap::bitor);
+                    .fold(RoaringBitmap::new(), RoaringBitmap::bitor);
                 // only keep allowed candidates
                 combinations_candidates &= &allowed_candidates;
                 // remove current candidates from allowed candidates
@@ -261,7 +275,7 @@ fn resolve_state(
             candidates_array.reverse();
 
             Ok((all_exact_candidates, Some(Remainings(candidates_array))))
-        },
+        }
         // pop remainings candidates until the emptiness
         Remainings(mut candidates_array) => {
             let candidates = candidates_array.pop().unwrap_or_default();
@@ -270,12 +284,15 @@ fn resolve_state(
             } else {
                 Ok((candidates, None))
             }
-        },
-
+        }
     }
 }
 
-fn attribute_start_with_docids(ctx: &dyn Context, attribute_id: u32, query: &[ExactQueryPart]) -> heed::Result<Vec<RoaringBitmap>> {
+fn attribute_start_with_docids(
+    ctx: &dyn Context,
+    attribute_id: u32,
+    query: &[ExactQueryPart],
+) -> heed::Result<Vec<RoaringBitmap>> {
     let lowest_level = TreeLevel::min_value();
     let mut attribute_candidates_array = Vec::new();
     // start from attribute first position
@@ -293,7 +310,7 @@ fn attribute_start_with_docids(ctx: &dyn Context, attribute_id: u32, query: &[Ex
                 }
                 attribute_candidates_array.push(synonyms_candidates);
                 pos += 1;
-            },
+            }
             Phrase(phrase) => {
                 for word in phrase {
                     let wc = ctx.word_level_position_docids(word, lowest_level, pos, pos)?;
@@ -325,24 +342,30 @@ pub enum ExactQueryPart {
 }
 
 impl ExactQueryPart {
-    fn from_primitive_query_part(ctx: &dyn Context, part: &PrimitiveQueryPart) -> heed::Result<Self> {
+    fn from_primitive_query_part(
+        ctx: &dyn Context,
+        part: &PrimitiveQueryPart,
+    ) -> heed::Result<Self> {
         let part = match part {
             PrimitiveQueryPart::Word(word, _) => {
                 match ctx.synonyms(word)? {
                     Some(synonyms) => {
-                        let mut synonyms: Vec<_> = synonyms.into_iter().filter_map(|mut array| {
-                            // keep 1 word synonyms only.
-                            match array.pop() {
-                                Some(word) if array.is_empty() => Some(word),
-                                _ => None,
-                            }
-                        }).collect();
+                        let mut synonyms: Vec<_> = synonyms
+                            .into_iter()
+                            .filter_map(|mut array| {
+                                // keep 1 word synonyms only.
+                                match array.pop() {
+                                    Some(word) if array.is_empty() => Some(word),
+                                    _ => None,
+                                }
+                            })
+                            .collect();
                         synonyms.push(word.clone());
                         ExactQueryPart::Synonyms(synonyms)
-                    },
+                    }
                     None => ExactQueryPart::Synonyms(vec![word.clone()]),
                 }
-            },
+            }
             PrimitiveQueryPart::Phrase(phrase) => ExactQueryPart::Phrase(phrase.clone()),
         };
 
diff --git a/milli/src/search/criteria/final.rs b/milli/src/search/criteria/final.rs
index 645a3a5d7..bd3244143 100644
--- a/milli/src/search/criteria/final.rs
+++ b/milli/src/search/criteria/final.rs
@@ -1,10 +1,10 @@
 use log::debug;
 use roaring::RoaringBitmap;
 
-use crate::Result;
+use super::{resolve_query_tree, Context, Criterion, CriterionParameters, CriterionResult};
 use crate::search::query_tree::Operation;
 use crate::search::WordDerivationsCache;
-use super::{resolve_query_tree, Criterion, CriterionResult, CriterionParameters, Context};
+use crate::Result;
 
 /// The result of a call to the fetcher.
 #[derive(Debug, Clone, PartialEq)]
@@ -26,7 +26,12 @@ pub struct Final<'t> {
 
 impl<'t> Final<'t> {
     pub fn new(ctx: &'t dyn Context<'t>, parent: Box<dyn Criterion + 't>) -> Final<'t> {
-        Final { ctx, parent, wdcache: WordDerivationsCache::new(), returned_candidates: RoaringBitmap::new() }
+        Final {
+            ctx,
+            parent,
+            wdcache: WordDerivationsCache::new(),
+            returned_candidates: RoaringBitmap::new(),
+        }
     }
 
     #[logging_timer::time("Final::{}")]
@@ -40,10 +45,17 @@ impl<'t> Final<'t> {
         };
 
         match self.parent.next(&mut criterion_parameters)? {
-            Some(CriterionResult { query_tree, candidates, filtered_candidates, bucket_candidates }) => {
+            Some(CriterionResult {
+                query_tree,
+                candidates,
+                filtered_candidates,
+                bucket_candidates,
+            }) => {
                 let mut candidates = match (candidates, query_tree.as_ref()) {
                     (Some(candidates), _) => candidates,
-                    (None, Some(qt)) => resolve_query_tree(self.ctx, qt, &mut self.wdcache)? - excluded_candidates,
+                    (None, Some(qt)) => {
+                        resolve_query_tree(self.ctx, qt, &mut self.wdcache)? - excluded_candidates
+                    }
                     (None, None) => self.ctx.documents_ids()? - excluded_candidates,
                 };
 
@@ -56,7 +68,7 @@ impl<'t> Final<'t> {
                 self.returned_candidates |= &candidates;
 
                 Ok(Some(FinalResult { query_tree, candidates, bucket_candidates }))
-            },
+            }
             None => Ok(None),
         }
     }
diff --git a/milli/src/search/criteria/initial.rs b/milli/src/search/criteria/initial.rs
index e6d0a17f7..514dbff96 100644
--- a/milli/src/search/criteria/initial.rs
+++ b/milli/src/search/criteria/initial.rs
@@ -1,15 +1,18 @@
 use roaring::RoaringBitmap;
 
-use crate::Result;
+use super::{Criterion, CriterionParameters, CriterionResult};
 use crate::search::query_tree::Operation;
-use super::{Criterion, CriterionResult, CriterionParameters};
+use crate::Result;
 
 pub struct Initial {
-    answer: Option<CriterionResult>
+    answer: Option<CriterionResult>,
 }
 
 impl Initial {
-    pub fn new(query_tree: Option<Operation>, filtered_candidates: Option<RoaringBitmap>) -> Initial {
+    pub fn new(
+        query_tree: Option<Operation>,
+        filtered_candidates: Option<RoaringBitmap>,
+    ) -> Initial {
         let answer = CriterionResult {
             query_tree,
             candidates: None,
diff --git a/milli/src/search/criteria/mod.rs b/milli/src/search/criteria/mod.rs
index 48af0b8aa..228d48bd7 100644
--- a/milli/src/search/criteria/mod.rs
+++ b/milli/src/search/criteria/mod.rs
@@ -1,29 +1,28 @@
-use std::collections::HashMap;
 use std::borrow::Cow;
+use std::collections::HashMap;
 
 use roaring::RoaringBitmap;
 
-use crate::{FieldId, TreeLevel, search::{word_derivations, WordDerivationsCache}};
-use crate::{Index, DocumentId, Result};
-
-use super::query_tree::{Operation, PrimitiveQueryPart, Query, QueryKind};
 use self::asc_desc::AscDesc;
 use self::attribute::Attribute;
 use self::exactness::Exactness;
-use self::r#final::Final;
 use self::initial::Initial;
 use self::proximity::Proximity;
+use self::r#final::Final;
 use self::typo::Typo;
 use self::words::Words;
+use super::query_tree::{Operation, PrimitiveQueryPart, Query, QueryKind};
+use crate::search::{word_derivations, WordDerivationsCache};
+use crate::{DocumentId, FieldId, Index, Result, TreeLevel};
 
 mod asc_desc;
 mod attribute;
 mod exactness;
+pub mod r#final;
 mod initial;
 mod proximity;
 mod typo;
 mod words;
-pub mod r#final;
 
 pub trait Criterion {
     fn next(&mut self, params: &mut CriterionParameters) -> Result<Option<CriterionResult>>;
@@ -55,7 +54,7 @@ pub struct CriterionParameters<'a> {
 #[derive(Debug)]
 enum Candidates {
     Allowed(RoaringBitmap),
-    Forbidden(RoaringBitmap)
+    Forbidden(RoaringBitmap),
 }
 
 impl Default for Candidates {
@@ -68,17 +67,55 @@ pub trait Context<'c> {
     fn documents_ids(&self) -> heed::Result<RoaringBitmap>;
     fn word_docids(&self, word: &str) -> heed::Result<Option<RoaringBitmap>>;
     fn word_prefix_docids(&self, word: &str) -> heed::Result<Option<RoaringBitmap>>;
-    fn word_pair_proximity_docids(&self, left: &str, right: &str, proximity: u8) -> heed::Result<Option<RoaringBitmap>>;
-    fn word_prefix_pair_proximity_docids(&self, left: &str, right: &str, proximity: u8) -> heed::Result<Option<RoaringBitmap>>;
+    fn word_pair_proximity_docids(
+        &self,
+        left: &str,
+        right: &str,
+        proximity: u8,
+    ) -> heed::Result<Option<RoaringBitmap>>;
+    fn word_prefix_pair_proximity_docids(
+        &self,
+        left: &str,
+        right: &str,
+        proximity: u8,
+    ) -> heed::Result<Option<RoaringBitmap>>;
     fn words_fst<'t>(&self) -> &'t fst::Set<Cow<[u8]>>;
     fn in_prefix_cache(&self, word: &str) -> bool;
-    fn docid_words_positions(&self, docid: DocumentId) -> heed::Result<HashMap<String, RoaringBitmap>>;
-    fn word_position_iterator(&self, word: &str, level: TreeLevel, in_prefix_cache: bool, left: Option<u32>, right: Option<u32>) -> heed::Result<Box<dyn Iterator<Item =heed::Result<((&'c str, TreeLevel, u32, u32), RoaringBitmap)>> + 'c>>;
-    fn word_position_last_level(&self, word: &str, in_prefix_cache: bool) -> heed::Result<Option<TreeLevel>>;
+    fn docid_words_positions(
+        &self,
+        docid: DocumentId,
+    ) -> heed::Result<HashMap<String, RoaringBitmap>>;
+    fn word_position_iterator(
+        &self,
+        word: &str,
+        level: TreeLevel,
+        in_prefix_cache: bool,
+        left: Option<u32>,
+        right: Option<u32>,
+    ) -> heed::Result<
+        Box<
+            dyn Iterator<Item = heed::Result<((&'c str, TreeLevel, u32, u32), RoaringBitmap)>> + 'c,
+        >,
+    >;
+    fn word_position_last_level(
+        &self,
+        word: &str,
+        in_prefix_cache: bool,
+    ) -> heed::Result<Option<TreeLevel>>;
     fn synonyms(&self, word: &str) -> heed::Result<Option<Vec<Vec<String>>>>;
-    fn searchable_fields_ids(&self) ->  Result<Vec<FieldId>>;
-    fn field_id_word_count_docids(&self, field_id: FieldId, word_count: u8) -> heed::Result<Option<RoaringBitmap>>;
-    fn word_level_position_docids(&self, word: &str, level: TreeLevel, left: u32, right: u32) -> heed::Result<Option<RoaringBitmap>>;
+    fn searchable_fields_ids(&self) -> Result<Vec<FieldId>>;
+    fn field_id_word_count_docids(
+        &self,
+        field_id: FieldId,
+        word_count: u8,
+    ) -> heed::Result<Option<RoaringBitmap>>;
+    fn word_level_position_docids(
+        &self,
+        word: &str,
+        level: TreeLevel,
+        left: u32,
+        right: u32,
+    ) -> heed::Result<Option<RoaringBitmap>>;
 }
 
 pub struct CriteriaBuilder<'t> {
@@ -101,12 +138,22 @@ impl<'c> Context<'c> for CriteriaBuilder<'c> {
         self.index.word_prefix_docids.get(self.rtxn, &word)
     }
 
-    fn word_pair_proximity_docids(&self, left: &str, right: &str, proximity: u8) -> heed::Result<Option<RoaringBitmap>> {
+    fn word_pair_proximity_docids(
+        &self,
+        left: &str,
+        right: &str,
+        proximity: u8,
+    ) -> heed::Result<Option<RoaringBitmap>> {
         let key = (left, right, proximity);
         self.index.word_pair_proximity_docids.get(self.rtxn, &key)
     }
 
-    fn word_prefix_pair_proximity_docids(&self, left: &str, right: &str, proximity: u8) -> heed::Result<Option<RoaringBitmap>> {
+    fn word_prefix_pair_proximity_docids(
+        &self,
+        left: &str,
+        right: &str,
+        proximity: u8,
+    ) -> heed::Result<Option<RoaringBitmap>> {
         let key = (left, right, proximity);
         self.index.word_prefix_pair_proximity_docids.get(self.rtxn, &key)
     }
@@ -119,7 +166,10 @@ impl<'c> Context<'c> for CriteriaBuilder<'c> {
         self.words_prefixes_fst.contains(word)
     }
 
-    fn docid_words_positions(&self, docid: DocumentId) -> heed::Result<HashMap<String, RoaringBitmap>> {
+    fn docid_words_positions(
+        &self,
+        docid: DocumentId,
+    ) -> heed::Result<HashMap<String, RoaringBitmap>> {
         let mut words_positions = HashMap::new();
         for result in self.index.docid_word_positions.prefix_iter(self.rtxn, &(docid, ""))? {
             let ((_, word), positions) = result?;
@@ -134,9 +184,12 @@ impl<'c> Context<'c> for CriteriaBuilder<'c> {
         level: TreeLevel,
         in_prefix_cache: bool,
         left: Option<u32>,
-        right: Option<u32>
-    ) -> heed::Result<Box<dyn Iterator<Item = heed::Result<((&'c str, TreeLevel, u32, u32), RoaringBitmap)>> + 'c>>
-    {
+        right: Option<u32>,
+    ) -> heed::Result<
+        Box<
+            dyn Iterator<Item = heed::Result<((&'c str, TreeLevel, u32, u32), RoaringBitmap)>> + 'c,
+        >,
+    > {
         let range = {
             let left = left.unwrap_or(u32::min_value());
             let right = right.unwrap_or(u32::max_value());
@@ -152,7 +205,11 @@ impl<'c> Context<'c> for CriteriaBuilder<'c> {
         Ok(Box::new(db.range(self.rtxn, &range)?))
     }
 
-    fn word_position_last_level(&self, word: &str, in_prefix_cache: bool) -> heed::Result<Option<TreeLevel>> {
+    fn word_position_last_level(
+        &self,
+        word: &str,
+        in_prefix_cache: bool,
+    ) -> heed::Result<Option<TreeLevel>> {
         let range = {
             let left = (word, TreeLevel::min_value(), u32::min_value(), u32::min_value());
             let right = (word, TreeLevel::max_value(), u32::max_value(), u32::max_value());
@@ -164,7 +221,9 @@ impl<'c> Context<'c> for CriteriaBuilder<'c> {
         };
         let last_level = db
             .remap_data_type::<heed::types::DecodeIgnore>()
-            .range(self.rtxn, &range)?.last().transpose()?
+            .range(self.rtxn, &range)?
+            .last()
+            .transpose()?
             .map(|((_, level, _, _), _)| level);
 
         Ok(last_level)
@@ -181,12 +240,22 @@ impl<'c> Context<'c> for CriteriaBuilder<'c> {
         }
     }
 
-    fn field_id_word_count_docids(&self, field_id: FieldId, word_count: u8) -> heed::Result<Option<RoaringBitmap>> {
+    fn field_id_word_count_docids(
+        &self,
+        field_id: FieldId,
+        word_count: u8,
+    ) -> heed::Result<Option<RoaringBitmap>> {
         let key = (field_id, word_count);
         self.index.field_id_word_count_docids.get(self.rtxn, &key)
     }
 
-    fn word_level_position_docids(&self, word: &str, level: TreeLevel, left: u32, right: u32) -> heed::Result<Option<RoaringBitmap>> {
+    fn word_level_position_docids(
+        &self,
+        word: &str,
+        level: TreeLevel,
+        left: u32,
+        right: u32,
+    ) -> heed::Result<Option<RoaringBitmap>> {
         let key = (word, level, left, right);
         self.index.word_level_position_docids.get(self.rtxn, &key)
     }
@@ -204,13 +273,13 @@ impl<'t> CriteriaBuilder<'t> {
         query_tree: Option<Operation>,
         primitive_query: Option<Vec<PrimitiveQueryPart>>,
         filtered_candidates: Option<RoaringBitmap>,
-    ) -> Result<Final<'t>>
-    {
+    ) -> Result<Final<'t>> {
         use crate::criterion::Criterion as Name;
 
         let primitive_query = primitive_query.unwrap_or_default();
 
-        let mut criterion = Box::new(Initial::new(query_tree, filtered_candidates)) as Box<dyn Criterion>;
+        let mut criterion =
+            Box::new(Initial::new(query_tree, filtered_candidates)) as Box<dyn Criterion>;
         for name in self.index.criteria(&self.rtxn)? {
             criterion = match name {
                 Name::Typo => Box::new(Typo::new(self, criterion)),
@@ -218,8 +287,12 @@ impl<'t> CriteriaBuilder<'t> {
                 Name::Proximity => Box::new(Proximity::new(self, criterion)),
                 Name::Attribute => Box::new(Attribute::new(self, criterion)),
                 Name::Exactness => Box::new(Exactness::new(self, criterion, &primitive_query)?),
-                Name::Asc(field) => Box::new(AscDesc::asc(&self.index, &self.rtxn, criterion, field)?),
-                Name::Desc(field) => Box::new(AscDesc::desc(&self.index, &self.rtxn, criterion, field)?),
+                Name::Asc(field) => {
+                    Box::new(AscDesc::asc(&self.index, &self.rtxn, criterion, field)?)
+                }
+                Name::Desc(field) => {
+                    Box::new(AscDesc::desc(&self.index, &self.rtxn, criterion, field)?)
+                }
             };
         }
 
@@ -231,21 +304,20 @@ pub fn resolve_query_tree<'t>(
     ctx: &'t dyn Context,
     query_tree: &Operation,
     wdcache: &mut WordDerivationsCache,
-) -> Result<RoaringBitmap>
-{
+) -> Result<RoaringBitmap> {
     fn resolve_operation<'t>(
         ctx: &'t dyn Context,
         query_tree: &Operation,
         wdcache: &mut WordDerivationsCache,
-    ) -> Result<RoaringBitmap>
-    {
-        use Operation::{And, Phrase, Or, Query};
+    ) -> Result<RoaringBitmap> {
+        use Operation::{And, Or, Phrase, Query};
 
         match query_tree {
             And(ops) => {
-                let mut ops = ops.iter().map(|op| {
-                    resolve_operation(ctx, op, wdcache)
-                }).collect::<Result<Vec<_>>>()?;
+                let mut ops = ops
+                    .iter()
+                    .map(|op| resolve_operation(ctx, op, wdcache))
+                    .collect::<Result<Vec<_>>>()?;
 
                 ops.sort_unstable_by_key(|cds| cds.len());
 
@@ -260,7 +332,7 @@ pub fn resolve_query_tree<'t>(
                     }
                 }
                 Ok(candidates)
-            },
+            }
             Phrase(words) => {
                 let mut candidates = RoaringBitmap::new();
                 let mut first_loop = true;
@@ -276,12 +348,12 @@ pub fn resolve_query_tree<'t>(
                             } else {
                                 candidates &= pair_docids;
                             }
-                        },
-                        None => return Ok(RoaringBitmap::new())
+                        }
+                        None => return Ok(RoaringBitmap::new()),
                     }
                 }
                 Ok(candidates)
-            },
+            }
             Or(_, ops) => {
                 let mut candidates = RoaringBitmap::new();
                 for op in ops {
@@ -289,7 +361,7 @@ pub fn resolve_query_tree<'t>(
                     candidates.union_with(&docids);
                 }
                 Ok(candidates)
-            },
+            }
             Query(q) => Ok(query_docids(ctx, q, wdcache)?),
         }
     }
@@ -297,18 +369,18 @@ pub fn resolve_query_tree<'t>(
     resolve_operation(ctx, query_tree, wdcache)
 }
 
-
 fn all_word_pair_proximity_docids<T: AsRef<str>, U: AsRef<str>>(
     ctx: &dyn Context,
     left_words: &[(T, u8)],
     right_words: &[(U, u8)],
-    proximity: u8
-) -> Result<RoaringBitmap>
-{
+    proximity: u8,
+) -> Result<RoaringBitmap> {
     let mut docids = RoaringBitmap::new();
     for (left, _l_typo) in left_words {
         for (right, _r_typo) in right_words {
-            let current_docids = ctx.word_pair_proximity_docids(left.as_ref(), right.as_ref(), proximity)?.unwrap_or_default();
+            let current_docids = ctx
+                .word_pair_proximity_docids(left.as_ref(), right.as_ref(), proximity)?
+                .unwrap_or_default();
             docids.union_with(&current_docids);
         }
     }
@@ -319,8 +391,7 @@ fn query_docids(
     ctx: &dyn Context,
     query: &Query,
     wdcache: &mut WordDerivationsCache,
-) -> Result<RoaringBitmap>
-{
+) -> Result<RoaringBitmap> {
     match &query.kind {
         QueryKind::Exact { word, .. } => {
             if query.prefix && ctx.in_prefix_cache(&word) {
@@ -336,7 +407,7 @@ fn query_docids(
             } else {
                 Ok(ctx.word_docids(&word)?.unwrap_or_default())
             }
-        },
+        }
         QueryKind::Tolerant { typo, word } => {
             let words = word_derivations(&word, query.prefix, *typo, ctx.words_fst(), wdcache)?;
             let mut docids = RoaringBitmap::new();
@@ -345,7 +416,7 @@ fn query_docids(
                 docids.union_with(&current_docids);
             }
             Ok(docids)
-        },
+        }
     }
 }
 
@@ -355,8 +426,7 @@ fn query_pair_proximity_docids(
     right: &Query,
     proximity: u8,
     wdcache: &mut WordDerivationsCache,
-) -> Result<RoaringBitmap>
-{
+) -> Result<RoaringBitmap> {
     if proximity >= 8 {
         let mut candidates = query_docids(ctx, left, wdcache)?;
         let right_candidates = query_docids(ctx, right, wdcache)?;
@@ -368,20 +438,31 @@ fn query_pair_proximity_docids(
     match (&left.kind, &right.kind) {
         (QueryKind::Exact { word: left, .. }, QueryKind::Exact { word: right, .. }) => {
             if prefix && ctx.in_prefix_cache(&right) {
-                Ok(ctx.word_prefix_pair_proximity_docids(left.as_str(), right.as_str(), proximity)?.unwrap_or_default())
+                Ok(ctx
+                    .word_prefix_pair_proximity_docids(left.as_str(), right.as_str(), proximity)?
+                    .unwrap_or_default())
             } else if prefix {
                 let r_words = word_derivations(&right, true, 0, ctx.words_fst(), wdcache)?;
                 all_word_pair_proximity_docids(ctx, &[(left, 0)], &r_words, proximity)
             } else {
-                Ok(ctx.word_pair_proximity_docids(left.as_str(), right.as_str(), proximity)?.unwrap_or_default())
+                Ok(ctx
+                    .word_pair_proximity_docids(left.as_str(), right.as_str(), proximity)?
+                    .unwrap_or_default())
             }
-        },
+        }
         (QueryKind::Tolerant { typo, word: left }, QueryKind::Exact { word: right, .. }) => {
-            let l_words = word_derivations(&left, false, *typo, ctx.words_fst(), wdcache)?.to_owned();
+            let l_words =
+                word_derivations(&left, false, *typo, ctx.words_fst(), wdcache)?.to_owned();
             if prefix && ctx.in_prefix_cache(&right) {
                 let mut docids = RoaringBitmap::new();
                 for (left, _) in l_words {
-                    let current_docids = ctx.word_prefix_pair_proximity_docids(left.as_ref(), right.as_ref(), proximity)?.unwrap_or_default();
+                    let current_docids = ctx
+                        .word_prefix_pair_proximity_docids(
+                            left.as_ref(),
+                            right.as_ref(),
+                            proximity,
+                        )?
+                        .unwrap_or_default();
                     docids.union_with(&current_docids);
                 }
                 Ok(docids)
@@ -391,28 +472,36 @@ fn query_pair_proximity_docids(
             } else {
                 all_word_pair_proximity_docids(ctx, &l_words, &[(right, 0)], proximity)
             }
-        },
+        }
         (QueryKind::Exact { word: left, .. }, QueryKind::Tolerant { typo, word: right }) => {
             let r_words = word_derivations(&right, prefix, *typo, ctx.words_fst(), wdcache)?;
             all_word_pair_proximity_docids(ctx, &[(left, 0)], &r_words, proximity)
-        },
-        (QueryKind::Tolerant { typo: l_typo, word: left }, QueryKind::Tolerant { typo: r_typo, word: right }) => {
-            let l_words = word_derivations(&left, false, *l_typo, ctx.words_fst(), wdcache)?.to_owned();
+        }
+        (
+            QueryKind::Tolerant { typo: l_typo, word: left },
+            QueryKind::Tolerant { typo: r_typo, word: right },
+        ) => {
+            let l_words =
+                word_derivations(&left, false, *l_typo, ctx.words_fst(), wdcache)?.to_owned();
             let r_words = word_derivations(&right, prefix, *r_typo, ctx.words_fst(), wdcache)?;
             all_word_pair_proximity_docids(ctx, &l_words, &r_words, proximity)
-        },
+        }
     }
 }
 
 #[cfg(test)]
 pub mod test {
-    use maplit::hashmap;
-    use rand::{Rng, SeedableRng, rngs::StdRng};
-
-    use super::*;
     use std::collections::HashMap;
 
-    fn s(s: &str) -> String { s.to_string() }
+    use maplit::hashmap;
+    use rand::rngs::StdRng;
+    use rand::{Rng, SeedableRng};
+
+    use super::*;
+
+    fn s(s: &str) -> String {
+        s.to_string()
+    }
     pub struct TestContext<'t> {
         words_fst: fst::Set<Cow<'t, [u8]>>,
         word_docids: HashMap<String, RoaringBitmap>,
@@ -435,12 +524,22 @@ pub mod test {
             Ok(self.word_prefix_docids.get(&word.to_string()).cloned())
         }
 
-        fn word_pair_proximity_docids(&self, left: &str, right: &str, proximity: u8) -> heed::Result<Option<RoaringBitmap>> {
+        fn word_pair_proximity_docids(
+            &self,
+            left: &str,
+            right: &str,
+            proximity: u8,
+        ) -> heed::Result<Option<RoaringBitmap>> {
             let key = (left.to_string(), right.to_string(), proximity.into());
             Ok(self.word_pair_proximity_docids.get(&key).cloned())
         }
 
-        fn word_prefix_pair_proximity_docids(&self, left: &str, right: &str, proximity: u8) -> heed::Result<Option<RoaringBitmap>> {
+        fn word_prefix_pair_proximity_docids(
+            &self,
+            left: &str,
+            right: &str,
+            proximity: u8,
+        ) -> heed::Result<Option<RoaringBitmap>> {
             let key = (left.to_string(), right.to_string(), proximity.into());
             Ok(self.word_prefix_pair_proximity_docids.get(&key).cloned())
         }
@@ -453,24 +552,44 @@ pub mod test {
             self.word_prefix_docids.contains_key(&word.to_string())
         }
 
-        fn docid_words_positions(&self, docid: DocumentId) -> heed::Result<HashMap<String, RoaringBitmap>> {
+        fn docid_words_positions(
+            &self,
+            docid: DocumentId,
+        ) -> heed::Result<HashMap<String, RoaringBitmap>> {
             if let Some(docid_words) = self.docid_words.get(&docid) {
                 Ok(docid_words
                     .iter()
                     .enumerate()
-                    .map(|(i,w)| (w.clone(), RoaringBitmap::from_sorted_iter(std::iter::once(i as u32))))
-                    .collect()
-                )
+                    .map(|(i, w)| {
+                        (w.clone(), RoaringBitmap::from_sorted_iter(std::iter::once(i as u32)))
+                    })
+                    .collect())
             } else {
                 Ok(HashMap::new())
             }
         }
 
-        fn word_position_iterator(&self, _word: &str, _level: TreeLevel, _in_prefix_cache: bool, _left: Option<u32>, _right: Option<u32>) -> heed::Result<Box<dyn Iterator<Item =heed::Result<((&'c str, TreeLevel, u32, u32), RoaringBitmap)>> + 'c>> {
+        fn word_position_iterator(
+            &self,
+            _word: &str,
+            _level: TreeLevel,
+            _in_prefix_cache: bool,
+            _left: Option<u32>,
+            _right: Option<u32>,
+        ) -> heed::Result<
+            Box<
+                dyn Iterator<Item = heed::Result<((&'c str, TreeLevel, u32, u32), RoaringBitmap)>>
+                    + 'c,
+            >,
+        > {
             todo!()
         }
 
-        fn word_position_last_level(&self, _word: &str, _in_prefix_cache: bool) -> heed::Result<Option<TreeLevel>> {
+        fn word_position_last_level(
+            &self,
+            _word: &str,
+            _in_prefix_cache: bool,
+        ) -> heed::Result<Option<TreeLevel>> {
             todo!()
         }
 
@@ -478,15 +597,25 @@ pub mod test {
             todo!()
         }
 
-        fn searchable_fields_ids(&self) ->  Result<Vec<FieldId>> {
+        fn searchable_fields_ids(&self) -> Result<Vec<FieldId>> {
             todo!()
         }
 
-        fn word_level_position_docids(&self, _word: &str, _level: TreeLevel, _left: u32, _right: u32) -> heed::Result<Option<RoaringBitmap>> {
+        fn word_level_position_docids(
+            &self,
+            _word: &str,
+            _level: TreeLevel,
+            _left: u32,
+            _right: u32,
+        ) -> heed::Result<Option<RoaringBitmap>> {
             todo!()
         }
 
-        fn field_id_word_count_docids(&self, _field_id: FieldId, _word_count: u8) -> heed::Result<Option<RoaringBitmap>> {
+        fn field_id_word_count_docids(
+            &self,
+            _field_id: FieldId,
+            _word_count: u8,
+        ) -> heed::Result<Option<RoaringBitmap>> {
             todo!()
         }
     }
@@ -506,7 +635,7 @@ pub mod test {
                 RoaringBitmap::from_sorted_iter(values.into_iter())
             }
 
-            let word_docids = hashmap!{
+            let word_docids = hashmap! {
                 s("hello")      => random_postings(rng,   1500),
                 s("hi")         => random_postings(rng,   4000),
                 s("word")       => random_postings(rng,   2500),
@@ -530,7 +659,7 @@ pub mod test {
                 }
             }
 
-            let word_prefix_docids = hashmap!{
+            let word_prefix_docids = hashmap! {
                 s("h")   => &word_docids[&s("hello")] | &word_docids[&s("hi")],
                 s("wor") => &word_docids[&s("word")]  | &word_docids[&s("world")],
                 s("20")  => &word_docids[&s("2020")]  | &word_docids[&s("2021")],
@@ -540,7 +669,9 @@ pub mod test {
             let mut word_prefix_pair_proximity_docids = HashMap::new();
             for (lword, lcandidates) in &word_docids {
                 for (rword, rcandidates) in &word_docids {
-                    if lword == rword { continue }
+                    if lword == rword {
+                        continue;
+                    }
                     let candidates = lcandidates & rcandidates;
                     for candidate in candidates {
                         if let Some(docid_words) = docid_words.get(&candidate) {
@@ -551,24 +682,31 @@ pub mod test {
                             } else {
                                 (s(lword), s(rword), (lposition - rposition + 1) as i32)
                             };
-                            let docids = word_pair_proximity_docids.entry(key).or_insert(RoaringBitmap::new());
+                            let docids = word_pair_proximity_docids
+                                .entry(key)
+                                .or_insert(RoaringBitmap::new());
                             docids.push(candidate);
                         }
                     }
                 }
                 for (pword, pcandidates) in &word_prefix_docids {
-                    if lword.starts_with(pword) { continue }
+                    if lword.starts_with(pword) {
+                        continue;
+                    }
                     let candidates = lcandidates & pcandidates;
                     for candidate in candidates {
                         if let Some(docid_words) = docid_words.get(&candidate) {
                             let lposition = docid_words.iter().position(|w| w == lword).unwrap();
-                            let rposition = docid_words.iter().position(|w| w.starts_with(pword)).unwrap();
+                            let rposition =
+                                docid_words.iter().position(|w| w.starts_with(pword)).unwrap();
                             let key = if lposition < rposition {
                                 (s(lword), s(pword), (rposition - lposition) as i32)
                             } else {
                                 (s(lword), s(pword), (lposition - rposition + 1) as i32)
                             };
-                            let docids = word_prefix_pair_proximity_docids.entry(key).or_insert(RoaringBitmap::new());
+                            let docids = word_prefix_pair_proximity_docids
+                                .entry(key)
+                                .or_insert(RoaringBitmap::new());
                             docids.push(candidate);
                         }
                     }
diff --git a/milli/src/search/criteria/proximity.rs b/milli/src/search/criteria/proximity.rs
index c3c8027cb..3e8196e93 100644
--- a/milli/src/search/criteria/proximity.rs
+++ b/milli/src/search/criteria/proximity.rs
@@ -2,22 +2,16 @@ use std::collections::btree_map::{self, BTreeMap};
 use std::collections::hash_map::HashMap;
 use std::mem::take;
 
-use roaring::RoaringBitmap;
 use log::debug;
+use roaring::RoaringBitmap;
 
-use crate::search::query_tree::{maximum_proximity, Operation, Query};
-use crate::search::{build_dfa, WordDerivationsCache};
-use crate::search::{query_tree::QueryKind};
-use crate::{DocumentId, Position, Result};
 use super::{
-    Context,
-    Criterion,
-    CriterionParameters,
-    CriterionResult,
-    query_docids,
-    query_pair_proximity_docids,
-    resolve_query_tree,
+    query_docids, query_pair_proximity_docids, resolve_query_tree, Context, Criterion,
+    CriterionParameters, CriterionResult,
 };
+use crate::search::query_tree::{maximum_proximity, Operation, Query, QueryKind};
+use crate::search::{build_dfa, WordDerivationsCache};
+use crate::{DocumentId, Position, Result};
 
 type Cache = HashMap<(Operation, u8), Vec<(Query, Query, RoaringBitmap)>>;
 
@@ -63,28 +57,33 @@ impl<'t> Criterion for Proximity<'t> {
         }
 
         loop {
-            debug!("Proximity at iteration {} (max prox {:?}) ({:?})",
+            debug!(
+                "Proximity at iteration {} (max prox {:?}) ({:?})",
                 self.proximity,
                 self.state.as_ref().map(|(mp, _, _)| mp),
                 self.state.as_ref().map(|(_, _, cd)| cd),
             );
 
             match &mut self.state {
-                Some((max_prox, _, allowed_candidates)) if allowed_candidates.is_empty() || self.proximity > *max_prox => {
+                Some((max_prox, _, allowed_candidates))
+                    if allowed_candidates.is_empty() || self.proximity > *max_prox =>
+                {
                     self.state = None; // reset state
-                },
+                }
                 Some((_, query_tree, allowed_candidates)) => {
-                    let mut new_candidates = if allowed_candidates.len() <= CANDIDATES_THRESHOLD && self.proximity > PROXIMITY_THRESHOLD {
+                    let mut new_candidates = if allowed_candidates.len() <= CANDIDATES_THRESHOLD
+                        && self.proximity > PROXIMITY_THRESHOLD
+                    {
                         if let Some(cache) = self.plane_sweep_cache.as_mut() {
                             match cache.next() {
                                 Some((p, candidates)) => {
                                     self.proximity = p;
                                     candidates
-                                },
+                                }
                                 None => {
                                     self.state = None; // reset state
-                                    continue
-                                },
+                                    continue;
+                                }
                             }
                         } else {
                             let cache = resolve_plane_sweep_candidates(
@@ -95,9 +94,10 @@ impl<'t> Criterion for Proximity<'t> {
                             )?;
                             self.plane_sweep_cache = Some(cache.into_iter());
 
-                            continue
+                            continue;
                         }
-                    } else { // use set theory based algorithm
+                    } else {
+                        // use set theory based algorithm
                         resolve_candidates(
                             self.ctx,
                             &query_tree,
@@ -117,39 +117,50 @@ impl<'t> Criterion for Proximity<'t> {
                         filtered_candidates: None,
                         bucket_candidates: Some(take(&mut self.bucket_candidates)),
                     }));
-                },
-                None => {
-                    match self.parent.next(params)? {
-                        Some(CriterionResult { query_tree: Some(query_tree), candidates, filtered_candidates, bucket_candidates }) => {
-                            let mut candidates = match candidates {
-                                Some(candidates) => candidates,
-                                None => resolve_query_tree(self.ctx, &query_tree, params.wdcache)? - params.excluded_candidates,
-                            };
-
-                            if let Some(filtered_candidates) = filtered_candidates {
-                                candidates &= filtered_candidates;
+                }
+                None => match self.parent.next(params)? {
+                    Some(CriterionResult {
+                        query_tree: Some(query_tree),
+                        candidates,
+                        filtered_candidates,
+                        bucket_candidates,
+                    }) => {
+                        let mut candidates = match candidates {
+                            Some(candidates) => candidates,
+                            None => {
+                                resolve_query_tree(self.ctx, &query_tree, params.wdcache)?
+                                    - params.excluded_candidates
                             }
+                        };
 
-                            match bucket_candidates {
-                                Some(bucket_candidates) => self.bucket_candidates |= bucket_candidates,
-                                None => self.bucket_candidates |= &candidates,
-                            }
+                        if let Some(filtered_candidates) = filtered_candidates {
+                            candidates &= filtered_candidates;
+                        }
 
-                            let maximum_proximity = maximum_proximity(&query_tree);
-                            self.state = Some((maximum_proximity as u8, query_tree, candidates));
-                            self.proximity = 0;
-                            self.plane_sweep_cache = None;
-                        },
-                        Some(CriterionResult { query_tree: None, candidates, filtered_candidates, bucket_candidates }) => {
-                            return Ok(Some(CriterionResult {
-                                query_tree: None,
-                                candidates,
-                                filtered_candidates,
-                                bucket_candidates,
-                            }));
-                        },
-                        None => return Ok(None),
+                        match bucket_candidates {
+                            Some(bucket_candidates) => self.bucket_candidates |= bucket_candidates,
+                            None => self.bucket_candidates |= &candidates,
+                        }
+
+                        let maximum_proximity = maximum_proximity(&query_tree);
+                        self.state = Some((maximum_proximity as u8, query_tree, candidates));
+                        self.proximity = 0;
+                        self.plane_sweep_cache = None;
                     }
+                    Some(CriterionResult {
+                        query_tree: None,
+                        candidates,
+                        filtered_candidates,
+                        bucket_candidates,
+                    }) => {
+                        return Ok(Some(CriterionResult {
+                            query_tree: None,
+                            candidates,
+                            filtered_candidates,
+                            bucket_candidates,
+                        }));
+                    }
+                    None => return Ok(None),
                 },
             }
         }
@@ -162,46 +173,48 @@ fn resolve_candidates<'t>(
     proximity: u8,
     cache: &mut Cache,
     wdcache: &mut WordDerivationsCache,
-) -> Result<RoaringBitmap>
-{
+) -> Result<RoaringBitmap> {
     fn resolve_operation<'t>(
         ctx: &'t dyn Context,
         query_tree: &Operation,
         proximity: u8,
         cache: &mut Cache,
         wdcache: &mut WordDerivationsCache,
-    ) -> Result<Vec<(Query, Query, RoaringBitmap)>>
-    {
-        use Operation::{And, Phrase, Or};
+    ) -> Result<Vec<(Query, Query, RoaringBitmap)>> {
+        use Operation::{And, Or, Phrase};
 
         let result = match query_tree {
             And(ops) => mdfs(ctx, ops, proximity, cache, wdcache)?,
-            Phrase(words) => if proximity == 0 {
-                let most_left = words.first().map(|w| Query { prefix: false, kind: QueryKind::exact(w.clone()) });
-                let most_right = words.last().map(|w| Query { prefix: false, kind: QueryKind::exact(w.clone()) });
-                let mut candidates = None;
-                for slice in words.windows(2) {
-                    let (left, right) = (&slice[0], &slice[1]);
-                    match ctx.word_pair_proximity_docids(left, right, 1)? {
-                        Some(pair_docids) => {
-                            match candidates.as_mut() {
+            Phrase(words) => {
+                if proximity == 0 {
+                    let most_left = words
+                        .first()
+                        .map(|w| Query { prefix: false, kind: QueryKind::exact(w.clone()) });
+                    let most_right = words
+                        .last()
+                        .map(|w| Query { prefix: false, kind: QueryKind::exact(w.clone()) });
+                    let mut candidates = None;
+                    for slice in words.windows(2) {
+                        let (left, right) = (&slice[0], &slice[1]);
+                        match ctx.word_pair_proximity_docids(left, right, 1)? {
+                            Some(pair_docids) => match candidates.as_mut() {
                                 Some(candidates) => *candidates &= pair_docids,
                                 None => candidates = Some(pair_docids),
+                            },
+                            None => {
+                                candidates = None;
+                                break;
                             }
-                        },
-                        None => {
-                            candidates = None;
-                            break;
                         }
                     }
+                    match (most_left, most_right, candidates) {
+                        (Some(l), Some(r), Some(c)) => vec![(l, r, c)],
+                        _otherwise => Default::default(),
+                    }
+                } else {
+                    Default::default()
                 }
-                match (most_left, most_right, candidates) {
-                    (Some(l), Some(r), Some(c)) => vec![(l, r, c)],
-                    _otherwise => Default::default(),
-                }
-            } else {
-                Default::default()
-            },
+            }
             Or(_, ops) => {
                 let mut output = Vec::new();
                 for op in ops {
@@ -209,13 +222,15 @@ fn resolve_candidates<'t>(
                     output.extend(result);
                 }
                 output
-            },
-            Operation::Query(q) => if proximity == 0 {
-                let candidates = query_docids(ctx, q, wdcache)?;
-                vec![(q.clone(), q.clone(), candidates)]
-            } else {
-                Default::default()
-            },
+            }
+            Operation::Query(q) => {
+                if proximity == 0 {
+                    let candidates = query_docids(ctx, q, wdcache)?;
+                    vec![(q.clone(), q.clone(), candidates)]
+                } else {
+                    Default::default()
+                }
+            }
         };
 
         Ok(result)
@@ -228,8 +243,7 @@ fn resolve_candidates<'t>(
         proximity: u8,
         cache: &mut Cache,
         wdcache: &mut WordDerivationsCache,
-    ) -> Result<Vec<(Query, Query, RoaringBitmap)>>
-    {
+    ) -> Result<Vec<(Query, Query, RoaringBitmap)>> {
         fn pair_combinations(mana: u8, left_max: u8) -> impl Iterator<Item = (u8, u8)> {
             (0..=mana.min(left_max)).map(move |m| (m, mana - m))
         }
@@ -257,7 +271,8 @@ fn resolve_candidates<'t>(
 
                 for (ll, lr, lcandidates) in lefts {
                     for (rl, rr, rcandidates) in rights {
-                        let mut candidates = query_pair_proximity_docids(ctx, lr, rl, pair_p + 1, wdcache)?;
+                        let mut candidates =
+                            query_pair_proximity_docids(ctx, lr, rl, pair_p + 1, wdcache)?;
                         if lcandidates.len() < rcandidates.len() {
                             candidates.intersect_with(lcandidates);
                             candidates.intersect_with(rcandidates);
@@ -282,22 +297,26 @@ fn resolve_candidates<'t>(
         proximity: u8,
         cache: &mut Cache,
         wdcache: &mut WordDerivationsCache,
-    ) -> Result<Vec<(Query, Query, RoaringBitmap)>>
-    {
+    ) -> Result<Vec<(Query, Query, RoaringBitmap)>> {
         // Extract the first two elements but gives the tail
         // that is just after the first element.
-        let next = branches.split_first().map(|(h1, t)| {
-            (h1, t.split_first().map(|(h2, _)| (h2, t)))
-        });
+        let next =
+            branches.split_first().map(|(h1, t)| (h1, t.split_first().map(|(h2, _)| (h2, t))));
 
         match next {
-            Some((head1, Some((head2, [_])))) => mdfs_pair(ctx, head1, head2, proximity, cache, wdcache),
+            Some((head1, Some((head2, [_])))) => {
+                mdfs_pair(ctx, head1, head2, proximity, cache, wdcache)
+            }
             Some((head1, Some((head2, tail)))) => {
                 let mut output = Vec::new();
                 for p in 0..=proximity {
-                    for (lhead, _, head_candidates) in mdfs_pair(ctx, head1, head2, p, cache, wdcache)? {
+                    for (lhead, _, head_candidates) in
+                        mdfs_pair(ctx, head1, head2, p, cache, wdcache)?
+                    {
                         if !head_candidates.is_empty() {
-                            for (_, rtail, mut candidates) in mdfs(ctx, tail, proximity - p, cache, wdcache)? {
+                            for (_, rtail, mut candidates) in
+                                mdfs(ctx, tail, proximity - p, cache, wdcache)?
+                            {
                                 candidates.intersect_with(&head_candidates);
                                 if !candidates.is_empty() {
                                     output.push((lhead.clone(), rtail, candidates));
@@ -307,7 +326,7 @@ fn resolve_candidates<'t>(
                     }
                 }
                 Ok(output)
-            },
+            }
             Some((head1, None)) => resolve_operation(ctx, head1, proximity, cache, wdcache),
             None => Ok(Default::default()),
         }
@@ -325,47 +344,48 @@ fn resolve_plane_sweep_candidates(
     query_tree: &Operation,
     allowed_candidates: &RoaringBitmap,
     wdcache: &mut WordDerivationsCache,
-) -> Result<BTreeMap<u8, RoaringBitmap>>
-{
+) -> Result<BTreeMap<u8, RoaringBitmap>> {
     /// FIXME may be buggy with query like "new new york"
     fn plane_sweep(
         groups_positions: Vec<Vec<(Position, u8, Position)>>,
         consecutive: bool,
-    ) -> Result<Vec<(Position, u8, Position)>>
-    {
+    ) -> Result<Vec<(Position, u8, Position)>> {
         fn compute_groups_proximity(
             groups: &[(usize, (Position, u8, Position))],
             consecutive: bool,
-        ) -> Option<(Position, u8, Position)>
-        {
+        ) -> Option<(Position, u8, Position)> {
             // take the inner proximity of the first group as initial
             let (_, (_, mut proximity, _)) = groups.first()?;
             let (_, (left_most_pos, _, _)) = groups.first()?;
-            let (_, (_, _, right_most_pos)) = groups.iter().max_by_key(|(_, (_, _, right_most_pos))| right_most_pos)?;
+            let (_, (_, _, right_most_pos)) =
+                groups.iter().max_by_key(|(_, (_, _, right_most_pos))| right_most_pos)?;
 
             for pair in groups.windows(2) {
                 if let [(i1, (lpos1, _, rpos1)), (i2, (lpos2, prox2, rpos2))] = pair {
                     // if two positions are equal, meaning that they share at least a word, we return None
                     if rpos1 == rpos2 || lpos1 == lpos2 || rpos1 == lpos2 || lpos1 == rpos2 {
-                        return None
+                        return None;
                     }
 
                     let pair_proximity = {
                         // if intervals are disjoint [..].(..)
-                        if lpos2 > rpos1 { lpos2 - rpos1 }
+                        if lpos2 > rpos1 {
+                            lpos2 - rpos1
+                        }
                         // if the second interval is a subset of the first [.(..).]
-                        else if rpos2 < rpos1 { (lpos2 - lpos1).min(rpos1 - rpos2) }
+                        else if rpos2 < rpos1 {
+                            (lpos2 - lpos1).min(rpos1 - rpos2)
+                        }
                         // if intervals overlaps [.(..].)
-                        else { (lpos2 - lpos1).min(rpos2 - rpos1) }
+                        else {
+                            (lpos2 - lpos1).min(rpos2 - rpos1)
+                        }
                     };
 
                     // if groups are in the good order (query order) we remove 1 to the proximity
                     // the proximity is clamped to 7
-                    let pair_proximity = if i1 < i2 {
-                        (pair_proximity - 1).min(7)
-                    } else {
-                        pair_proximity.min(7)
-                    };
+                    let pair_proximity =
+                        if i1 < i2 { (pair_proximity - 1).min(7) } else { pair_proximity.min(7) };
 
                     proximity += pair_proximity as u8 + prox2;
                 }
@@ -381,7 +401,8 @@ fn resolve_plane_sweep_candidates(
 
         let groups_len = groups_positions.len();
 
-        let mut groups_positions: Vec<_> = groups_positions.into_iter().map(|pos| pos.into_iter()).collect();
+        let mut groups_positions: Vec<_> =
+            groups_positions.into_iter().map(|pos| pos.into_iter()).collect();
 
         // Pop top elements of each list.
         let mut current = Vec::with_capacity(groups_len);
@@ -452,9 +473,8 @@ fn resolve_plane_sweep_candidates(
         rocache: &mut HashMap<&'a Operation, Vec<(Position, u8, Position)>>,
         words_positions: &HashMap<String, RoaringBitmap>,
         wdcache: &mut WordDerivationsCache,
-    ) -> Result<Vec<(Position, u8, Position)>>
-    {
-        use Operation::{And, Phrase, Or};
+    ) -> Result<Vec<(Position, u8, Position)>> {
+        use Operation::{And, Or, Phrase};
 
         if let Some(result) = rocache.get(query_tree) {
             return Ok(result.clone());
@@ -462,13 +482,20 @@ fn resolve_plane_sweep_candidates(
 
         let result = match query_tree {
             And(ops) => {
-                 let mut groups_positions = Vec::with_capacity(ops.len());
+                let mut groups_positions = Vec::with_capacity(ops.len());
                 for operation in ops {
-                    let positions = resolve_operation(ctx, operation, docid, rocache, words_positions, wdcache)?;
+                    let positions = resolve_operation(
+                        ctx,
+                        operation,
+                        docid,
+                        rocache,
+                        words_positions,
+                        wdcache,
+                    )?;
                     groups_positions.push(positions);
                 }
                 plane_sweep(groups_positions, false)?
-            },
+            }
             Phrase(words) => {
                 let mut groups_positions = Vec::with_capacity(words.len());
                 for word in words {
@@ -479,16 +506,23 @@ fn resolve_plane_sweep_candidates(
                     groups_positions.push(positions);
                 }
                 plane_sweep(groups_positions, true)?
-            },
+            }
             Or(_, ops) => {
                 let mut result = Vec::new();
                 for op in ops {
-                    result.extend(resolve_operation(ctx, op, docid, rocache, words_positions, wdcache)?)
+                    result.extend(resolve_operation(
+                        ctx,
+                        op,
+                        docid,
+                        rocache,
+                        words_positions,
+                        wdcache,
+                    )?)
                 }
 
                 result.sort_unstable();
                 result
-            },
+            }
             Operation::Query(Query { prefix, kind }) => {
                 let mut result = Vec::new();
                 match kind {
@@ -498,9 +532,9 @@ fn resolve_plane_sweep_candidates(
                                 .flat_map(|positions| positions.iter().map(|p| (p, 0, p)));
                             result.extend(iter);
                         } else if let Some(positions) = words_positions.get(word) {
-                                result.extend(positions.iter().map(|p| (p, 0, p)));
+                            result.extend(positions.iter().map(|p| (p, 0, p)));
                         }
-                    },
+                    }
                     QueryKind::Tolerant { typo, word } => {
                         let iter = word_derivations(word, *prefix, *typo, &words_positions)
                             .flat_map(|positions| positions.iter().map(|p| (p, 0, p)));
@@ -522,8 +556,7 @@ fn resolve_plane_sweep_candidates(
         is_prefix: bool,
         max_typo: u8,
         words_positions: &'a HashMap<String, RoaringBitmap>,
-    ) -> impl Iterator<Item = &'a RoaringBitmap>
-    {
+    ) -> impl Iterator<Item = &'a RoaringBitmap> {
         let dfa = build_dfa(word, max_typo, is_prefix);
         words_positions.iter().filter_map(move |(document_word, positions)| {
             use levenshtein_automata::Distance;
@@ -539,7 +572,7 @@ fn resolve_plane_sweep_candidates(
     for docid in allowed_candidates {
         let words_positions = ctx.docid_words_positions(docid)?;
         resolve_operation_cache.clear();
-        let positions =  resolve_operation(
+        let positions = resolve_operation(
             ctx,
             query_tree,
             docid,
diff --git a/milli/src/search/criteria/typo.rs b/milli/src/search/criteria/typo.rs
index 436f4affd..f4ae15f0a 100644
--- a/milli/src/search/criteria/typo.rs
+++ b/milli/src/search/criteria/typo.rs
@@ -1,20 +1,17 @@
-use std::{borrow::Cow, collections::HashMap, mem::take};
+use std::borrow::Cow;
+use std::collections::HashMap;
+use std::mem::take;
 
 use log::debug;
 use roaring::RoaringBitmap;
 
+use super::{
+    query_docids, resolve_query_tree, Candidates, Context, Criterion, CriterionParameters,
+    CriterionResult,
+};
 use crate::search::query_tree::{maximum_typo, Operation, Query, QueryKind};
 use crate::search::{word_derivations, WordDerivationsCache};
 use crate::Result;
-use super::{
-    Candidates,
-    Context,
-    Criterion,
-    CriterionParameters,
-    CriterionResult,
-    query_docids,
-    resolve_query_tree,
-};
 
 /// Maximum number of typo for a word of any length.
 const MAX_TYPOS_PER_WORD: u8 = 2;
@@ -54,7 +51,8 @@ impl<'t> Criterion for Typo<'t> {
         }
 
         loop {
-            debug!("Typo at iteration {} (max typos {:?}) ({:?})",
+            debug!(
+                "Typo at iteration {} (max typos {:?}) ({:?})",
                 self.typos,
                 self.state.as_ref().map(|(mt, _, _)| mt),
                 self.state.as_ref().map(|(_, _, cd)| cd),
@@ -63,29 +61,42 @@ impl<'t> Criterion for Typo<'t> {
             match self.state.as_mut() {
                 Some((max_typos, _, _)) if self.typos > *max_typos => {
                     self.state = None; // reset state
-                },
+                }
                 Some((_, _, Allowed(allowed_candidates))) if allowed_candidates.is_empty() => {
                     self.state = None; // reset state
-                },
+                }
                 Some((_, query_tree, candidates_authorization)) => {
                     let fst = self.ctx.words_fst();
                     let new_query_tree = match self.typos {
-                        typos if typos < MAX_TYPOS_PER_WORD => {
-                            alterate_query_tree(&fst, query_tree.clone(), self.typos, params.wdcache)?
-                        },
+                        typos if typos < MAX_TYPOS_PER_WORD => alterate_query_tree(
+                            &fst,
+                            query_tree.clone(),
+                            self.typos,
+                            params.wdcache,
+                        )?,
                         MAX_TYPOS_PER_WORD => {
                             // When typos >= MAX_TYPOS_PER_WORD, no more alteration of the query tree is possible,
                             // we keep the altered query tree
-                            *query_tree = alterate_query_tree(&fst, query_tree.clone(), self.typos, params.wdcache)?;
+                            *query_tree = alterate_query_tree(
+                                &fst,
+                                query_tree.clone(),
+                                self.typos,
+                                params.wdcache,
+                            )?;
                             // we compute the allowed candidates
-                            let query_tree_allowed_candidates = resolve_query_tree(self.ctx, query_tree, params.wdcache)?;
+                            let query_tree_allowed_candidates =
+                                resolve_query_tree(self.ctx, query_tree, params.wdcache)?;
                             // we assign the allowed candidates to the candidates authorization.
                             *candidates_authorization = match take(candidates_authorization) {
-                                Allowed(allowed_candidates) => Allowed(query_tree_allowed_candidates & allowed_candidates),
-                                Forbidden(forbidden_candidates) => Allowed(query_tree_allowed_candidates - forbidden_candidates),
+                                Allowed(allowed_candidates) => {
+                                    Allowed(query_tree_allowed_candidates & allowed_candidates)
+                                }
+                                Forbidden(forbidden_candidates) => {
+                                    Allowed(query_tree_allowed_candidates - forbidden_candidates)
+                                }
                             };
                             query_tree.clone()
-                        },
+                        }
                         _otherwise => query_tree.clone(),
                     };
 
@@ -101,11 +112,11 @@ impl<'t> Criterion for Typo<'t> {
                         Allowed(allowed_candidates) => {
                             candidates &= &*allowed_candidates;
                             *allowed_candidates -= &candidates;
-                        },
+                        }
                         Forbidden(forbidden_candidates) => {
                             candidates -= &*forbidden_candidates;
                             *forbidden_candidates |= &candidates;
-                        },
+                        }
                     }
 
                     let bucket_candidates = match self.bucket_candidates.as_mut() {
@@ -121,35 +132,45 @@ impl<'t> Criterion for Typo<'t> {
                         filtered_candidates: None,
                         bucket_candidates: Some(bucket_candidates),
                     }));
-                },
-                None => {
-                    match self.parent.next(params)? {
-                        Some(CriterionResult { query_tree: Some(query_tree), candidates, filtered_candidates, bucket_candidates }) => {
-                            self.bucket_candidates = match (self.bucket_candidates.take(), bucket_candidates) {
+                }
+                None => match self.parent.next(params)? {
+                    Some(CriterionResult {
+                        query_tree: Some(query_tree),
+                        candidates,
+                        filtered_candidates,
+                        bucket_candidates,
+                    }) => {
+                        self.bucket_candidates =
+                            match (self.bucket_candidates.take(), bucket_candidates) {
                                 (Some(self_bc), Some(parent_bc)) => Some(self_bc | parent_bc),
                                 (self_bc, parent_bc) => self_bc.or(parent_bc),
                             };
 
-                            let candidates = match candidates.or(filtered_candidates) {
-                                Some(candidates) => Candidates::Allowed(candidates - params.excluded_candidates),
-                                None => Candidates::Forbidden(params.excluded_candidates.clone()),
-                            };
+                        let candidates = match candidates.or(filtered_candidates) {
+                            Some(candidates) => {
+                                Candidates::Allowed(candidates - params.excluded_candidates)
+                            }
+                            None => Candidates::Forbidden(params.excluded_candidates.clone()),
+                        };
 
-                            let maximum_typos = maximum_typo(&query_tree) as u8;
-                            self.state = Some((maximum_typos, query_tree, candidates));
-                            self.typos = 0;
-
-                        },
-                        Some(CriterionResult { query_tree: None, candidates, filtered_candidates, bucket_candidates }) => {
-                            return Ok(Some(CriterionResult {
-                                query_tree: None,
-                                candidates,
-                                filtered_candidates,
-                                bucket_candidates,
-                            }));
-                        },
-                        None => return Ok(None),
+                        let maximum_typos = maximum_typo(&query_tree) as u8;
+                        self.state = Some((maximum_typos, query_tree, candidates));
+                        self.typos = 0;
                     }
+                    Some(CriterionResult {
+                        query_tree: None,
+                        candidates,
+                        filtered_candidates,
+                        bucket_candidates,
+                    }) => {
+                        return Ok(Some(CriterionResult {
+                            query_tree: None,
+                            candidates,
+                            filtered_candidates,
+                            bucket_candidates,
+                        }));
+                    }
+                    None => return Ok(None),
                 },
             }
         }
@@ -164,21 +185,19 @@ fn alterate_query_tree(
     mut query_tree: Operation,
     number_typos: u8,
     wdcache: &mut WordDerivationsCache,
-) -> Result<Operation>
-{
+) -> Result<Operation> {
     fn recurse(
         words_fst: &fst::Set<Cow<[u8]>>,
         operation: &mut Operation,
         number_typos: u8,
         wdcache: &mut WordDerivationsCache,
-    ) -> Result<()>
-    {
-        use Operation::{And, Phrase, Or};
+    ) -> Result<()> {
+        use Operation::{And, Or, Phrase};
 
         match operation {
             And(ops) | Or(_, ops) => {
                 ops.iter_mut().try_for_each(|op| recurse(words_fst, op, number_typos, wdcache))
-            },
+            }
             // Because Phrases don't allow typos, no alteration can be done.
             Phrase(_words) => return Ok(()),
             Operation::Query(q) => {
@@ -193,19 +212,25 @@ fn alterate_query_tree(
                     } else {
                         let typo = *typo.min(&number_typos);
                         let words = word_derivations(word, q.prefix, typo, words_fst, wdcache)?;
-                        let queries = words.iter().map(|(word, typo)| {
-                            Operation::Query(Query {
-                                prefix: false,
-                                kind: QueryKind::Exact { original_typo: *typo, word: word.to_string() },
+                        let queries = words
+                            .iter()
+                            .map(|(word, typo)| {
+                                Operation::Query(Query {
+                                    prefix: false,
+                                    kind: QueryKind::Exact {
+                                        original_typo: *typo,
+                                        word: word.to_string(),
+                                    },
+                                })
                             })
-                        }).collect();
+                            .collect();
 
                         *operation = Operation::or(false, queries);
                     }
                 }
 
                 Ok(())
-            },
+            }
         }
     }
 
@@ -219,22 +244,18 @@ fn resolve_candidates<'t>(
     number_typos: u8,
     cache: &mut HashMap<(Operation, u8), RoaringBitmap>,
     wdcache: &mut WordDerivationsCache,
-) -> Result<RoaringBitmap>
-{
+) -> Result<RoaringBitmap> {
     fn resolve_operation<'t>(
         ctx: &'t dyn Context,
         query_tree: &Operation,
         number_typos: u8,
         cache: &mut HashMap<(Operation, u8), RoaringBitmap>,
         wdcache: &mut WordDerivationsCache,
-    ) -> Result<RoaringBitmap>
-    {
-        use Operation::{And, Phrase, Or, Query};
+    ) -> Result<RoaringBitmap> {
+        use Operation::{And, Or, Phrase, Query};
 
         match query_tree {
-            And(ops) => {
-                mdfs(ctx, ops, number_typos, cache, wdcache)
-            },
+            And(ops) => mdfs(ctx, ops, number_typos, cache, wdcache),
             Phrase(words) => {
                 let mut candidates = RoaringBitmap::new();
                 let mut first_loop = true;
@@ -250,12 +271,12 @@ fn resolve_candidates<'t>(
                             } else {
                                 candidates &= pair_docids;
                             }
-                        },
-                        None => return Ok(RoaringBitmap::new())
+                        }
+                        None => return Ok(RoaringBitmap::new()),
                     }
                 }
                 Ok(candidates)
-            },
+            }
             Or(_, ops) => {
                 let mut candidates = RoaringBitmap::new();
                 for op in ops {
@@ -263,12 +284,14 @@ fn resolve_candidates<'t>(
                     candidates.union_with(&docids);
                 }
                 Ok(candidates)
-            },
-            Query(q) => if q.kind.typo() == number_typos {
-                Ok(query_docids(ctx, q, wdcache)?)
-            } else {
-                Ok(RoaringBitmap::new())
-            },
+            }
+            Query(q) => {
+                if q.kind.typo() == number_typos {
+                    Ok(query_docids(ctx, q, wdcache)?)
+                } else {
+                    Ok(RoaringBitmap::new())
+                }
+            }
         }
     }
 
@@ -278,8 +301,7 @@ fn resolve_candidates<'t>(
         mana: u8,
         cache: &mut HashMap<(Operation, u8), RoaringBitmap>,
         wdcache: &mut WordDerivationsCache,
-    ) -> Result<RoaringBitmap>
-    {
+    ) -> Result<RoaringBitmap> {
         match branches.split_first() {
             Some((head, [])) => {
                 let cache_key = (head.clone(), mana);
@@ -290,7 +312,7 @@ fn resolve_candidates<'t>(
                     cache.insert(cache_key, candidates.clone());
                     Ok(candidates)
                 }
-            },
+            }
             Some((head, tail)) => {
                 let mut candidates = RoaringBitmap::new();
 
@@ -313,7 +335,7 @@ fn resolve_candidates<'t>(
                 }
 
                 Ok(candidates)
-            },
+            }
             None => Ok(RoaringBitmap::new()),
         }
     }
@@ -323,9 +345,9 @@ fn resolve_candidates<'t>(
 
 #[cfg(test)]
 mod test {
-    use super::*;
     use super::super::initial::Initial;
     use super::super::test::TestContext;
+    use super::*;
 
     #[test]
     fn initial_placeholder_no_facets() {
@@ -348,13 +370,23 @@ mod test {
     #[test]
     fn initial_query_tree_no_facets() {
         let context = TestContext::default();
-        let query_tree = Operation::Or(false, vec![
-            Operation::And(vec![
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact("split".to_string()) }),
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact("this".to_string()) }),
-                Operation::Query(Query { prefix: false, kind: QueryKind::tolerant(1, "world".to_string()) }),
-            ])
-        ]);
+        let query_tree = Operation::Or(
+            false,
+            vec![Operation::And(vec![
+                Operation::Query(Query {
+                    prefix: false,
+                    kind: QueryKind::exact("split".to_string()),
+                }),
+                Operation::Query(Query {
+                    prefix: false,
+                    kind: QueryKind::exact("this".to_string()),
+                }),
+                Operation::Query(Query {
+                    prefix: false,
+                    kind: QueryKind::tolerant(1, "world".to_string()),
+                }),
+            ])],
+        );
 
         let facet_candidates = None;
 
@@ -369,13 +401,23 @@ mod test {
             & context.word_docids("this").unwrap().unwrap()
             & context.word_docids("world").unwrap().unwrap();
         let expected_1 = CriterionResult {
-            query_tree: Some(Operation::Or(false, vec![
-                Operation::And(vec![
-                    Operation::Query(Query { prefix: false, kind: QueryKind::exact("split".to_string()) }),
-                    Operation::Query(Query { prefix: false, kind: QueryKind::exact("this".to_string()) }),
-                    Operation::Query(Query { prefix: false, kind: QueryKind::exact("world".to_string()) }),
-                ]),
-            ])),
+            query_tree: Some(Operation::Or(
+                false,
+                vec![Operation::And(vec![
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("split".to_string()),
+                    }),
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("this".to_string()),
+                    }),
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("world".to_string()),
+                    }),
+                ])],
+            )),
             candidates: Some(candidates_1.clone()),
             bucket_candidates: Some(candidates_1),
             filtered_candidates: None,
@@ -383,22 +425,37 @@ mod test {
 
         assert_eq!(criteria.next(&mut criterion_parameters).unwrap(), Some(expected_1));
 
-        let candidates_2 = (
-                context.word_docids("split").unwrap().unwrap()
-                & context.word_docids("this").unwrap().unwrap()
-                & context.word_docids("word").unwrap().unwrap()
-            ) - context.word_docids("world").unwrap().unwrap();
+        let candidates_2 = (context.word_docids("split").unwrap().unwrap()
+            & context.word_docids("this").unwrap().unwrap()
+            & context.word_docids("word").unwrap().unwrap())
+            - context.word_docids("world").unwrap().unwrap();
         let expected_2 = CriterionResult {
-            query_tree: Some(Operation::Or(false, vec![
-                Operation::And(vec![
-                    Operation::Query(Query { prefix: false, kind: QueryKind::exact("split".to_string()) }),
-                    Operation::Query(Query { prefix: false, kind: QueryKind::exact("this".to_string()) }),
-                    Operation::Or(false, vec![
-                        Operation::Query(Query { prefix: false, kind: QueryKind::exact_with_typo(1, "word".to_string()) }),
-                        Operation::Query(Query { prefix: false, kind: QueryKind::exact("world".to_string()) }),
-                    ]),
-                ]),
-            ])),
+            query_tree: Some(Operation::Or(
+                false,
+                vec![Operation::And(vec![
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("split".to_string()),
+                    }),
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("this".to_string()),
+                    }),
+                    Operation::Or(
+                        false,
+                        vec![
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::exact_with_typo(1, "word".to_string()),
+                            }),
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::exact("world".to_string()),
+                            }),
+                        ],
+                    ),
+                ])],
+            )),
             candidates: Some(candidates_2.clone()),
             bucket_candidates: Some(candidates_2),
             filtered_candidates: None,
@@ -437,17 +494,26 @@ mod test {
     #[test]
     fn initial_query_tree_with_facets() {
         let context = TestContext::default();
-        let query_tree = Operation::Or(false, vec![
-            Operation::And(vec![
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact("split".to_string()) }),
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact("this".to_string()) }),
-                Operation::Query(Query { prefix: false, kind: QueryKind::tolerant(1, "world".to_string()) }),
-            ])
-        ]);
+        let query_tree = Operation::Or(
+            false,
+            vec![Operation::And(vec![
+                Operation::Query(Query {
+                    prefix: false,
+                    kind: QueryKind::exact("split".to_string()),
+                }),
+                Operation::Query(Query {
+                    prefix: false,
+                    kind: QueryKind::exact("this".to_string()),
+                }),
+                Operation::Query(Query {
+                    prefix: false,
+                    kind: QueryKind::tolerant(1, "world".to_string()),
+                }),
+            ])],
+        );
 
         let facet_candidates = context.word_docids("earth").unwrap().unwrap();
 
-
         let mut criterion_parameters = CriterionParameters {
             wdcache: &mut WordDerivationsCache::new(),
             excluded_candidates: &RoaringBitmap::new(),
@@ -459,13 +525,23 @@ mod test {
             & context.word_docids("this").unwrap().unwrap()
             & context.word_docids("world").unwrap().unwrap();
         let expected_1 = CriterionResult {
-            query_tree: Some(Operation::Or(false, vec![
-                Operation::And(vec![
-                    Operation::Query(Query { prefix: false, kind: QueryKind::exact("split".to_string()) }),
-                    Operation::Query(Query { prefix: false, kind: QueryKind::exact("this".to_string()) }),
-                    Operation::Query(Query { prefix: false, kind: QueryKind::exact("world".to_string()) }),
-                ]),
-            ])),
+            query_tree: Some(Operation::Or(
+                false,
+                vec![Operation::And(vec![
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("split".to_string()),
+                    }),
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("this".to_string()),
+                    }),
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("world".to_string()),
+                    }),
+                ])],
+            )),
             candidates: Some(&candidates_1 & &facet_candidates),
             bucket_candidates: Some(&candidates_1 & &facet_candidates),
             filtered_candidates: None,
@@ -473,22 +549,37 @@ mod test {
 
         assert_eq!(criteria.next(&mut criterion_parameters).unwrap(), Some(expected_1));
 
-        let candidates_2 = (
-                context.word_docids("split").unwrap().unwrap()
-                & context.word_docids("this").unwrap().unwrap()
-                & context.word_docids("word").unwrap().unwrap()
-            ) - context.word_docids("world").unwrap().unwrap();
+        let candidates_2 = (context.word_docids("split").unwrap().unwrap()
+            & context.word_docids("this").unwrap().unwrap()
+            & context.word_docids("word").unwrap().unwrap())
+            - context.word_docids("world").unwrap().unwrap();
         let expected_2 = CriterionResult {
-            query_tree: Some(Operation::Or(false, vec![
-                Operation::And(vec![
-                    Operation::Query(Query { prefix: false, kind: QueryKind::exact("split".to_string()) }),
-                    Operation::Query(Query { prefix: false, kind: QueryKind::exact("this".to_string()) }),
-                    Operation::Or(false, vec![
-                        Operation::Query(Query { prefix: false, kind: QueryKind::exact_with_typo(1, "word".to_string()) }),
-                        Operation::Query(Query { prefix: false, kind: QueryKind::exact("world".to_string()) }),
-                    ]),
-                ]),
-            ])),
+            query_tree: Some(Operation::Or(
+                false,
+                vec![Operation::And(vec![
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("split".to_string()),
+                    }),
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("this".to_string()),
+                    }),
+                    Operation::Or(
+                        false,
+                        vec![
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::exact_with_typo(1, "word".to_string()),
+                            }),
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::exact("world".to_string()),
+                            }),
+                        ],
+                    ),
+                ])],
+            )),
             candidates: Some(&candidates_2 & &facet_candidates),
             bucket_candidates: Some(&candidates_2 & &facet_candidates),
             filtered_candidates: None,
diff --git a/milli/src/search/criteria/words.rs b/milli/src/search/criteria/words.rs
index add90d80d..ccc6c0617 100644
--- a/milli/src/search/criteria/words.rs
+++ b/milli/src/search/criteria/words.rs
@@ -3,9 +3,9 @@ use std::mem::take;
 use log::debug;
 use roaring::RoaringBitmap;
 
+use super::{resolve_query_tree, Context, Criterion, CriterionParameters, CriterionResult};
 use crate::search::query_tree::Operation;
 use crate::Result;
-use super::{Context, Criterion, CriterionParameters, CriterionResult, resolve_query_tree};
 
 pub struct Words<'t> {
     ctx: &'t dyn Context<'t>,
@@ -44,11 +44,12 @@ impl<'t> Criterion for Words<'t> {
                 Some(query_tree) => {
                     let candidates = match self.candidates.as_mut() {
                         Some(allowed_candidates) => {
-                            let mut candidates = resolve_query_tree(self.ctx, &query_tree, params.wdcache)?;
+                            let mut candidates =
+                                resolve_query_tree(self.ctx, &query_tree, params.wdcache)?;
                             candidates &= &*allowed_candidates;
                             *allowed_candidates -= &candidates;
                             Some(candidates)
-                        },
+                        }
                         None => None,
                     };
 
@@ -63,29 +64,38 @@ impl<'t> Criterion for Words<'t> {
                         filtered_candidates: self.filtered_candidates.clone(),
                         bucket_candidates,
                     }));
-                },
-                None => {
-                    match self.parent.next(params)? {
-                        Some(CriterionResult { query_tree: Some(query_tree), candidates, filtered_candidates, bucket_candidates }) => {
-                            self.query_trees = explode_query_tree(query_tree);
-                            self.candidates = candidates;
-                            self.filtered_candidates = filtered_candidates;
+                }
+                None => match self.parent.next(params)? {
+                    Some(CriterionResult {
+                        query_tree: Some(query_tree),
+                        candidates,
+                        filtered_candidates,
+                        bucket_candidates,
+                    }) => {
+                        self.query_trees = explode_query_tree(query_tree);
+                        self.candidates = candidates;
+                        self.filtered_candidates = filtered_candidates;
 
-                            self.bucket_candidates = match (self.bucket_candidates.take(), bucket_candidates) {
+                        self.bucket_candidates =
+                            match (self.bucket_candidates.take(), bucket_candidates) {
                                 (Some(self_bc), Some(parent_bc)) => Some(self_bc | parent_bc),
                                 (self_bc, parent_bc) => self_bc.or(parent_bc),
                             };
-                        },
-                        Some(CriterionResult { query_tree: None, candidates, filtered_candidates, bucket_candidates }) => {
-                            return Ok(Some(CriterionResult {
-                                query_tree: None,
-                                candidates,
-                                filtered_candidates,
-                                bucket_candidates,
-                            }));
-                        },
-                        None => return Ok(None),
                     }
+                    Some(CriterionResult {
+                        query_tree: None,
+                        candidates,
+                        filtered_candidates,
+                        bucket_candidates,
+                    }) => {
+                        return Ok(Some(CriterionResult {
+                            query_tree: None,
+                            candidates,
+                            filtered_candidates,
+                            bucket_candidates,
+                        }));
+                    }
+                    None => return Ok(None),
                 },
             }
         }
diff --git a/milli/src/search/distinct/facet_distinct.rs b/milli/src/search/distinct/facet_distinct.rs
index b9ffd9d90..290a7602f 100644
--- a/milli/src/search/distinct/facet_distinct.rs
+++ b/milli/src/search/distinct/facet_distinct.rs
@@ -3,11 +3,11 @@ use std::mem::size_of;
 use heed::types::ByteSlice;
 use roaring::RoaringBitmap;
 
+use super::{Distinct, DocIter};
 use crate::error::InternalError;
 use crate::heed_codec::facet::*;
 use crate::index::db_name;
 use crate::{DocumentId, FieldId, Index, Result};
-use super::{Distinct, DocIter};
 
 const FID_SIZE: usize = size_of::<FieldId>();
 const DOCID_SIZE: usize = size_of::<DocumentId>();
@@ -28,11 +28,7 @@ pub struct FacetDistinct<'a> {
 
 impl<'a> FacetDistinct<'a> {
     pub fn new(distinct: FieldId, index: &'a Index, txn: &'a heed::RoTxn<'a>) -> Self {
-        Self {
-            distinct,
-            index,
-            txn,
-        }
+        Self { distinct, index, txn }
     }
 }
 
@@ -47,16 +43,12 @@ pub struct FacetDistinctIter<'a> {
 
 impl<'a> FacetDistinctIter<'a> {
     fn facet_string_docids(&self, key: &str) -> heed::Result<Option<RoaringBitmap>> {
-        self.index
-            .facet_id_string_docids
-            .get(self.txn, &(self.distinct, key))
+        self.index.facet_id_string_docids.get(self.txn, &(self.distinct, key))
     }
 
     fn facet_number_docids(&self, key: f64) -> heed::Result<Option<RoaringBitmap>> {
         // get facet docids on level 0
-        self.index
-            .facet_id_f64_docids
-            .get(self.txn, &(self.distinct, 0, key, key))
+        self.index.facet_id_f64_docids.get(self.txn, &(self.distinct, 0, key, key))
     }
 
     fn distinct_string(&mut self, id: DocumentId) -> Result<()> {
@@ -64,9 +56,8 @@ impl<'a> FacetDistinctIter<'a> {
 
         for item in iter {
             let ((_, _, value), _) = item?;
-            let facet_docids = self
-                .facet_string_docids(value)?
-                .ok_or(InternalError::DatabaseMissingEntry {
+            let facet_docids =
+                self.facet_string_docids(value)?.ok_or(InternalError::DatabaseMissingEntry {
                     db_name: db_name::FACET_ID_STRING_DOCIDS,
                     key: None,
                 })?;
@@ -83,9 +74,8 @@ impl<'a> FacetDistinctIter<'a> {
 
         for item in iter {
             let ((_, _, value), _) = item?;
-            let facet_docids = self
-                .facet_number_docids(value)?
-                .ok_or(InternalError::DatabaseMissingEntry {
+            let facet_docids =
+                self.facet_number_docids(value)?.ok_or(InternalError::DatabaseMissingEntry {
                     db_name: db_name::FACET_ID_F64_DOCIDS,
                     key: None,
                 })?;
diff --git a/milli/src/search/distinct/mod.rs b/milli/src/search/distinct/mod.rs
index 99bc74be0..ae3fdb91e 100644
--- a/milli/src/search/distinct/mod.rs
+++ b/milli/src/search/distinct/mod.rs
@@ -1,11 +1,11 @@
 mod facet_distinct;
 mod noop_distinct;
 
+pub use facet_distinct::FacetDistinct;
+pub use noop_distinct::NoopDistinct;
 use roaring::RoaringBitmap;
 
 use crate::{DocumentId, Result};
-pub use facet_distinct::FacetDistinct;
-pub use noop_distinct::NoopDistinct;
 
 /// A trait implemented by document interators that are returned by calls to `Distinct::distinct`.
 /// It provides a way to get back the ownership to the excluded set.
@@ -29,13 +29,15 @@ mod test {
     use std::collections::HashSet;
 
     use once_cell::sync::Lazy;
-    use rand::{seq::SliceRandom, Rng};
+    use rand::seq::SliceRandom;
+    use rand::Rng;
     use roaring::RoaringBitmap;
     use serde_json::{json, Value};
 
-    use crate::index::{Index, tests::TempIndex};
+    use crate::index::tests::TempIndex;
+    use crate::index::Index;
     use crate::update::{IndexDocumentsMethod, UpdateBuilder, UpdateFormat};
-    use crate::{BEU32, FieldId, DocumentId};
+    use crate::{DocumentId, FieldId, BEU32};
 
     static JSON: Lazy<Value> = Lazy::new(generate_json);
 
@@ -89,9 +91,7 @@ mod test {
         addition.index_documents_method(IndexDocumentsMethod::ReplaceDocuments);
         addition.update_format(UpdateFormat::Json);
 
-        addition
-            .execute(JSON.to_string().as_bytes(), |_, _| ())
-            .unwrap();
+        addition.execute(JSON.to_string().as_bytes(), |_, _| ()).unwrap();
 
         let fields_map = index.fields_ids_map(&txn).unwrap();
         let fid = fields_map.id(&distinct).unwrap();
@@ -103,13 +103,12 @@ mod test {
         (index, fid, map)
     }
 
-
     /// Checks that all the candidates are distinct, and returns the candidates number.
     pub(crate) fn validate_distinct_candidates(
         candidates: impl Iterator<Item = crate::Result<DocumentId>>,
         distinct: FieldId,
         index: &Index,
-        ) -> usize {
+    ) -> usize {
         fn test(seen: &mut HashSet<String>, value: &Value) {
             match value {
                 Value::Null | Value::Object(_) | Value::Bool(_) => (),
@@ -117,7 +116,7 @@ mod test {
                     let s = value.to_string();
                     assert!(seen.insert(s));
                 }
-                Value::Array(values) => {values.into_iter().for_each(|value| test(seen, value))}
+                Value::Array(values) => values.into_iter().for_each(|value| test(seen, value)),
             }
         }
 
diff --git a/milli/src/search/distinct/noop_distinct.rs b/milli/src/search/distinct/noop_distinct.rs
index 812701794..96a1f7d5d 100644
--- a/milli/src/search/distinct/noop_distinct.rs
+++ b/milli/src/search/distinct/noop_distinct.rs
@@ -1,7 +1,8 @@
-use roaring::{RoaringBitmap, bitmap::IntoIter};
+use roaring::bitmap::IntoIter;
+use roaring::RoaringBitmap;
 
+use super::{Distinct, DocIter};
 use crate::{DocumentId, Result};
-use super::{DocIter, Distinct};
 
 /// A distinct implementer that does not perform any distinct,
 /// and simply returns an iterator to the candidates.
@@ -30,10 +31,7 @@ impl Distinct for NoopDistinct {
     type Iter = NoopDistinctIter;
 
     fn distinct(&mut self, candidates: RoaringBitmap, excluded: RoaringBitmap) -> Self::Iter {
-        NoopDistinctIter {
-            candidates: candidates.into_iter(),
-            excluded,
-        }
+        NoopDistinctIter { candidates: candidates.into_iter(), excluded }
     }
 }
 
diff --git a/milli/src/search/facet/facet_distribution.rs b/milli/src/search/facet/facet_distribution.rs
index 265a8ffeb..0a2036494 100644
--- a/milli/src/search/facet/facet_distribution.rs
+++ b/milli/src/search/facet/facet_distribution.rs
@@ -1,16 +1,16 @@
-use std::collections::{HashSet, BTreeMap};
+use std::collections::{BTreeMap, HashSet};
 use std::ops::Bound::Unbounded;
 use std::{cmp, fmt};
 
-use heed::{Database, BytesDecode};
 use heed::types::{ByteSlice, Unit};
+use heed::{BytesDecode, Database};
 use roaring::RoaringBitmap;
 
 use crate::error::FieldIdMapMissingEntry;
 use crate::facet::FacetType;
 use crate::heed_codec::facet::FacetValueStringCodec;
 use crate::search::facet::{FacetIter, FacetRange};
-use crate::{Index, FieldId, DocumentId, Result};
+use crate::{DocumentId, FieldId, Index, Result};
 
 /// The default number of values by facets that will
 /// be fetched from the key-value store.
@@ -43,7 +43,7 @@ impl<'a> FacetDistribution<'a> {
         }
     }
 
-    pub fn facets<I: IntoIterator<Item=A>, A: AsRef<str>>(&mut self, names: I) -> &mut Self {
+    pub fn facets<I: IntoIterator<Item = A>, A: AsRef<str>>(&mut self, names: I) -> &mut Self {
         self.facets = Some(names.into_iter().map(|s| s.as_ref().to_string()).collect());
         self
     }
@@ -66,8 +66,7 @@ impl<'a> FacetDistribution<'a> {
         facet_type: FacetType,
         candidates: &RoaringBitmap,
         distribution: &mut BTreeMap<String, u64>,
-    ) -> heed::Result<()>
-    {
+    ) -> heed::Result<()> {
         fn fetch_facet_values<'t, KC, K: 't>(
             rtxn: &'t heed::RoTxn,
             db: Database<KC, Unit>,
@@ -102,7 +101,7 @@ impl<'a> FacetDistribution<'a> {
             FacetType::Number => {
                 let db = self.index.field_id_docid_facet_f64s;
                 fetch_facet_values(self.rtxn, db, field_id, candidates, distribution)
-            },
+            }
             FacetType::String => {
                 let db = self.index.field_id_docid_facet_strings;
                 fetch_facet_values(self.rtxn, db, field_id, candidates, distribution)
@@ -117,11 +116,9 @@ impl<'a> FacetDistribution<'a> {
         field_id: FieldId,
         candidates: &RoaringBitmap,
         distribution: &mut BTreeMap<String, u64>,
-    ) -> heed::Result<()>
-    {
-        let iter = FacetIter::new_non_reducing(
-            self.rtxn, self.index, field_id, candidates.clone(),
-        )?;
+    ) -> heed::Result<()> {
+        let iter =
+            FacetIter::new_non_reducing(self.rtxn, self.index, field_id, candidates.clone())?;
 
         for result in iter {
             let (value, mut docids) = result?;
@@ -142,8 +139,7 @@ impl<'a> FacetDistribution<'a> {
     fn facet_values_from_raw_facet_database(
         &self,
         field_id: FieldId,
-    ) -> heed::Result<BTreeMap<String, u64>>
-    {
+    ) -> heed::Result<BTreeMap<String, u64>> {
         let mut distribution = BTreeMap::new();
 
         let db = self.index.facet_id_f64_docids;
@@ -157,7 +153,8 @@ impl<'a> FacetDistribution<'a> {
             }
         }
 
-        let iter = self.index
+        let iter = self
+            .index
             .facet_id_string_docids
             .remap_key_type::<ByteSlice>()
             .prefix_iter(self.rtxn, &[field_id])?
@@ -182,11 +179,30 @@ impl<'a> FacetDistribution<'a> {
             // to those candidates. We also enter here for facet strings for performance reasons.
             let mut distribution = BTreeMap::new();
             if candidates.len() <= CANDIDATES_THRESHOLD {
-                self.facet_distribution_from_documents(field_id, Number, candidates, &mut distribution)?;
-                self.facet_distribution_from_documents(field_id, String, candidates, &mut distribution)?;
+                self.facet_distribution_from_documents(
+                    field_id,
+                    Number,
+                    candidates,
+                    &mut distribution,
+                )?;
+                self.facet_distribution_from_documents(
+                    field_id,
+                    String,
+                    candidates,
+                    &mut distribution,
+                )?;
             } else {
-                self.facet_numbers_distribution_from_facet_levels(field_id, candidates, &mut distribution)?;
-                self.facet_distribution_from_documents(field_id, String, candidates, &mut distribution)?;
+                self.facet_numbers_distribution_from_facet_levels(
+                    field_id,
+                    candidates,
+                    &mut distribution,
+                )?;
+                self.facet_distribution_from_documents(
+                    field_id,
+                    String,
+                    candidates,
+                    &mut distribution,
+                )?;
             }
 
             Ok(distribution)
@@ -201,10 +217,11 @@ impl<'a> FacetDistribution<'a> {
 
         let mut distribution = BTreeMap::new();
         for name in filterable_fields {
-            let fid = fields_ids_map.id(&name).ok_or_else(|| FieldIdMapMissingEntry::FieldName {
-                field_name: name.clone(),
-                process: "FacetDistribution::execute",
-            })?;
+            let fid =
+                fields_ids_map.id(&name).ok_or_else(|| FieldIdMapMissingEntry::FieldName {
+                    field_name: name.clone(),
+                    process: "FacetDistribution::execute",
+                })?;
             let values = self.facet_values(fid)?;
             distribution.insert(name, values);
         }
@@ -215,13 +232,7 @@ impl<'a> FacetDistribution<'a> {
 
 impl fmt::Debug for FacetDistribution<'_> {
     fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
-        let FacetDistribution {
-            facets,
-            candidates,
-            max_values_by_facet,
-            rtxn: _,
-            index: _,
-        } = self;
+        let FacetDistribution { facets, candidates, max_values_by_facet, rtxn: _, index: _ } = self;
 
         f.debug_struct("FacetDistribution")
             .field("facets", facets)
diff --git a/milli/src/search/facet/filter_condition.rs b/milli/src/search/facet/filter_condition.rs
index 424118f77..31fc6018c 100644
--- a/milli/src/search/facet/filter_condition.rs
+++ b/milli/src/search/facet/filter_condition.rs
@@ -1,6 +1,6 @@
 use std::collections::HashSet;
 use std::fmt::Debug;
-use std::ops::Bound::{self, Included, Excluded};
+use std::ops::Bound::{self, Excluded, Included};
 use std::result::Result as StdResult;
 use std::str::FromStr;
 
@@ -12,16 +12,13 @@ use pest::iterators::{Pair, Pairs};
 use pest::Parser;
 use roaring::RoaringBitmap;
 
-use crate::error::UserError;
-use crate::heed_codec::facet::{FacetValueStringCodec, FacetLevelValueF64Codec};
-use crate::{Index, FieldId, FieldsIdsMap, CboRoaringBitmapCodec, Result};
-
-use super::FacetRange;
-use super::parser::Rule;
-use super::parser::{PREC_CLIMBER, FilterParser};
-
 use self::FilterCondition::*;
 use self::Operator::*;
+use super::parser::{FilterParser, Rule, PREC_CLIMBER};
+use super::FacetRange;
+use crate::error::UserError;
+use crate::heed_codec::facet::{FacetLevelValueF64Codec, FacetValueStringCodec};
+use crate::{CboRoaringBitmapCodec, FieldId, FieldsIdsMap, Index, Result};
 
 #[derive(Debug, Clone, PartialEq)]
 pub enum Operator {
@@ -39,13 +36,13 @@ impl Operator {
     /// an OR operation for the between case (i.e. `TO`).
     fn negate(self) -> (Self, Option<Self>) {
         match self {
-            GreaterThan(n)        => (LowerThanOrEqual(n), None),
+            GreaterThan(n) => (LowerThanOrEqual(n), None),
             GreaterThanOrEqual(n) => (LowerThan(n), None),
-            Equal(n, s)           => (NotEqual(n, s), None),
-            NotEqual(n, s)        => (Equal(n, s), None),
-            LowerThan(n)          => (GreaterThanOrEqual(n), None),
-            LowerThanOrEqual(n)   => (GreaterThan(n), None),
-            Between(n, m)         => (LowerThan(n), Some(GreaterThan(m))),
+            Equal(n, s) => (NotEqual(n, s), None),
+            NotEqual(n, s) => (Equal(n, s), None),
+            LowerThan(n) => (GreaterThanOrEqual(n), None),
+            LowerThanOrEqual(n) => (GreaterThan(n), None),
+            Between(n, m) => (LowerThan(n), Some(GreaterThan(m))),
         }
     }
 }
@@ -63,10 +60,11 @@ impl FilterCondition {
         index: &Index,
         array: I,
     ) -> Result<Option<FilterCondition>>
-    where I: IntoIterator<Item=Either<J, B>>,
-          J: IntoIterator<Item=A>,
-          A: AsRef<str>,
-          B: AsRef<str>,
+    where
+        I: IntoIterator<Item = Either<J, B>>,
+        J: IntoIterator<Item = A>,
+        A: AsRef<str>,
+        B: AsRef<str>,
     {
         let mut ands = None;
 
@@ -88,7 +86,7 @@ impl FilterCondition {
                             None => Some(rule),
                         };
                     }
-                },
+                }
                 Either::Right(rule) => {
                     let condition = FilterCondition::from_str(rtxn, index, rule.as_ref())?;
                     ands = match ands.take() {
@@ -106,11 +104,11 @@ impl FilterCondition {
         rtxn: &heed::RoTxn,
         index: &Index,
         expression: &str,
-    ) -> Result<FilterCondition>
-    {
+    ) -> Result<FilterCondition> {
         let fields_ids_map = index.fields_ids_map(rtxn)?;
         let filterable_fields = index.filterable_fields_ids(rtxn)?;
-        let lexed = FilterParser::parse(Rule::prgm, expression).map_err(UserError::InvalidFilter)?;
+        let lexed =
+            FilterParser::parse(Rule::prgm, expression).map_err(UserError::InvalidFilter)?;
         FilterCondition::from_pairs(&fields_ids_map, &filterable_fields, lexed)
     }
 
@@ -118,8 +116,7 @@ impl FilterCondition {
         fim: &FieldsIdsMap,
         ff: &HashSet<FieldId>,
         expression: Pairs<Rule>,
-    ) -> Result<Self>
-    {
+    ) -> Result<Self> {
         PREC_CLIMBER.climb(
             expression,
             |pair: Pair<Rule>| match pair.as_rule() {
@@ -135,12 +132,10 @@ impl FilterCondition {
                 Rule::term => Self::from_pairs(fim, ff, pair.into_inner()),
                 _ => unreachable!(),
             },
-            |lhs: Result<Self>, op: Pair<Rule>, rhs: Result<Self>| {
-                match op.as_rule() {
-                    Rule::or => Ok(Or(Box::new(lhs?), Box::new(rhs?))),
-                    Rule::and => Ok(And(Box::new(lhs?), Box::new(rhs?))),
-                    _ => unreachable!(),
-                }
+            |lhs: Result<Self>, op: Pair<Rule>, rhs: Result<Self>| match op.as_rule() {
+                Rule::or => Ok(Or(Box::new(lhs?), Box::new(rhs?))),
+                Rule::and => Ok(And(Box::new(lhs?), Box::new(rhs?))),
+                _ => unreachable!(),
             },
         )
     }
@@ -160,8 +155,7 @@ impl FilterCondition {
         fields_ids_map: &FieldsIdsMap,
         filterable_fields: &HashSet<FieldId>,
         item: Pair<Rule>,
-    ) -> Result<FilterCondition>
-    {
+    ) -> Result<FilterCondition> {
         let mut items = item.into_inner();
         let fid = field_id(fields_ids_map, filterable_fields, &mut items)
             .map_err(UserError::InvalidFilterAttribute)?;
@@ -179,8 +173,7 @@ impl FilterCondition {
         fields_ids_map: &FieldsIdsMap,
         filterable_fields: &HashSet<FieldId>,
         item: Pair<Rule>,
-    ) -> Result<FilterCondition>
-    {
+    ) -> Result<FilterCondition> {
         let mut items = item.into_inner();
         let fid = field_id(fields_ids_map, filterable_fields, &mut items)
             .map_err(UserError::InvalidFilterAttribute)?;
@@ -196,8 +189,7 @@ impl FilterCondition {
         fields_ids_map: &FieldsIdsMap,
         filterable_fields: &HashSet<FieldId>,
         item: Pair<Rule>,
-    ) -> Result<FilterCondition>
-    {
+    ) -> Result<FilterCondition> {
         let mut items = item.into_inner();
         let fid = field_id(fields_ids_map, filterable_fields, &mut items)
             .map_err(UserError::InvalidFilterAttribute)?;
@@ -213,8 +205,7 @@ impl FilterCondition {
         fields_ids_map: &FieldsIdsMap,
         filterable_fields: &HashSet<FieldId>,
         item: Pair<Rule>,
-    ) -> Result<FilterCondition>
-    {
+    ) -> Result<FilterCondition> {
         let mut items = item.into_inner();
         let fid = field_id(fields_ids_map, filterable_fields, &mut items)
             .map_err(UserError::InvalidFilterAttribute)?;
@@ -230,8 +221,7 @@ impl FilterCondition {
         fields_ids_map: &FieldsIdsMap,
         filterable_fields: &HashSet<FieldId>,
         item: Pair<Rule>,
-    ) -> Result<FilterCondition>
-    {
+    ) -> Result<FilterCondition> {
         let mut items = item.into_inner();
         let fid = field_id(fields_ids_map, filterable_fields, &mut items)
             .map_err(UserError::InvalidFilterAttribute)?;
@@ -247,8 +237,7 @@ impl FilterCondition {
         fields_ids_map: &FieldsIdsMap,
         filterable_fields: &HashSet<FieldId>,
         item: Pair<Rule>,
-    ) -> Result<FilterCondition>
-    {
+    ) -> Result<FilterCondition> {
         let mut items = item.into_inner();
         let fid = field_id(fields_ids_map, filterable_fields, &mut items)
             .map_err(UserError::InvalidFilterAttribute)?;
@@ -272,13 +261,14 @@ impl FilterCondition {
         left: Bound<f64>,
         right: Bound<f64>,
         output: &mut RoaringBitmap,
-    ) -> Result<()>
-    {
+    ) -> Result<()> {
         match (left, right) {
             // If the request is an exact value we must go directly to the deepest level.
             (Included(l), Included(r)) if l == r && level > 0 => {
-                return Self::explore_facet_number_levels(rtxn, db, field_id, 0, left, right, output);
-            },
+                return Self::explore_facet_number_levels(
+                    rtxn, db, field_id, 0, left, right, output,
+                );
+            }
             // lower TO upper when lower > upper must return no result
             (Included(l), Included(r)) if l > r => return Ok(()),
             (Included(l), Excluded(r)) if l >= r => return Ok(()),
@@ -301,7 +291,9 @@ impl FilterCondition {
             debug!("{:?} to {:?} (level {}) found {} documents", l, r, level, docids.len());
             output.union_with(&docids);
             // We save the leftest and rightest bounds we actually found at this level.
-            if i == 0 { left_found = Some(l); }
+            if i == 0 {
+                left_found = Some(l);
+            }
             right_found = Some(r);
         }
 
@@ -318,20 +310,50 @@ impl FilterCondition {
                 // If the bound is satisfied we avoid calling this function again.
                 if !matches!(left, Included(l) if l == left_found) {
                     let sub_right = Excluded(left_found);
-                    debug!("calling left with {:?} to {:?} (level {})",  left, sub_right, deeper_level);
-                    Self::explore_facet_number_levels(rtxn, db, field_id, deeper_level, left, sub_right, output)?;
+                    debug!(
+                        "calling left with {:?} to {:?} (level {})",
+                        left, sub_right, deeper_level
+                    );
+                    Self::explore_facet_number_levels(
+                        rtxn,
+                        db,
+                        field_id,
+                        deeper_level,
+                        left,
+                        sub_right,
+                        output,
+                    )?;
                 }
                 if !matches!(right, Included(r) if r == right_found) {
                     let sub_left = Excluded(right_found);
-                    debug!("calling right with {:?} to {:?} (level {})", sub_left, right, deeper_level);
-                    Self::explore_facet_number_levels(rtxn, db, field_id, deeper_level, sub_left, right, output)?;
+                    debug!(
+                        "calling right with {:?} to {:?} (level {})",
+                        sub_left, right, deeper_level
+                    );
+                    Self::explore_facet_number_levels(
+                        rtxn,
+                        db,
+                        field_id,
+                        deeper_level,
+                        sub_left,
+                        right,
+                        output,
+                    )?;
                 }
-            },
+            }
             None => {
                 // If we found nothing at this level it means that we must find
                 // the same bounds but at a deeper, more precise level.
-                Self::explore_facet_number_levels(rtxn, db, field_id, deeper_level, left, right, output)?;
-            },
+                Self::explore_facet_number_levels(
+                    rtxn,
+                    db,
+                    field_id,
+                    deeper_level,
+                    left,
+                    right,
+                    output,
+                )?;
+            }
         }
 
         Ok(())
@@ -344,27 +366,34 @@ impl FilterCondition {
         strings_db: heed::Database<FacetValueStringCodec, CboRoaringBitmapCodec>,
         field_id: FieldId,
         operator: &Operator,
-    ) -> Result<RoaringBitmap>
-    {
+    ) -> Result<RoaringBitmap> {
         // Make sure we always bound the ranges with the field id and the level,
         // as the facets values are all in the same database and prefixed by the
         // field id and the level.
         let (left, right) = match operator {
-            GreaterThan(val)        => (Excluded(*val), Included(f64::MAX)),
+            GreaterThan(val) => (Excluded(*val), Included(f64::MAX)),
             GreaterThanOrEqual(val) => (Included(*val), Included(f64::MAX)),
-            Equal(number, string)   => {
+            Equal(number, string) => {
                 let string_docids = strings_db.get(rtxn, &(field_id, &string))?.unwrap_or_default();
                 let number_docids = match number {
                     Some(n) => {
                         let n = Included(*n);
                         let mut output = RoaringBitmap::new();
-                        Self::explore_facet_number_levels(rtxn, numbers_db, field_id, 0, n, n, &mut output)?;
+                        Self::explore_facet_number_levels(
+                            rtxn,
+                            numbers_db,
+                            field_id,
+                            0,
+                            n,
+                            n,
+                            &mut output,
+                        )?;
                         output
-                    },
+                    }
                     None => RoaringBitmap::new(),
                 };
                 return Ok(string_docids | number_docids);
-            },
+            }
             NotEqual(number, string) => {
                 let all_numbers_ids = if number.is_some() {
                     index.number_faceted_documents_ids(rtxn, field_id)?
@@ -373,12 +402,14 @@ impl FilterCondition {
                 };
                 let all_strings_ids = index.string_faceted_documents_ids(rtxn, field_id)?;
                 let operator = Equal(*number, string.clone());
-                let docids = Self::evaluate_operator(rtxn, index, numbers_db, strings_db, field_id, &operator)?;
+                let docids = Self::evaluate_operator(
+                    rtxn, index, numbers_db, strings_db, field_id, &operator,
+                )?;
                 return Ok((all_numbers_ids | all_strings_ids) - docids);
-            },
-            LowerThan(val)        => (Included(f64::MIN), Excluded(*val)),
+            }
+            LowerThan(val) => (Included(f64::MIN), Excluded(*val)),
             LowerThanOrEqual(val) => (Included(f64::MIN), Included(*val)),
-            Between(left, right)  => (Included(*left),    Included(*right)),
+            Between(left, right) => (Included(*left), Included(*right)),
         };
 
         // Ask for the biggest value that can exist for this specific field, if it exists
@@ -391,36 +422,39 @@ impl FilterCondition {
         match biggest_level {
             Some(level) => {
                 let mut output = RoaringBitmap::new();
-                Self::explore_facet_number_levels(rtxn, numbers_db, field_id, level, left, right, &mut output)?;
+                Self::explore_facet_number_levels(
+                    rtxn,
+                    numbers_db,
+                    field_id,
+                    level,
+                    left,
+                    right,
+                    &mut output,
+                )?;
                 Ok(output)
-            },
+            }
             None => Ok(RoaringBitmap::new()),
         }
     }
 
-    pub fn evaluate(
-        &self,
-        rtxn: &heed::RoTxn,
-        index: &Index,
-    ) -> Result<RoaringBitmap>
-    {
+    pub fn evaluate(&self, rtxn: &heed::RoTxn, index: &Index) -> Result<RoaringBitmap> {
         let numbers_db = index.facet_id_f64_docids;
         let strings_db = index.facet_id_string_docids;
 
         match self {
             Operator(fid, op) => {
                 Self::evaluate_operator(rtxn, index, numbers_db, strings_db, *fid, op)
-            },
+            }
             Or(lhs, rhs) => {
                 let lhs = lhs.evaluate(rtxn, index)?;
                 let rhs = rhs.evaluate(rtxn, index)?;
                 Ok(lhs | rhs)
-            },
+            }
             And(lhs, rhs) => {
                 let lhs = lhs.evaluate(rtxn, index)?;
                 let rhs = rhs.evaluate(rtxn, index)?;
                 Ok(lhs & rhs)
-            },
+            }
         }
     }
 }
@@ -434,23 +468,24 @@ fn field_id(
     fields_ids_map: &FieldsIdsMap,
     filterable_fields: &HashSet<FieldId>,
     items: &mut Pairs<Rule>,
-) -> StdResult<FieldId, PestError<Rule>>
-{
+) -> StdResult<FieldId, PestError<Rule>> {
     // lexing ensures that we at least have a key
     let key = items.next().unwrap();
 
     let field_id = match fields_ids_map.id(key.as_str()) {
         Some(field_id) => field_id,
-        None => return Err(PestError::new_from_span(
-            ErrorVariant::CustomError {
-                message: format!(
-                    "attribute `{}` not found, available attributes are: {}",
-                    key.as_str(),
-                    fields_ids_map.iter().map(|(_, n)| n).collect::<Vec<_>>().join(", "),
-                ),
-            },
-            key.as_span(),
-        )),
+        None => {
+            return Err(PestError::new_from_span(
+                ErrorVariant::CustomError {
+                    message: format!(
+                        "attribute `{}` not found, available attributes are: {}",
+                        key.as_str(),
+                        fields_ids_map.iter().map(|(_, n)| n).collect::<Vec<_>>().join(", "),
+                    ),
+                },
+                key.as_span(),
+            ))
+        }
     };
 
     if !filterable_fields.contains(&field_id) {
@@ -459,9 +494,11 @@ fn field_id(
                 message: format!(
                     "attribute `{}` is not filterable, available filterable attributes are: {}",
                     key.as_str(),
-                    filterable_fields.iter().flat_map(|id| {
-                        fields_ids_map.name(*id)
-                    }).collect::<Vec<_>>().join(", "),
+                    filterable_fields
+                        .iter()
+                        .flat_map(|id| { fields_ids_map.name(*id) })
+                        .collect::<Vec<_>>()
+                        .join(", "),
                 ),
             },
             key.as_span(),
@@ -476,8 +513,9 @@ fn field_id(
 ///
 /// Returns the parsing error associated with the span if the conversion fails.
 fn pest_parse<T>(pair: Pair<Rule>) -> (StdResult<T, pest::error::Error<Rule>>, String)
-where T: FromStr,
-      T::Err: ToString,
+where
+    T: FromStr,
+    T::Err: ToString,
 {
     let result = match pair.as_str().parse::<T>() {
         Ok(value) => Ok(value),
@@ -492,11 +530,12 @@ where T: FromStr,
 
 #[cfg(test)]
 mod tests {
-    use super::*;
-    use crate::update::Settings;
+    use big_s::S;
     use heed::EnvOpenOptions;
     use maplit::hashset;
-    use big_s::S;
+
+    use super::*;
+    use crate::update::Settings;
 
     #[test]
     fn string() {
@@ -508,7 +547,7 @@ mod tests {
         // Set the filterable fields to be the channel.
         let mut wtxn = index.write_txn().unwrap();
         let mut builder = Settings::new(&mut wtxn, &index, 0);
-        builder.set_filterable_fields(hashset!{ S("channel") });
+        builder.set_filterable_fields(hashset! { S("channel") });
         builder.execute(|_, _| ()).unwrap();
         wtxn.commit().unwrap();
 
@@ -537,7 +576,7 @@ mod tests {
         // Set the filterable fields to be the channel.
         let mut wtxn = index.write_txn().unwrap();
         let mut builder = Settings::new(&mut wtxn, &index, 0);
-        builder.set_filterable_fields(hashset!{ "timestamp".into() });
+        builder.set_filterable_fields(hashset! { "timestamp".into() });
         builder.execute(|_, _| ()).unwrap();
         wtxn.commit().unwrap();
 
@@ -548,10 +587,8 @@ mod tests {
         assert_eq!(condition, expected);
 
         let condition = FilterCondition::from_str(&rtxn, &index, "NOT timestamp 22 TO 44").unwrap();
-        let expected = Or(
-            Box::new(Operator(0, LowerThan(22.0))),
-            Box::new(Operator(0, GreaterThan(44.0))),
-        );
+        let expected =
+            Or(Box::new(Operator(0, LowerThan(22.0))), Box::new(Operator(0, GreaterThan(44.0))));
         assert_eq!(condition, expected);
     }
 
@@ -566,29 +603,33 @@ mod tests {
         let mut wtxn = index.write_txn().unwrap();
         let mut builder = Settings::new(&mut wtxn, &index, 0);
         builder.set_searchable_fields(vec![S("channel"), S("timestamp")]); // to keep the fields order
-        builder.set_filterable_fields(hashset!{ S("channel"), S("timestamp") });
+        builder.set_filterable_fields(hashset! { S("channel"), S("timestamp") });
         builder.execute(|_, _| ()).unwrap();
         wtxn.commit().unwrap();
 
         // Test that the facet condition is correctly generated.
         let rtxn = index.read_txn().unwrap();
         let condition = FilterCondition::from_str(
-            &rtxn, &index,
+            &rtxn,
+            &index,
             "channel = gotaga OR (timestamp 22 TO 44 AND channel != ponce)",
-        ).unwrap();
+        )
+        .unwrap();
         let expected = Or(
             Box::new(Operator(0, Operator::Equal(None, S("gotaga")))),
             Box::new(And(
                 Box::new(Operator(1, Between(22.0, 44.0))),
                 Box::new(Operator(0, Operator::NotEqual(None, S("ponce")))),
-            ))
+            )),
         );
         assert_eq!(condition, expected);
 
         let condition = FilterCondition::from_str(
-            &rtxn, &index,
+            &rtxn,
+            &index,
             "channel = gotaga OR NOT (timestamp 22 TO 44 AND channel != ponce)",
-        ).unwrap();
+        )
+        .unwrap();
         let expected = Or(
             Box::new(Operator(0, Operator::Equal(None, S("gotaga")))),
             Box::new(Or(
@@ -613,20 +654,28 @@ mod tests {
         let mut wtxn = index.write_txn().unwrap();
         let mut builder = Settings::new(&mut wtxn, &index, 0);
         builder.set_searchable_fields(vec![S("channel"), S("timestamp")]); // to keep the fields order
-        builder.set_filterable_fields(hashset!{ S("channel"), S("timestamp") });
+        builder.set_filterable_fields(hashset! { S("channel"), S("timestamp") });
         builder.execute(|_, _| ()).unwrap();
         wtxn.commit().unwrap();
 
         // Test that the facet condition is correctly generated.
         let rtxn = index.read_txn().unwrap();
         let condition = FilterCondition::from_array(
-            &rtxn, &index,
-            vec![Either::Right("channel = gotaga"), Either::Left(vec!["timestamp = 44", "channel != ponce"])],
-        ).unwrap().unwrap();
+            &rtxn,
+            &index,
+            vec![
+                Either::Right("channel = gotaga"),
+                Either::Left(vec!["timestamp = 44", "channel != ponce"]),
+            ],
+        )
+        .unwrap()
+        .unwrap();
         let expected = FilterCondition::from_str(
-            &rtxn, &index,
+            &rtxn,
+            &index,
             "channel = gotaga AND (timestamp = 44 OR channel != ponce)",
-        ).unwrap();
+        )
+        .unwrap();
         assert_eq!(condition, expected);
     }
 }
diff --git a/milli/src/search/facet/mod.rs b/milli/src/search/facet/mod.rs
index a1a03dba3..240d99ccc 100644
--- a/milli/src/search/facet/mod.rs
+++ b/milli/src/search/facet/mod.rs
@@ -1,20 +1,19 @@
-use std::ops::Bound::{self, Included, Excluded, Unbounded};
+use std::ops::Bound::{self, Excluded, Included, Unbounded};
 
 use either::Either::{self, Left, Right};
-use heed::types::{DecodeIgnore, ByteSlice};
-use heed::{Database, RoRange, RoRevRange, LazyDecode};
+use heed::types::{ByteSlice, DecodeIgnore};
+use heed::{Database, LazyDecode, RoRange, RoRevRange};
 use roaring::RoaringBitmap;
 
-use crate::heed_codec::CboRoaringBitmapCodec;
-use crate::heed_codec::facet::FacetLevelValueF64Codec;
-use crate::{Index, FieldId};
-
 pub use self::facet_distribution::FacetDistribution;
 pub use self::filter_condition::{FilterCondition, Operator};
 pub(crate) use self::parser::Rule as ParserRule;
+use crate::heed_codec::facet::FacetLevelValueF64Codec;
+use crate::heed_codec::CboRoaringBitmapCodec;
+use crate::{FieldId, Index};
 
-mod filter_condition;
 mod facet_distribution;
+mod filter_condition;
 mod parser;
 
 pub struct FacetRange<'t> {
@@ -30,8 +29,7 @@ impl<'t> FacetRange<'t> {
         level: u8,
         left: Bound<f64>,
         right: Bound<f64>,
-    ) -> heed::Result<FacetRange<'t>>
-    {
+    ) -> heed::Result<FacetRange<'t>> {
         let left_bound = match left {
             Included(left) => Included((field_id, level, left, f64::MIN)),
             Excluded(left) => Excluded((field_id, level, left, f64::MIN)),
@@ -62,7 +60,7 @@ impl<'t> Iterator for FacetRange<'t> {
                 } else {
                     None
                 }
-            },
+            }
             Some(Err(e)) => Some(Err(e)),
             None => None,
         }
@@ -82,8 +80,7 @@ impl<'t> FacetRevRange<'t> {
         level: u8,
         left: Bound<f64>,
         right: Bound<f64>,
-    ) -> heed::Result<FacetRevRange<'t>>
-    {
+    ) -> heed::Result<FacetRevRange<'t>> {
         let left_bound = match left {
             Included(left) => Included((field_id, level, left, f64::MIN)),
             Excluded(left) => Excluded((field_id, level, left, f64::MIN)),
@@ -114,7 +111,7 @@ impl<'t> Iterator for FacetRevRange<'t> {
                         }
                     }
                     continue;
-                },
+                }
                 Some(Err(e)) => return Some(Err(e)),
                 None => return None,
             }
@@ -139,11 +136,11 @@ impl<'t> FacetIter<'t> {
         index: &'t Index,
         field_id: FieldId,
         documents_ids: RoaringBitmap,
-    ) -> heed::Result<FacetIter<'t>>
-    {
+    ) -> heed::Result<FacetIter<'t>> {
         let db = index.facet_id_f64_docids.remap_key_type::<FacetLevelValueF64Codec>();
         let highest_level = Self::highest_level(rtxn, db, field_id)?.unwrap_or(0);
-        let highest_iter = FacetRange::new(rtxn, db, field_id, highest_level, Unbounded, Unbounded)?;
+        let highest_iter =
+            FacetRange::new(rtxn, db, field_id, highest_level, Unbounded, Unbounded)?;
         let level_iters = vec![(documents_ids, Left(highest_iter))];
         Ok(FacetIter { rtxn, db, field_id, level_iters, must_reduce: true })
     }
@@ -156,11 +153,11 @@ impl<'t> FacetIter<'t> {
         index: &'t Index,
         field_id: FieldId,
         documents_ids: RoaringBitmap,
-    ) -> heed::Result<FacetIter<'t>>
-    {
+    ) -> heed::Result<FacetIter<'t>> {
         let db = index.facet_id_f64_docids.remap_key_type::<FacetLevelValueF64Codec>();
         let highest_level = Self::highest_level(rtxn, db, field_id)?.unwrap_or(0);
-        let highest_iter = FacetRevRange::new(rtxn, db, field_id, highest_level, Unbounded, Unbounded)?;
+        let highest_iter =
+            FacetRevRange::new(rtxn, db, field_id, highest_level, Unbounded, Unbounded)?;
         let level_iters = vec![(documents_ids, Right(highest_iter))];
         Ok(FacetIter { rtxn, db, field_id, level_iters, must_reduce: true })
     }
@@ -174,11 +171,11 @@ impl<'t> FacetIter<'t> {
         index: &'t Index,
         field_id: FieldId,
         documents_ids: RoaringBitmap,
-    ) -> heed::Result<FacetIter<'t>>
-    {
+    ) -> heed::Result<FacetIter<'t>> {
         let db = index.facet_id_f64_docids.remap_key_type::<FacetLevelValueF64Codec>();
         let highest_level = Self::highest_level(rtxn, db, field_id)?.unwrap_or(0);
-        let highest_iter = FacetRange::new(rtxn, db, field_id, highest_level, Unbounded, Unbounded)?;
+        let highest_iter =
+            FacetRange::new(rtxn, db, field_id, highest_level, Unbounded, Unbounded)?;
         let level_iters = vec![(documents_ids, Left(highest_iter))];
         Ok(FacetIter { rtxn, db, field_id, level_iters, must_reduce: false })
     }
@@ -187,12 +184,13 @@ impl<'t> FacetIter<'t> {
         rtxn: &'t heed::RoTxn,
         db: Database<FacetLevelValueF64Codec, X>,
         fid: FieldId,
-    ) -> heed::Result<Option<u8>>
-    {
-        let level = db.remap_types::<ByteSlice, DecodeIgnore>()
+    ) -> heed::Result<Option<u8>> {
+        let level = db
+            .remap_types::<ByteSlice, DecodeIgnore>()
             .prefix_iter(rtxn, &[fid][..])?
             .remap_key_type::<FacetLevelValueF64Codec>()
-            .last().transpose()?
+            .last()
+            .transpose()?
             .map(|((_, level, _, _), _)| level);
         Ok(level)
     }
@@ -215,7 +213,6 @@ impl<'t> Iterator for FacetIter<'t> {
 
                 match result {
                     Ok(((_fid, level, left, right), mut docids)) => {
-
                         docids.intersect_with(&documents_ids);
                         if !docids.is_empty() {
                             if self.must_reduce {
@@ -242,11 +239,11 @@ impl<'t> Iterator for FacetIter<'t> {
                                 Ok(iter) => {
                                     self.level_iters.push((docids, iter));
                                     continue 'outer;
-                                },
+                                }
                                 Err(e) => return Some(Err(e)),
                             }
                         }
-                    },
+                    }
                     Err(e) => return Some(Err(e)),
                 }
             }
diff --git a/milli/src/search/facet/parser.rs b/milli/src/search/facet/parser.rs
index 0e8bd23ac..1bff27cfb 100644
--- a/milli/src/search/facet/parser.rs
+++ b/milli/src/search/facet/parser.rs
@@ -1,5 +1,5 @@
 use once_cell::sync::Lazy;
-use pest::prec_climber::{Operator, Assoc, PrecClimber};
+use pest::prec_climber::{Assoc, Operator, PrecClimber};
 
 pub static PREC_CLIMBER: Lazy<PrecClimber<Rule>> = Lazy::new(|| {
     use Assoc::*;
diff --git a/milli/src/search/matching_words.rs b/milli/src/search/matching_words.rs
index c56db4e96..cd8e404b8 100644
--- a/milli/src/search/matching_words.rs
+++ b/milli/src/search/matching_words.rs
@@ -1,13 +1,11 @@
-use std::collections::HashSet;
 use std::cmp::{min, Reverse};
-use std::collections::BTreeMap;
+use std::collections::{BTreeMap, HashSet};
 use std::ops::{Index, IndexMut};
 
-use levenshtein_automata::{DFA, Distance};
-
-use crate::search::query_tree::{Operation, Query};
+use levenshtein_automata::{Distance, DFA};
 
 use super::build_dfa;
+use crate::search::query_tree::{Operation, Query};
 
 type IsPrefix = bool;
 
@@ -28,7 +26,9 @@ impl MatchingWords {
             .collect();
         // Sort word by len in DESC order prioritizing the longuest word,
         // in order to highlight the longuest part of the matched word.
-        dfas.sort_unstable_by_key(|(_dfa, query_word, _typo, _is_prefix)| Reverse(query_word.len()));
+        dfas.sort_unstable_by_key(|(_dfa, query_word, _typo, _is_prefix)| {
+            Reverse(query_word.len())
+        });
         Self { dfas }
     }
 
@@ -37,12 +37,13 @@ impl MatchingWords {
         self.dfas.iter().find_map(|(dfa, query_word, typo, is_prefix)| match dfa.eval(word) {
             Distance::Exact(t) if t <= *typo => {
                 if *is_prefix {
-                    let (_dist, len) = prefix_damerau_levenshtein(query_word.as_bytes(), word.as_bytes());
+                    let (_dist, len) =
+                        prefix_damerau_levenshtein(query_word.as_bytes(), word.as_bytes());
                     Some(len)
                 } else {
                     Some(word.len())
                 }
-            },
+            }
             _otherwise => None,
         })
     }
@@ -54,11 +55,11 @@ fn fetch_queries(tree: &Operation) -> HashSet<(&str, u8, IsPrefix)> {
         match tree {
             Operation::Or(_, ops) | Operation::And(ops) => {
                 ops.as_slice().iter().for_each(|op| resolve_ops(op, out));
-            },
+            }
             Operation::Query(Query { prefix, kind }) => {
                 let typo = if kind.is_exact() { 0 } else { kind.typo() };
                 out.insert((kind.word(), typo, *prefix));
-            },
+            }
             Operation::Phrase(words) => {
                 for word in words {
                     out.insert((word, 0, false));
@@ -80,10 +81,7 @@ struct N2Array<T> {
 
 impl<T: Clone> N2Array<T> {
     fn new(x: usize, y: usize, value: T) -> N2Array<T> {
-        N2Array {
-            y_size: y,
-            buf: vec![value; x * y],
-        }
+        N2Array { y_size: y, buf: vec![value; x * y] }
     }
 }
 
@@ -178,9 +176,8 @@ fn prefix_damerau_levenshtein(source: &[u8], target: &[u8]) -> (u32, usize) {
 #[cfg(test)]
 mod tests {
     use super::*;
-
-    use crate::MatchingWords;
     use crate::search::query_tree::{Operation, Query, QueryKind};
+    use crate::MatchingWords;
 
     #[test]
     fn matched_length() {
@@ -194,13 +191,23 @@ mod tests {
 
     #[test]
     fn matching_words() {
-        let query_tree = Operation::Or(false, vec![
-            Operation::And(vec![
-                Operation::Query(Query { prefix: true, kind: QueryKind::exact("split".to_string()) }),
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact("this".to_string()) }),
-                Operation::Query(Query { prefix: true, kind: QueryKind::tolerant(1, "world".to_string()) }),
-            ]),
-        ]);
+        let query_tree = Operation::Or(
+            false,
+            vec![Operation::And(vec![
+                Operation::Query(Query {
+                    prefix: true,
+                    kind: QueryKind::exact("split".to_string()),
+                }),
+                Operation::Query(Query {
+                    prefix: false,
+                    kind: QueryKind::exact("this".to_string()),
+                }),
+                Operation::Query(Query {
+                    prefix: true,
+                    kind: QueryKind::tolerant(1, "world".to_string()),
+                }),
+            ])],
+        );
 
         let matching_words = MatchingWords::from_query_tree(&query_tree);
 
diff --git a/milli/src/search/mod.rs b/milli/src/search/mod.rs
index 3c85796bc..f692df173 100644
--- a/milli/src/search/mod.rs
+++ b/milli/src/search/mod.rs
@@ -6,6 +6,7 @@ use std::result::Result as StdResult;
 use std::str::Utf8Error;
 use std::time::Instant;
 
+use distinct::{Distinct, DocIter, FacetDistinct, NoopDistinct};
 use fst::{IntoStreamer, Streamer};
 use levenshtein_automata::{LevenshteinAutomatonBuilder as LevBuilder, DFA};
 use log::debug;
@@ -13,16 +14,13 @@ use meilisearch_tokenizer::{Analyzer, AnalyzerConfig};
 use once_cell::sync::Lazy;
 use roaring::bitmap::RoaringBitmap;
 
+pub(crate) use self::facet::ParserRule;
+pub use self::facet::{FacetDistribution, FacetIter, FilterCondition, Operator};
+pub use self::matching_words::MatchingWords;
+use self::query_tree::QueryTreeBuilder;
 use crate::error::FieldIdMapMissingEntry;
 use crate::search::criteria::r#final::{Final, FinalResult};
-use crate::{Index, DocumentId, Result};
-
-pub use self::facet::{FilterCondition, FacetDistribution, FacetIter, Operator};
-pub use self::matching_words::MatchingWords;
-pub(crate) use self::facet::ParserRule;
-use self::query_tree::QueryTreeBuilder;
-
-use distinct::{Distinct, DocIter, FacetDistinct, NoopDistinct};
+use crate::{DocumentId, Index, Result};
 
 // Building these factories is not free.
 static LEVDIST0: Lazy<LevBuilder> = Lazy::new(|| LevBuilder::new(0, true));
@@ -32,8 +30,8 @@ static LEVDIST2: Lazy<LevBuilder> = Lazy::new(|| LevBuilder::new(2, true));
 mod criteria;
 mod distinct;
 mod facet;
-mod query_tree;
 mod matching_words;
+mod query_tree;
 
 pub struct Search<'a> {
     query: Option<String>,
@@ -117,7 +115,7 @@ impl<'a> Search<'a> {
                 let result = analyzer.analyze(query);
                 let tokens = result.tokens();
                 builder.build(tokens)?.map_or((None, None), |(qt, pq)| (Some(qt), Some(pq)))
-            },
+            }
             None => (None, None),
         };
 
@@ -144,10 +142,11 @@ impl<'a> Search<'a> {
             None => self.perform_sort(NoopDistinct, matching_words, criteria),
             Some(name) => {
                 let field_ids_map = self.index.fields_ids_map(self.rtxn)?;
-                let id = field_ids_map.id(name).ok_or_else(|| FieldIdMapMissingEntry::FieldName {
-                    field_name: name.to_string(),
-                    process: "distinct attribute",
-                })?;
+                let id =
+                    field_ids_map.id(name).ok_or_else(|| FieldIdMapMissingEntry::FieldName {
+                        field_name: name.to_string(),
+                        process: "distinct attribute",
+                    })?;
                 let distinct = FacetDistinct::new(id, self.index, self.rtxn);
                 self.perform_sort(distinct, matching_words, criteria)
             }
@@ -159,14 +158,15 @@ impl<'a> Search<'a> {
         mut distinct: D,
         matching_words: MatchingWords,
         mut criteria: Final,
-    ) -> Result<SearchResult>
-    {
+    ) -> Result<SearchResult> {
         let mut offset = self.offset;
         let mut initial_candidates = RoaringBitmap::new();
         let mut excluded_candidates = RoaringBitmap::new();
         let mut documents_ids = Vec::with_capacity(self.limit);
 
-        while let Some(FinalResult { candidates, bucket_candidates, .. }) = criteria.next(&excluded_candidates)? {
+        while let Some(FinalResult { candidates, bucket_candidates, .. }) =
+            criteria.next(&excluded_candidates)?
+        {
             debug!("Number of candidates found {}", candidates.len());
 
             let excluded = take(&mut excluded_candidates);
@@ -183,7 +183,9 @@ impl<'a> Search<'a> {
             for candidate in candidates.by_ref().take(self.limit - documents_ids.len()) {
                 documents_ids.push(candidate?);
             }
-            if documents_ids.len() == self.limit { break }
+            if documents_ids.len() == self.limit {
+                break;
+            }
             excluded_candidates = candidates.into_excluded();
         }
 
@@ -247,7 +249,7 @@ pub fn word_derivations<'c>(
             }
 
             Ok(entry.insert(derived_words))
-        },
+        }
     }
 }
 
diff --git a/milli/src/search/query_tree.rs b/milli/src/search/query_tree.rs
index c371b07d4..8fa24b9d3 100644
--- a/milli/src/search/query_tree.rs
+++ b/milli/src/search/query_tree.rs
@@ -1,4 +1,4 @@
-use std::{fmt, cmp, mem};
+use std::{cmp, fmt, mem};
 
 use fst::Set;
 use meilisearch_tokenizer::token::SeparatorKind;
@@ -28,18 +28,18 @@ impl fmt::Debug for Operation {
                 Operation::And(children) => {
                     writeln!(f, "{:1$}AND", "", depth * 2)?;
                     children.iter().try_for_each(|c| pprint_tree(f, c, depth + 1))
-                },
+                }
                 Operation::Phrase(children) => {
                     writeln!(f, "{:2$}PHRASE {:?}", "", children, depth * 2)
-                },
+                }
                 Operation::Or(true, children) => {
                     writeln!(f, "{:1$}OR(WORD)", "", depth * 2)?;
                     children.iter().try_for_each(|c| pprint_tree(f, c, depth + 1))
-                },
+                }
                 Operation::Or(false, children) => {
                     writeln!(f, "{:1$}OR", "", depth * 2)?;
                     children.iter().try_for_each(|c| pprint_tree(f, c, depth + 1))
-                },
+                }
                 Operation::Query(query) => writeln!(f, "{:2$}{:?}", "", query, depth * 2),
             }
         }
@@ -136,10 +136,12 @@ impl fmt::Debug for Query {
         match kind {
             QueryKind::Exact { word, .. } => {
                 f.debug_struct(&(prefix + "Exact")).field("word", &word).finish()
-            },
-            QueryKind::Tolerant { typo, word } => {
-                f.debug_struct(&(prefix + "Tolerant")).field("word", &word).field("max typo", &typo).finish()
-            },
+            }
+            QueryKind::Tolerant { typo, word } => f
+                .debug_struct(&(prefix + "Tolerant"))
+                .field("word", &word)
+                .field("max typo", &typo)
+                .finish(),
         }
     }
 }
@@ -223,7 +225,12 @@ impl<'a> QueryTreeBuilder<'a> {
         let stop_words = self.index.stop_words(self.rtxn)?;
         let primitive_query = create_primitive_query(query, stop_words, self.words_limit);
         if !primitive_query.is_empty() {
-            let qt = create_query_tree(self, self.optional_words, self.authorize_typos, &primitive_query)?;
+            let qt = create_query_tree(
+                self,
+                self.optional_words,
+                self.authorize_typos,
+                &primitive_query,
+            )?;
             Ok(Some((qt, primitive_query)))
         } else {
             Ok(None)
@@ -248,12 +255,7 @@ fn split_best_frequency(ctx: &impl Context, word: &str) -> heed::Result<Option<O
         }
     }
 
-    Ok(best.map(|(_, left, right)| Operation::Phrase(
-        vec![
-            left.to_string(),
-            right.to_string()
-        ]
-    )))
+    Ok(best.map(|(_, left, right)| Operation::Phrase(vec![left.to_string(), right.to_string()])))
 }
 
 /// Return the `QueryKind` of a word depending on `authorize_typos`
@@ -263,7 +265,7 @@ fn typos(word: String, authorize_typos: bool) -> QueryKind {
         match word.len() {
             0..=4 => QueryKind::exact(word),
             5..=8 => QueryKind::tolerant(1, word),
-            _     => QueryKind::tolerant(2, word),
+            _ => QueryKind::tolerant(2, word),
         }
     } else {
         QueryKind::exact(word)
@@ -276,12 +278,18 @@ fn synonyms(ctx: &impl Context, word: &[&str]) -> heed::Result<Option<Vec<Operat
     let synonyms = ctx.synonyms(word)?;
 
     Ok(synonyms.map(|synonyms| {
-        synonyms.into_iter().map(|synonym| {
-            let words = synonym.into_iter().map(|word| {
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact(word) })
-            }).collect();
-            Operation::and(words)
-        }).collect()
+        synonyms
+            .into_iter()
+            .map(|synonym| {
+                let words = synonym
+                    .into_iter()
+                    .map(|word| {
+                        Operation::Query(Query { prefix: false, kind: QueryKind::exact(word) })
+                    })
+                    .collect();
+                Operation::and(words)
+            })
+            .collect()
     }))
 }
 
@@ -291,15 +299,13 @@ fn create_query_tree(
     optional_words: bool,
     authorize_typos: bool,
     query: &[PrimitiveQueryPart],
-) -> Result<Operation>
-{
+) -> Result<Operation> {
     /// Matches on the `PrimitiveQueryPart` and create an operation from it.
     fn resolve_primitive_part(
         ctx: &impl Context,
         authorize_typos: bool,
         part: PrimitiveQueryPart,
-    ) -> Result<Operation>
-    {
+    ) -> Result<Operation> {
         match part {
             // 1. try to split word in 2
             // 2. try to fetch synonyms
@@ -310,13 +316,12 @@ fn create_query_tree(
                 if let Some(child) = split_best_frequency(ctx, &word)? {
                     children.push(child);
                 }
-                children.push(Operation::Query(Query { prefix, kind: typos(word, authorize_typos) }));
+                children
+                    .push(Operation::Query(Query { prefix, kind: typos(word, authorize_typos) }));
                 Ok(Operation::or(false, children))
-            },
+            }
             // create a CONSECUTIVE operation wrapping all word in the phrase
-            PrimitiveQueryPart::Phrase(words) => {
-                Ok(Operation::phrase(words))
-            },
+            PrimitiveQueryPart::Phrase(words) => Ok(Operation::phrase(words)),
         }
     }
 
@@ -325,8 +330,7 @@ fn create_query_tree(
         ctx: &impl Context,
         authorize_typos: bool,
         query: &[PrimitiveQueryPart],
-    ) -> Result<Operation>
-    {
+    ) -> Result<Operation> {
         const MAX_NGRAM: usize = 3;
         let mut op_children = Vec::new();
 
@@ -341,21 +345,26 @@ fn create_query_tree(
 
                     match group {
                         [part] => {
-                            let operation = resolve_primitive_part(ctx, authorize_typos, part.clone())?;
+                            let operation =
+                                resolve_primitive_part(ctx, authorize_typos, part.clone())?;
                             and_op_children.push(operation);
-                        },
+                        }
                         words => {
                             let is_prefix = words.last().map_or(false, |part| part.is_prefix());
-                            let words: Vec<_> = words.iter().filter_map(|part| {
-                                if let PrimitiveQueryPart::Word(word, _) = part {
-                                    Some(word.as_str())
-                                } else {
-                                    None
-                                }
-                            }).collect();
+                            let words: Vec<_> = words
+                                .iter()
+                                .filter_map(|part| {
+                                    if let PrimitiveQueryPart::Word(word, _) = part {
+                                        Some(word.as_str())
+                                    } else {
+                                        None
+                                    }
+                                })
+                                .collect();
                             let mut operations = synonyms(ctx, &words)?.unwrap_or_default();
                             let concat = words.concat();
-                            let query = Query { prefix: is_prefix, kind: typos(concat, authorize_typos) };
+                            let query =
+                                Query { prefix: is_prefix, kind: typos(concat, authorize_typos) };
                             operations.push(Operation::Query(query));
                             and_op_children.push(Operation::or(false, operations));
                         }
@@ -379,26 +388,27 @@ fn create_query_tree(
         ctx: &impl Context,
         authorize_typos: bool,
         query: PrimitiveQuery,
-    ) -> Result<Operation>
-    {
+    ) -> Result<Operation> {
         let number_phrases = query.iter().filter(|p| p.is_phrase()).count();
         let mut operation_children = Vec::new();
 
         let start = number_phrases + (number_phrases == 0) as usize;
         for len in start..=query.len() {
             let mut word_count = len - number_phrases;
-            let query: Vec<_> = query.iter().filter(|p| {
-                if p.is_phrase() {
-                    true
-                } else if word_count != 0 {
-                    word_count -= 1;
-                    true
-                } else {
-                    false
-                }
-            })
-            .cloned()
-            .collect();
+            let query: Vec<_> = query
+                .iter()
+                .filter(|p| {
+                    if p.is_phrase() {
+                        true
+                    } else if word_count != 0 {
+                        word_count -= 1;
+                        true
+                    } else {
+                        false
+                    }
+                })
+                .cloned()
+                .collect();
 
             let ngrams = ngrams(ctx, authorize_typos, &query)?;
             operation_children.push(ngrams);
@@ -434,7 +444,11 @@ impl PrimitiveQueryPart {
 
 /// Create primitive query from tokenized query string,
 /// the primitive query is an intermediate state to build the query tree.
-fn create_primitive_query(query: TokenStream, stop_words: Option<Set<&[u8]>>, words_limit: Option<usize>) -> PrimitiveQuery {
+fn create_primitive_query(
+    query: TokenStream,
+    stop_words: Option<Set<&[u8]>>,
+    words_limit: Option<usize>,
+) -> PrimitiveQuery {
     let mut primitive_query = Vec::new();
     let mut phrase = Vec::new();
     let mut quoted = false;
@@ -444,23 +458,29 @@ fn create_primitive_query(query: TokenStream, stop_words: Option<Set<&[u8]>>, wo
     let mut peekable = query.peekable();
     while let Some(token) = peekable.next() {
         // early return if word limit is exceeded
-        if primitive_query.len() >= parts_limit { return primitive_query }
+        if primitive_query.len() >= parts_limit {
+            return primitive_query;
+        }
 
         match token.kind {
-            TokenKind::Word | TokenKind::StopWord  => {
+            TokenKind::Word | TokenKind::StopWord => {
                 // 1. if the word is quoted we push it in a phrase-buffer waiting for the ending quote,
                 // 2. if the word is not the last token of the query and is not a stop_word we push it as a non-prefix word,
                 // 3. if the word is the last token of the query we push it as a prefix word.
                 if quoted {
                     phrase.push(token.word.to_string());
                 } else if peekable.peek().is_some() {
-                     if !stop_words.as_ref().map_or(false, |swords| swords.contains(token.word.as_ref())) {
-                         primitive_query.push(PrimitiveQueryPart::Word(token.word.to_string(), false));
-                     }
+                    if !stop_words
+                        .as_ref()
+                        .map_or(false, |swords| swords.contains(token.word.as_ref()))
+                    {
+                        primitive_query
+                            .push(PrimitiveQueryPart::Word(token.word.to_string(), false));
+                    }
                 } else {
                     primitive_query.push(PrimitiveQueryPart::Word(token.word.to_string(), true));
                 }
-            },
+            }
             TokenKind::Separator(separator_kind) => {
                 let quote_count = token.word.chars().filter(|&s| s == '"').count();
                 // swap quoted state if we encounter a double quote
@@ -468,10 +488,11 @@ fn create_primitive_query(query: TokenStream, stop_words: Option<Set<&[u8]>>, wo
                     quoted = !quoted;
                 }
                 // if there is a quote or a hard separator we close the phrase.
-                if !phrase.is_empty() && (quote_count > 0 || separator_kind == SeparatorKind::Hard) {
+                if !phrase.is_empty() && (quote_count > 0 || separator_kind == SeparatorKind::Hard)
+                {
                     primitive_query.push(PrimitiveQueryPart::Phrase(mem::take(&mut phrase)));
                 }
-            },
+            }
             _ => (),
         }
     }
@@ -486,7 +507,7 @@ fn create_primitive_query(query: TokenStream, stop_words: Option<Set<&[u8]>>, wo
 
 /// Returns the maximum number of typos that this Operation allows.
 pub fn maximum_typo(operation: &Operation) -> usize {
-    use Operation::{Or, And, Query, Phrase};
+    use Operation::{And, Or, Phrase, Query};
     match operation {
         Or(_, ops) => ops.iter().map(maximum_typo).max().unwrap_or(0),
         And(ops) => ops.iter().map(maximum_typo).sum::<usize>(),
@@ -498,13 +519,12 @@ pub fn maximum_typo(operation: &Operation) -> usize {
 
 /// Returns the maximum proximity that this Operation allows.
 pub fn maximum_proximity(operation: &Operation) -> usize {
-    use Operation::{Or, And, Query, Phrase};
+    use Operation::{And, Or, Phrase, Query};
     match operation {
         Or(_, ops) => ops.iter().map(maximum_proximity).max().unwrap_or(0),
         And(ops) => {
-            ops.iter().map(maximum_proximity).sum::<usize>()
-            + ops.len().saturating_sub(1) * 7
-        },
+            ops.iter().map(maximum_proximity).sum::<usize>() + ops.len().saturating_sub(1) * 7
+        }
         Query(_) | Phrase(_) => 0,
     }
 }
@@ -515,7 +535,8 @@ mod test {
 
     use maplit::hashmap;
     use meilisearch_tokenizer::{Analyzer, AnalyzerConfig};
-    use rand::{Rng, SeedableRng, rngs::StdRng};
+    use rand::rngs::StdRng;
+    use rand::{Rng, SeedableRng};
 
     use super::*;
 
@@ -532,11 +553,11 @@ mod test {
             authorize_typos: bool,
             words_limit: Option<usize>,
             query: TokenStream,
-        ) -> Result<Option<(Operation, PrimitiveQuery)>>
-        {
+        ) -> Result<Option<(Operation, PrimitiveQuery)>> {
             let primitive_query = create_primitive_query(query, None, words_limit);
             if !primitive_query.is_empty() {
-                let qt = create_query_tree(self, optional_words, authorize_typos, &primitive_query)?;
+                let qt =
+                    create_query_tree(self, optional_words, authorize_typos, &primitive_query)?;
                 Ok(Some((qt, primitive_query)))
             } else {
                 Ok(None)
@@ -571,7 +592,7 @@ mod test {
             }
 
             TestContext {
-                synonyms: hashmap!{
+                synonyms: hashmap! {
                     vec![String::from("hello")] => vec![
                         vec![String::from("hi")],
                         vec![String::from("good"), String::from("morning")],
@@ -594,7 +615,7 @@ mod test {
                         vec![String::from("new"), String::from("york")],
                     ],
                 },
-                postings: hashmap!{
+                postings: hashmap! {
                     String::from("hello")      => random_postings(rng,   1500),
                     String::from("hi")         => random_postings(rng,   4000),
                     String::from("word")       => random_postings(rng,   2500),
@@ -620,15 +641,28 @@ mod test {
         let result = analyzer.analyze(query);
         let tokens = result.tokens();
 
-        let expected = Operation::Or(false, vec![
-            Operation::And(vec![
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact("hey".to_string()) }),
-                Operation::Query(Query { prefix: true, kind: QueryKind::tolerant(1, "friends".to_string()) }),
-            ]),
-            Operation::Query(Query { prefix: true, kind: QueryKind::tolerant(2, "heyfriends".to_string()) }),
-        ]);
+        let expected = Operation::Or(
+            false,
+            vec![
+                Operation::And(vec![
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("hey".to_string()),
+                    }),
+                    Operation::Query(Query {
+                        prefix: true,
+                        kind: QueryKind::tolerant(1, "friends".to_string()),
+                    }),
+                ]),
+                Operation::Query(Query {
+                    prefix: true,
+                    kind: QueryKind::tolerant(2, "heyfriends".to_string()),
+                }),
+            ],
+        );
 
-        let (query_tree, _) = TestContext::default().build(false, true, None, tokens).unwrap().unwrap();
+        let (query_tree, _) =
+            TestContext::default().build(false, true, None, tokens).unwrap().unwrap();
 
         assert_eq!(expected, query_tree);
     }
@@ -640,15 +674,28 @@ mod test {
         let result = analyzer.analyze(query);
         let tokens = result.tokens();
 
-        let expected = Operation::Or(false, vec![
-            Operation::And(vec![
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact("hey".to_string()) }),
-                Operation::Query(Query { prefix: false, kind: QueryKind::tolerant(1, "friends".to_string()) }),
-            ]),
-            Operation::Query(Query { prefix: false, kind: QueryKind::tolerant(2, "heyfriends".to_string()) }),
-        ]);
+        let expected = Operation::Or(
+            false,
+            vec![
+                Operation::And(vec![
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("hey".to_string()),
+                    }),
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::tolerant(1, "friends".to_string()),
+                    }),
+                ]),
+                Operation::Query(Query {
+                    prefix: false,
+                    kind: QueryKind::tolerant(2, "heyfriends".to_string()),
+                }),
+            ],
+        );
 
-        let (query_tree, _) = TestContext::default().build(false, true, None, tokens).unwrap().unwrap();
+        let (query_tree, _) =
+            TestContext::default().build(false, true, None, tokens).unwrap().unwrap();
 
         assert_eq!(expected, query_tree);
     }
@@ -660,26 +707,60 @@ mod test {
         let result = analyzer.analyze(query);
         let tokens = result.tokens();
 
-        let expected = Operation::Or(false, vec![
-            Operation::And(vec![
-                Operation::Or(false, vec![
-                    Operation::Query(Query { prefix: false, kind: QueryKind::exact("hi".to_string()) }),
-                    Operation::And(vec![
-                        Operation::Query(Query { prefix: false, kind: QueryKind::exact("good".to_string()) }),
-                        Operation::Query(Query { prefix: false, kind: QueryKind::exact("morning".to_string()) }),
-                    ]),
-                    Operation::Query(Query { prefix: false, kind: QueryKind::tolerant(1, "hello".to_string()) }),
+        let expected = Operation::Or(
+            false,
+            vec![
+                Operation::And(vec![
+                    Operation::Or(
+                        false,
+                        vec![
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::exact("hi".to_string()),
+                            }),
+                            Operation::And(vec![
+                                Operation::Query(Query {
+                                    prefix: false,
+                                    kind: QueryKind::exact("good".to_string()),
+                                }),
+                                Operation::Query(Query {
+                                    prefix: false,
+                                    kind: QueryKind::exact("morning".to_string()),
+                                }),
+                            ]),
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::tolerant(1, "hello".to_string()),
+                            }),
+                        ],
+                    ),
+                    Operation::Or(
+                        false,
+                        vec![
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::exact("earth".to_string()),
+                            }),
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::exact("nature".to_string()),
+                            }),
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::tolerant(1, "world".to_string()),
+                            }),
+                        ],
+                    ),
                 ]),
-                Operation::Or(false, vec![
-                    Operation::Query(Query { prefix: false, kind: QueryKind::exact("earth".to_string()) }),
-                    Operation::Query(Query { prefix: false, kind: QueryKind::exact("nature".to_string()) }),
-                    Operation::Query(Query { prefix: false, kind: QueryKind::tolerant(1, "world".to_string()) }),
-                ]),
-            ]),
-            Operation::Query(Query { prefix: false, kind: QueryKind::tolerant(2, "helloworld".to_string()) }),
-        ]);
+                Operation::Query(Query {
+                    prefix: false,
+                    kind: QueryKind::tolerant(2, "helloworld".to_string()),
+                }),
+            ],
+        );
 
-        let (query_tree, _) = TestContext::default().build(false, true, None, tokens).unwrap().unwrap();
+        let (query_tree, _) =
+            TestContext::default().build(false, true, None, tokens).unwrap().unwrap();
 
         assert_eq!(expected, query_tree);
     }
@@ -691,40 +772,95 @@ mod test {
         let result = analyzer.analyze(query);
         let tokens = result.tokens();
 
-        let expected = Operation::Or(false, vec![
-            Operation::And(vec![
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact("new".to_string()) }),
-                Operation::Or(false, vec![
-                    Operation::And(vec![
-                        Operation::Query(Query { prefix: false, kind: QueryKind::exact("york".to_string()) }),
-                        Operation::Query(Query { prefix: false, kind: QueryKind::exact("city".to_string()) }),
-                    ]),
-                    Operation::Query(Query { prefix: false, kind: QueryKind::tolerant(1, "yorkcity".to_string()) }),
-                ]),
-            ]),
-            Operation::And(vec![
-                Operation::Or(false, vec![
-                    Operation::Query(Query { prefix: false, kind: QueryKind::exact("nyc".to_string()) }),
-                    Operation::And(vec![
-                        Operation::Query(Query { prefix: false, kind: QueryKind::exact("new".to_string()) }),
-                        Operation::Query(Query { prefix: false, kind: QueryKind::exact("york".to_string()) }),
-                        Operation::Query(Query { prefix: false, kind: QueryKind::exact("city".to_string()) }),
-                    ]),
-                    Operation::Query(Query { prefix: false, kind: QueryKind::tolerant(1, "newyork".to_string()) }),
-                ]),
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact("city".to_string()) }),
-            ]),
-            Operation::Or(false, vec![
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact("nyc".to_string()) }),
+        let expected = Operation::Or(
+            false,
+            vec![
                 Operation::And(vec![
-                    Operation::Query(Query { prefix: false, kind: QueryKind::exact("new".to_string()) }),
-                    Operation::Query(Query { prefix: false, kind: QueryKind::exact("york".to_string()) }),
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("new".to_string()),
+                    }),
+                    Operation::Or(
+                        false,
+                        vec![
+                            Operation::And(vec![
+                                Operation::Query(Query {
+                                    prefix: false,
+                                    kind: QueryKind::exact("york".to_string()),
+                                }),
+                                Operation::Query(Query {
+                                    prefix: false,
+                                    kind: QueryKind::exact("city".to_string()),
+                                }),
+                            ]),
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::tolerant(1, "yorkcity".to_string()),
+                            }),
+                        ],
+                    ),
                 ]),
-                Operation::Query(Query { prefix: false, kind: QueryKind::tolerant(2, "newyorkcity".to_string()) }),
-            ]),
-        ]);
+                Operation::And(vec![
+                    Operation::Or(
+                        false,
+                        vec![
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::exact("nyc".to_string()),
+                            }),
+                            Operation::And(vec![
+                                Operation::Query(Query {
+                                    prefix: false,
+                                    kind: QueryKind::exact("new".to_string()),
+                                }),
+                                Operation::Query(Query {
+                                    prefix: false,
+                                    kind: QueryKind::exact("york".to_string()),
+                                }),
+                                Operation::Query(Query {
+                                    prefix: false,
+                                    kind: QueryKind::exact("city".to_string()),
+                                }),
+                            ]),
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::tolerant(1, "newyork".to_string()),
+                            }),
+                        ],
+                    ),
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("city".to_string()),
+                    }),
+                ]),
+                Operation::Or(
+                    false,
+                    vec![
+                        Operation::Query(Query {
+                            prefix: false,
+                            kind: QueryKind::exact("nyc".to_string()),
+                        }),
+                        Operation::And(vec![
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::exact("new".to_string()),
+                            }),
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::exact("york".to_string()),
+                            }),
+                        ]),
+                        Operation::Query(Query {
+                            prefix: false,
+                            kind: QueryKind::tolerant(2, "newyorkcity".to_string()),
+                        }),
+                    ],
+                ),
+            ],
+        );
 
-        let (query_tree, _) = TestContext::default().build(false, true, None, tokens).unwrap().unwrap();
+        let (query_tree, _) =
+            TestContext::default().build(false, true, None, tokens).unwrap().unwrap();
 
         assert_eq!(expected, query_tree);
     }
@@ -736,15 +872,28 @@ mod test {
         let result = analyzer.analyze(query);
         let tokens = result.tokens();
 
-        let expected = Operation::Or(false, vec![
-            Operation::And(vec![
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact("n".to_string()) }),
-                Operation::Query(Query { prefix: false, kind: QueryKind::tolerant(1, "grams".to_string()) }),
-            ]),
-            Operation::Query(Query { prefix: false, kind: QueryKind::tolerant(1, "ngrams".to_string()) }),
-        ]);
+        let expected = Operation::Or(
+            false,
+            vec![
+                Operation::And(vec![
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("n".to_string()),
+                    }),
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::tolerant(1, "grams".to_string()),
+                    }),
+                ]),
+                Operation::Query(Query {
+                    prefix: false,
+                    kind: QueryKind::tolerant(1, "ngrams".to_string()),
+                }),
+            ],
+        );
 
-        let (query_tree, _) = TestContext::default().build(false, true, None, tokens).unwrap().unwrap();
+        let (query_tree, _) =
+            TestContext::default().build(false, true, None, tokens).unwrap().unwrap();
 
         assert_eq!(expected, query_tree);
     }
@@ -756,21 +905,34 @@ mod test {
         let result = analyzer.analyze(query);
         let tokens = result.tokens();
 
-        let expected = Operation::Or(false, vec![
-            Operation::And(vec![
-                Operation::Or(false, vec![
-                    Operation::Phrase(vec![
-                        "word".to_string(),
-                        "split".to_string(),
-                    ]),
-                    Operation::Query(Query { prefix: false, kind: QueryKind::tolerant(2, "wordsplit".to_string()) }),
+        let expected = Operation::Or(
+            false,
+            vec![
+                Operation::And(vec![
+                    Operation::Or(
+                        false,
+                        vec![
+                            Operation::Phrase(vec!["word".to_string(), "split".to_string()]),
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::tolerant(2, "wordsplit".to_string()),
+                            }),
+                        ],
+                    ),
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("fish".to_string()),
+                    }),
                 ]),
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact("fish".to_string()) })
-            ]),
-            Operation::Query(Query { prefix: false, kind: QueryKind::tolerant(2, "wordsplitfish".to_string()) }),
-        ]);
+                Operation::Query(Query {
+                    prefix: false,
+                    kind: QueryKind::tolerant(2, "wordsplitfish".to_string()),
+                }),
+            ],
+        );
 
-        let (query_tree, _) = TestContext::default().build(false, true, None, tokens).unwrap().unwrap();
+        let (query_tree, _) =
+            TestContext::default().build(false, true, None, tokens).unwrap().unwrap();
 
         assert_eq!(expected, query_tree);
     }
@@ -783,14 +945,12 @@ mod test {
         let tokens = result.tokens();
 
         let expected = Operation::And(vec![
-            Operation::Phrase(vec![
-                "hey".to_string(),
-                "friends".to_string(),
-            ]),
+            Operation::Phrase(vec!["hey".to_string(), "friends".to_string()]),
             Operation::Query(Query { prefix: false, kind: QueryKind::exact("wooop".to_string()) }),
         ]);
 
-        let (query_tree, _) = TestContext::default().build(false, true, None, tokens).unwrap().unwrap();
+        let (query_tree, _) =
+            TestContext::default().build(false, true, None, tokens).unwrap().unwrap();
 
         assert_eq!(expected, query_tree);
     }
@@ -803,17 +963,12 @@ mod test {
         let tokens = result.tokens();
 
         let expected = Operation::And(vec![
-            Operation::Phrase(vec![
-                "hey".to_string(),
-                "friends".to_string(),
-            ]),
-            Operation::Phrase(vec![
-                "wooop".to_string(),
-                "wooop".to_string(),
-            ]),
+            Operation::Phrase(vec!["hey".to_string(), "friends".to_string()]),
+            Operation::Phrase(vec!["wooop".to_string(), "wooop".to_string()]),
         ]);
 
-        let (query_tree, _) = TestContext::default().build(false, true, None, tokens).unwrap().unwrap();
+        let (query_tree, _) =
+            TestContext::default().build(false, true, None, tokens).unwrap().unwrap();
 
         assert_eq!(expected, query_tree);
     }
@@ -825,34 +980,80 @@ mod test {
         let result = analyzer.analyze(query);
         let tokens = result.tokens();
 
-        let expected = Operation::Or(true, vec![
-            Operation::Query(Query { prefix: false, kind: QueryKind::exact("hey".to_string()) }),
-            Operation::Or(false, vec![
-                Operation::And(vec![
-                    Operation::Query(Query { prefix: false, kind: QueryKind::exact("hey".to_string()) }),
-                    Operation::Query(Query { prefix: false, kind: QueryKind::exact("my".to_string()) }),
-                ]),
-                Operation::Query(Query { prefix: false, kind: QueryKind::tolerant(1, "heymy".to_string()) }),
-            ]),
-            Operation::Or(false, vec![
-                Operation::And(vec![
-                    Operation::Query(Query { prefix: false, kind: QueryKind::exact("hey".to_string()) }),
-                    Operation::Or(false, vec![
+        let expected = Operation::Or(
+            true,
+            vec![
+                Operation::Query(Query {
+                    prefix: false,
+                    kind: QueryKind::exact("hey".to_string()),
+                }),
+                Operation::Or(
+                    false,
+                    vec![
                         Operation::And(vec![
-                            Operation::Query(Query { prefix: false, kind: QueryKind::exact("my".to_string()) }),
-                            Operation::Query(Query { prefix: false, kind: QueryKind::tolerant(1, "friend".to_string()) }),
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::exact("hey".to_string()),
+                            }),
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::exact("my".to_string()),
+                            }),
                         ]),
-                        Operation::Query(Query { prefix: false, kind: QueryKind::tolerant(1, "myfriend".to_string()) })
-                    ])
-                ]),
-                Operation::And(vec![
-                    Operation::Query(Query { prefix: false, kind: QueryKind::tolerant(1, "heymy".to_string()) }),
-                    Operation::Query(Query { prefix: false, kind: QueryKind::tolerant(1, "friend".to_string()) }),
-                ]),
-                Operation::Query(Query { prefix: false, kind: QueryKind::tolerant(2, "heymyfriend".to_string()) }),
-            ]),
-        ]);
-        let (query_tree, _) = TestContext::default().build(true, true, None, tokens).unwrap().unwrap();
+                        Operation::Query(Query {
+                            prefix: false,
+                            kind: QueryKind::tolerant(1, "heymy".to_string()),
+                        }),
+                    ],
+                ),
+                Operation::Or(
+                    false,
+                    vec![
+                        Operation::And(vec![
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::exact("hey".to_string()),
+                            }),
+                            Operation::Or(
+                                false,
+                                vec![
+                                    Operation::And(vec![
+                                        Operation::Query(Query {
+                                            prefix: false,
+                                            kind: QueryKind::exact("my".to_string()),
+                                        }),
+                                        Operation::Query(Query {
+                                            prefix: false,
+                                            kind: QueryKind::tolerant(1, "friend".to_string()),
+                                        }),
+                                    ]),
+                                    Operation::Query(Query {
+                                        prefix: false,
+                                        kind: QueryKind::tolerant(1, "myfriend".to_string()),
+                                    }),
+                                ],
+                            ),
+                        ]),
+                        Operation::And(vec![
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::tolerant(1, "heymy".to_string()),
+                            }),
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::tolerant(1, "friend".to_string()),
+                            }),
+                        ]),
+                        Operation::Query(Query {
+                            prefix: false,
+                            kind: QueryKind::tolerant(2, "heymyfriend".to_string()),
+                        }),
+                    ],
+                ),
+            ],
+        );
+        let (query_tree, _) =
+            TestContext::default().build(true, true, None, tokens).unwrap().unwrap();
 
         assert_eq!(expected, query_tree);
     }
@@ -864,11 +1065,9 @@ mod test {
         let result = analyzer.analyze(query);
         let tokens = result.tokens();
 
-        let expected = Operation::Phrase(vec![
-            "hey".to_string(),
-            "my".to_string(),
-        ]);
-        let (query_tree, _) = TestContext::default().build(true, true, None, tokens).unwrap().unwrap();
+        let expected = Operation::Phrase(vec!["hey".to_string(), "my".to_string()]);
+        let (query_tree, _) =
+            TestContext::default().build(true, true, None, tokens).unwrap().unwrap();
 
         assert_eq!(expected, query_tree);
     }
@@ -880,29 +1079,66 @@ mod test {
         let result = analyzer.analyze(query);
         let tokens = result.tokens();
 
-        let expected = Operation::Or(true, vec![
-            Operation::And(vec![
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact("hey".to_string()) }),
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact("friend".to_string()) }),
-            ]),
-            Operation::And(vec![
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact("hey".to_string()) }),
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact("my".to_string()) }),
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact("friend".to_string()) }),
-            ]),
-            Operation::And(vec![
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact("hey".to_string()) }),
-                Operation::Or(false, vec![
-                    Operation::And(vec![
-                        Operation::Query(Query { prefix: false, kind: QueryKind::exact("my".to_string()) }),
-                        Operation::Query(Query { prefix: false, kind: QueryKind::exact("good".to_string()) }),
-                    ]),
-                    Operation::Query(Query { prefix: false, kind: QueryKind::tolerant(1, "mygood".to_string()) }),
+        let expected = Operation::Or(
+            true,
+            vec![
+                Operation::And(vec![
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("hey".to_string()),
+                    }),
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("friend".to_string()),
+                    }),
                 ]),
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact("friend".to_string()) }),
-            ]),
-        ]);
-        let (query_tree, _) = TestContext::default().build(true, true, None, tokens).unwrap().unwrap();
+                Operation::And(vec![
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("hey".to_string()),
+                    }),
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("my".to_string()),
+                    }),
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("friend".to_string()),
+                    }),
+                ]),
+                Operation::And(vec![
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("hey".to_string()),
+                    }),
+                    Operation::Or(
+                        false,
+                        vec![
+                            Operation::And(vec![
+                                Operation::Query(Query {
+                                    prefix: false,
+                                    kind: QueryKind::exact("my".to_string()),
+                                }),
+                                Operation::Query(Query {
+                                    prefix: false,
+                                    kind: QueryKind::exact("good".to_string()),
+                                }),
+                            ]),
+                            Operation::Query(Query {
+                                prefix: false,
+                                kind: QueryKind::tolerant(1, "mygood".to_string()),
+                            }),
+                        ],
+                    ),
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("friend".to_string()),
+                    }),
+                ]),
+            ],
+        );
+        let (query_tree, _) =
+            TestContext::default().build(true, true, None, tokens).unwrap().unwrap();
 
         assert_eq!(expected, query_tree);
     }
@@ -914,14 +1150,27 @@ mod test {
         let result = analyzer.analyze(query);
         let tokens = result.tokens();
 
-        let expected = Operation::Or(false, vec![
-            Operation::And(vec![
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact("hey".to_string()) }),
-                Operation::Query(Query { prefix: false, kind: QueryKind::exact("friends".to_string()) }),
-            ]),
-            Operation::Query(Query { prefix: false, kind: QueryKind::exact("heyfriends".to_string()) }),
-        ]);
-        let (query_tree, _) = TestContext::default().build(false, false, None, tokens).unwrap().unwrap();
+        let expected = Operation::Or(
+            false,
+            vec![
+                Operation::And(vec![
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("hey".to_string()),
+                    }),
+                    Operation::Query(Query {
+                        prefix: false,
+                        kind: QueryKind::exact("friends".to_string()),
+                    }),
+                ]),
+                Operation::Query(Query {
+                    prefix: false,
+                    kind: QueryKind::exact("heyfriends".to_string()),
+                }),
+            ],
+        );
+        let (query_tree, _) =
+            TestContext::default().build(false, false, None, tokens).unwrap().unwrap();
 
         assert_eq!(expected, query_tree);
     }
@@ -934,14 +1183,12 @@ mod test {
         let tokens = result.tokens();
 
         let expected = Operation::And(vec![
-            Operation::Phrase(vec![
-                "hey".to_string(),
-                "my".to_string(),
-            ]),
+            Operation::Phrase(vec!["hey".to_string(), "my".to_string()]),
             Operation::Query(Query { prefix: false, kind: QueryKind::exact("good".to_string()) }),
         ]);
 
-        let (query_tree, _) = TestContext::default().build(false, false, Some(2), tokens).unwrap().unwrap();
+        let (query_tree, _) =
+            TestContext::default().build(false, false, Some(2), tokens).unwrap().unwrap();
 
         assert_eq!(expected, query_tree);
     }
diff --git a/milli/src/update/available_documents_ids.rs b/milli/src/update/available_documents_ids.rs
index 34ff743f0..9e3fce75d 100644
--- a/milli/src/update/available_documents_ids.rs
+++ b/milli/src/update/available_documents_ids.rs
@@ -1,6 +1,7 @@
 use std::iter::{Chain, FromIterator};
 use std::ops::RangeInclusive;
-use roaring::bitmap::{RoaringBitmap, IntoIter};
+
+use roaring::bitmap::{IntoIter, RoaringBitmap};
 
 pub struct AvailableDocumentsIds {
     iter: Chain<IntoIter, RangeInclusive<u32>>,
@@ -18,16 +19,12 @@ impl AvailableDocumentsIds {
                     None => 1..=0, // empty range iterator
                 };
 
-                AvailableDocumentsIds {
-                    iter: available.into_iter().chain(iter),
-                }
-            },
+                AvailableDocumentsIds { iter: available.into_iter().chain(iter) }
+            }
             None => {
                 let empty = RoaringBitmap::new().into_iter();
-                AvailableDocumentsIds {
-                    iter: empty.chain(0..=u32::max_value()),
-                }
-            },
+                AvailableDocumentsIds { iter: empty.chain(0..=u32::max_value()) }
+            }
         }
     }
 }
diff --git a/milli/src/update/clear_documents.rs b/milli/src/update/clear_documents.rs
index 6e26bf027..42dd55443 100644
--- a/milli/src/update/clear_documents.rs
+++ b/milli/src/update/clear_documents.rs
@@ -1,7 +1,7 @@
 use chrono::Utc;
 use roaring::RoaringBitmap;
 
-use crate::{ExternalDocumentsIds, Index, FieldsDistribution, Result};
+use crate::{ExternalDocumentsIds, FieldsDistribution, Index, Result};
 
 pub struct ClearDocuments<'t, 'u, 'i> {
     wtxn: &'t mut heed::RwTxn<'i, 'u>,
@@ -13,9 +13,8 @@ impl<'t, 'u, 'i> ClearDocuments<'t, 'u, 'i> {
     pub fn new(
         wtxn: &'t mut heed::RwTxn<'i, 'u>,
         index: &'i Index,
-        update_id: u64
+        update_id: u64,
     ) -> ClearDocuments<'t, 'u, 'i> {
-
         ClearDocuments { wtxn, index, _update_id: update_id }
     }
 
@@ -80,8 +79,8 @@ impl<'t, 'u, 'i> ClearDocuments<'t, 'u, 'i> {
 mod tests {
     use heed::EnvOpenOptions;
 
-    use crate::update::{IndexDocuments, UpdateFormat};
     use super::*;
+    use crate::update::{IndexDocuments, UpdateFormat};
 
     #[test]
     fn clear_documents() {
diff --git a/milli/src/update/delete_documents.rs b/milli/src/update/delete_documents.rs
index 7fc7e5d77..dfb48dc58 100644
--- a/milli/src/update/delete_documents.rs
+++ b/milli/src/update/delete_documents.rs
@@ -1,5 +1,5 @@
-use std::collections::HashMap;
 use std::collections::hash_map::Entry;
+use std::collections::HashMap;
 
 use chrono::Utc;
 use fst::IntoStreamer;
@@ -7,11 +7,11 @@ use heed::types::{ByteSlice, Unit};
 use roaring::RoaringBitmap;
 use serde_json::Value;
 
-use crate::error::{InternalError, FieldIdMapMissingEntry, UserError};
+use super::ClearDocuments;
+use crate::error::{FieldIdMapMissingEntry, InternalError, UserError};
 use crate::heed_codec::CboRoaringBitmapCodec;
 use crate::index::{db_name, main_key};
-use crate::{Index, DocumentId, FieldId, BEU32, SmallString32, ExternalDocumentsIds, Result};
-use super::ClearDocuments;
+use crate::{DocumentId, ExternalDocumentsIds, FieldId, Index, Result, SmallString32, BEU32};
 
 pub struct DeleteDocuments<'t, 'u, 'i> {
     wtxn: &'t mut heed::RwTxn<'i, 'u>,
@@ -26,11 +26,8 @@ impl<'t, 'u, 'i> DeleteDocuments<'t, 'u, 'i> {
         wtxn: &'t mut heed::RwTxn<'i, 'u>,
         index: &'i Index,
         update_id: u64,
-    ) -> Result<DeleteDocuments<'t, 'u, 'i>>
-    {
-        let external_documents_ids = index
-            .external_documents_ids(wtxn)?
-            .into_static();
+    ) -> Result<DeleteDocuments<'t, 'u, 'i>> {
+        let external_documents_ids = index.external_documents_ids(wtxn)?.into_static();
 
         Ok(DeleteDocuments {
             wtxn,
@@ -84,12 +81,11 @@ impl<'t, 'u, 'i> DeleteDocuments<'t, 'u, 'i> {
                 key: Some(main_key::PRIMARY_KEY_KEY),
             }
         })?;
-        let id_field = fields_ids_map.id(primary_key).ok_or_else(|| {
-            FieldIdMapMissingEntry::FieldName {
+        let id_field =
+            fields_ids_map.id(primary_key).ok_or_else(|| FieldIdMapMissingEntry::FieldName {
                 field_name: primary_key.to_string(),
                 process: "DeleteDocuments::execute",
-            }
-        })?;
+            })?;
 
         let Index {
             env: _env,
@@ -130,7 +126,9 @@ impl<'t, 'u, 'i> DeleteDocuments<'t, 'u, 'i> {
                     let external_id = match serde_json::from_slice(content).unwrap() {
                         Value::String(string) => SmallString32::from(string.as_str()),
                         Value::Number(number) => SmallString32::from(number.to_string()),
-                        document_id => return Err(UserError::InvalidDocumentId { document_id }.into()),
+                        document_id => {
+                            return Err(UserError::InvalidDocumentId { document_id }.into())
+                        }
                     };
                     external_ids.push(external_id);
                 }
@@ -160,7 +158,7 @@ impl<'t, 'u, 'i> DeleteDocuments<'t, 'u, 'i> {
             if let Entry::Occupied(mut entry) = fields_distribution.entry(field_name.to_string()) {
                 match entry.get().checked_sub(count_diff) {
                     Some(0) | None => entry.remove(),
-                    Some(count) => entry.insert(count)
+                    Some(count) => entry.insert(count),
                 };
             }
         }
@@ -206,9 +204,16 @@ impl<'t, 'u, 'i> DeleteDocuments<'t, 'u, 'i> {
         }
 
         // We construct an FST set that contains the words to delete from the words FST.
-        let words_to_delete = words.iter().filter_map(|(word, must_remove)| {
-            if *must_remove { Some(word.as_ref()) } else { None }
-        });
+        let words_to_delete =
+            words.iter().filter_map(
+                |(word, must_remove)| {
+                    if *must_remove {
+                        Some(word.as_ref())
+                    } else {
+                        None
+                    }
+                },
+            );
         let words_to_delete = fst::Set::from_iter(words_to_delete)?;
 
         let new_words_fst = {
@@ -285,7 +290,8 @@ impl<'t, 'u, 'i> DeleteDocuments<'t, 'u, 'i> {
         // We delete the documents ids that are under the pairs of words,
         // it is faster and use no memory to iterate over all the words pairs than
         // to compute the cartesian product of every words of the deleted documents.
-        let mut iter = word_pair_proximity_docids.remap_key_type::<ByteSlice>().iter_mut(self.wtxn)?;
+        let mut iter =
+            word_pair_proximity_docids.remap_key_type::<ByteSlice>().iter_mut(self.wtxn)?;
         while let Some(result) = iter.next() {
             let (bytes, mut docids) = result?;
             let previous_len = docids.len();
@@ -300,7 +306,8 @@ impl<'t, 'u, 'i> DeleteDocuments<'t, 'u, 'i> {
         drop(iter);
 
         // We delete the documents ids that are under the word level position docids.
-        let mut iter = word_level_position_docids.iter_mut(self.wtxn)?.remap_key_type::<ByteSlice>();
+        let mut iter =
+            word_level_position_docids.iter_mut(self.wtxn)?.remap_key_type::<ByteSlice>();
         while let Some(result) = iter.next() {
             let (bytes, mut docids) = result?;
             let previous_len = docids.len();
@@ -315,7 +322,8 @@ impl<'t, 'u, 'i> DeleteDocuments<'t, 'u, 'i> {
         drop(iter);
 
         // We delete the documents ids that are under the word prefix level position docids.
-        let mut iter = word_prefix_level_position_docids.iter_mut(self.wtxn)?.remap_key_type::<ByteSlice>();
+        let mut iter =
+            word_prefix_level_position_docids.iter_mut(self.wtxn)?.remap_key_type::<ByteSlice>();
         while let Some(result) = iter.next() {
             let (bytes, mut docids) = result?;
             let previous_len = docids.len();
@@ -397,12 +405,11 @@ fn remove_docids_from_field_id_docid_facet_value<'a, C, K, F>(
     convert: F,
 ) -> heed::Result<()>
 where
-    C: heed::BytesDecode<'a, DItem=K> + heed::BytesEncode<'a, EItem=K>,
+    C: heed::BytesDecode<'a, DItem = K> + heed::BytesEncode<'a, EItem = K>,
     F: Fn(K) -> DocumentId,
 {
-    let mut iter = db.remap_key_type::<ByteSlice>()
-        .prefix_iter_mut(wtxn, &[field_id])?
-        .remap_key_type::<C>();
+    let mut iter =
+        db.remap_key_type::<ByteSlice>().prefix_iter_mut(wtxn, &[field_id])?.remap_key_type::<C>();
 
     while let Some(result) = iter.next() {
         let (key, ()) = result?;
@@ -441,8 +448,8 @@ where
 mod tests {
     use heed::EnvOpenOptions;
 
-    use crate::update::{IndexDocuments, UpdateFormat};
     use super::*;
+    use crate::update::{IndexDocuments, UpdateFormat};
 
     #[test]
     fn delete_documents_with_numbers_as_primary_key() {
diff --git a/milli/src/update/facets.rs b/milli/src/update/facets.rs
index 757cbe810..09f962bbc 100644
--- a/milli/src/update/facets.rs
+++ b/milli/src/update/facets.rs
@@ -3,17 +3,18 @@ use std::fs::File;
 use std::num::NonZeroUsize;
 
 use chrono::Utc;
-use grenad::{CompressionType, Reader, Writer, FileFuse};
+use grenad::{CompressionType, FileFuse, Reader, Writer};
 use heed::types::{ByteSlice, DecodeIgnore};
 use heed::{BytesEncode, Error};
 use log::debug;
 use roaring::RoaringBitmap;
 
 use crate::error::InternalError;
-use crate::heed_codec::CboRoaringBitmapCodec;
 use crate::heed_codec::facet::FacetLevelValueF64Codec;
-use crate::update::index_documents::WriteMethod;
-use crate::update::index_documents::{create_writer, writer_into_reader, write_into_lmdb_database};
+use crate::heed_codec::CboRoaringBitmapCodec;
+use crate::update::index_documents::{
+    create_writer, write_into_lmdb_database, writer_into_reader, WriteMethod,
+};
 use crate::{Index, Result};
 
 pub struct Facets<'t, 'u, 'i> {
@@ -32,8 +33,7 @@ impl<'t, 'u, 'i> Facets<'t, 'u, 'i> {
         wtxn: &'t mut heed::RwTxn<'i, 'u>,
         index: &'i Index,
         update_id: u64,
-    ) -> Facets<'t, 'u, 'i>
-    {
+    ) -> Facets<'t, 'u, 'i> {
         Facets {
             wtxn,
             index,
@@ -72,11 +72,7 @@ impl<'t, 'u, 'i> Facets<'t, 'u, 'i> {
             )?;
 
             // Clear the facet number levels.
-            clear_field_number_levels(
-                self.wtxn,
-                self.index.facet_id_f64_docids,
-                field_id,
-            )?;
+            clear_field_number_levels(self.wtxn, self.index.facet_id_f64_docids, field_id)?;
 
             // Compute and store the faceted numbers documents ids.
             let number_documents_ids = compute_faceted_documents_ids(
@@ -96,8 +92,16 @@ impl<'t, 'u, 'i> Facets<'t, 'u, 'i> {
                 field_id,
             )?;
 
-            self.index.put_string_faceted_documents_ids(self.wtxn, field_id, &string_documents_ids)?;
-            self.index.put_number_faceted_documents_ids(self.wtxn, field_id, &number_documents_ids)?;
+            self.index.put_string_faceted_documents_ids(
+                self.wtxn,
+                field_id,
+                &string_documents_ids,
+            )?;
+            self.index.put_number_faceted_documents_ids(
+                self.wtxn,
+                field_id,
+                &number_documents_ids,
+            )?;
 
             write_into_lmdb_database(
                 self.wtxn,
@@ -112,12 +116,11 @@ impl<'t, 'u, 'i> Facets<'t, 'u, 'i> {
     }
 }
 
-fn clear_field_number_levels<'t, >(
+fn clear_field_number_levels<'t>(
     wtxn: &'t mut heed::RwTxn,
     db: heed::Database<FacetLevelValueF64Codec, CboRoaringBitmapCodec>,
     field_id: u8,
-) -> heed::Result<()>
-{
+) -> heed::Result<()> {
     let left = (field_id, 1, f64::MIN, f64::MIN);
     let right = (field_id, u8::MAX, f64::MAX, f64::MAX);
     let range = left..=right;
@@ -133,8 +136,7 @@ fn compute_facet_number_levels<'t>(
     level_group_size: NonZeroUsize,
     min_level_size: NonZeroUsize,
     field_id: u8,
-) -> Result<Reader<FileFuse>>
-{
+) -> Result<Reader<FileFuse>> {
     let first_level_size = db
         .remap_key_type::<ByteSlice>()
         .prefix_iter(rtxn, &[field_id])?
@@ -143,9 +145,8 @@ fn compute_facet_number_levels<'t>(
 
     // It is forbidden to keep a cursor and write in a database at the same time with LMDB
     // therefore we write the facet levels entries into a grenad file before transfering them.
-    let mut writer = tempfile::tempfile().and_then(|file| {
-        create_writer(compression_type, compression_level, file)
-    })?;
+    let mut writer = tempfile::tempfile()
+        .and_then(|file| create_writer(compression_type, compression_level, file))?;
 
     let level_0_range = {
         let left = (field_id, 0, f64::MIN, f64::MIN);
@@ -196,8 +197,7 @@ fn compute_faceted_documents_ids(
     rtxn: &heed::RoTxn,
     db: heed::Database<ByteSlice, CboRoaringBitmapCodec>,
     field_id: u8,
-) -> Result<RoaringBitmap>
-{
+) -> Result<RoaringBitmap> {
     let mut documents_ids = RoaringBitmap::new();
 
     for result in db.prefix_iter(rtxn, &[field_id])? {
@@ -215,8 +215,7 @@ fn write_number_entry(
     left: f64,
     right: f64,
     ids: &RoaringBitmap,
-) -> Result<()>
-{
+) -> Result<()> {
     let key = (field_id, level, left, right);
     let key = FacetLevelValueF64Codec::bytes_encode(&key).ok_or(Error::Encoding)?;
     let data = CboRoaringBitmapCodec::bytes_encode(&ids).ok_or(Error::Encoding)?;
diff --git a/milli/src/update/index_documents/mod.rs b/milli/src/update/index_documents/mod.rs
index 51c8b948a..05242f540 100644
--- a/milli/src/update/index_documents/mod.rs
+++ b/milli/src/update/index_documents/mod.rs
@@ -1,7 +1,7 @@
 use std::borrow::Cow;
 use std::collections::HashSet;
 use std::fs::File;
-use std::io::{self, Seek, SeekFrom, BufReader, BufRead};
+use std::io::{self, BufRead, BufReader, Seek, SeekFrom};
 use std::num::{NonZeroU32, NonZeroUsize};
 use std::result::Result as StdResult;
 use std::str;
@@ -10,28 +10,26 @@ use std::time::Instant;
 
 use bstr::ByteSlice as _;
 use chrono::Utc;
-use grenad::{MergerIter, Writer, Sorter, Merger, Reader, FileFuse, CompressionType};
+use grenad::{CompressionType, FileFuse, Merger, MergerIter, Reader, Sorter, Writer};
 use heed::types::ByteSlice;
-use log::{debug, info, error};
+use log::{debug, error, info};
 use memmap::Mmap;
 use rayon::prelude::*;
 use rayon::ThreadPool;
-use serde::{Serialize, Deserialize};
+use serde::{Deserialize, Serialize};
 
-use crate::error::{Error, InternalError};
-use crate::{Index, Result};
-use crate::update::{
-    Facets, WordsLevelPositions, WordPrefixDocids, WordsPrefixesFst, UpdateIndexingStep,
-    WordPrefixPairProximityDocids,
-};
-use self::store::{Store, Readers};
 pub use self::merge_function::{
-    fst_merge, cbo_roaring_bitmap_merge, roaring_bitmap_merge, keep_first
+    cbo_roaring_bitmap_merge, fst_merge, keep_first, roaring_bitmap_merge,
 };
+use self::store::{Readers, Store};
 pub use self::transform::{Transform, TransformOutput};
-
-use crate::MergeFn;
 use super::UpdateBuilder;
+use crate::error::{Error, InternalError};
+use crate::update::{
+    Facets, UpdateIndexingStep, WordPrefixDocids, WordPrefixPairProximityDocids,
+    WordsLevelPositions, WordsPrefixesFst,
+};
+use crate::{Index, MergeFn, Result};
 
 mod merge_function;
 mod store;
@@ -48,7 +46,11 @@ pub enum WriteMethod {
     GetMergePut,
 }
 
-pub fn create_writer(typ: CompressionType, level: Option<u32>, file: File) -> io::Result<Writer<File>> {
+pub fn create_writer(
+    typ: CompressionType,
+    level: Option<u32>,
+    file: File,
+) -> io::Result<Writer<File>> {
     let mut builder = Writer::builder();
     builder.compression_type(typ);
     if let Some(level) = level {
@@ -64,8 +66,7 @@ pub fn create_sorter<E>(
     chunk_fusing_shrink_size: Option<u64>,
     max_nb_chunks: Option<usize>,
     max_memory: Option<usize>,
-) -> Sorter<MergeFn<E>>
-{
+) -> Sorter<MergeFn<E>> {
     let mut builder = Sorter::builder(merge);
     if let Some(shrink_size) = chunk_fusing_shrink_size {
         builder.file_fusing_shrink_size(shrink_size);
@@ -83,7 +84,10 @@ pub fn create_sorter<E>(
     builder.build()
 }
 
-pub fn writer_into_reader(writer: Writer<File>, shrink_size: Option<u64>) -> Result<Reader<FileFuse>> {
+pub fn writer_into_reader(
+    writer: Writer<File>,
+    shrink_size: Option<u64>,
+) -> Result<Reader<FileFuse>> {
     let mut file = writer.into_inner()?;
     file.seek(SeekFrom::Start(0))?;
     let file = if let Some(shrink_size) = shrink_size {
@@ -97,8 +101,7 @@ pub fn writer_into_reader(writer: Writer<File>, shrink_size: Option<u64>) -> Res
 pub fn merge_readers<E>(
     sources: Vec<Reader<FileFuse>>,
     merge: MergeFn<E>,
-) -> Merger<FileFuse, MergeFn<E>>
-{
+) -> Merger<FileFuse, MergeFn<E>> {
     let mut builder = Merger::builder(merge);
     builder.extend(sources);
     builder.build()
@@ -118,13 +121,7 @@ where
     let before = Instant::now();
 
     let merger = merge_readers(sources, merge);
-    merger_iter_into_lmdb_database(
-        wtxn,
-        database,
-        merger.into_merge_iter()?,
-        merge,
-        method,
-    )?;
+    merger_iter_into_lmdb_database(wtxn, database, merger.into_merge_iter()?, merge, method)?;
 
     debug!("MTBL stores merged in {:.02?}!", before.elapsed());
     Ok(())
@@ -149,7 +146,7 @@ where
             while let Some((k, v)) = reader.next()? {
                 out_iter.append(k, v)?;
             }
-        },
+        }
         WriteMethod::GetMergePut => {
             while let Some((k, v)) = reader.next()? {
                 let mut iter = database.prefix_iter_mut::<_, ByteSlice, ByteSlice>(wtxn, k)?;
@@ -158,11 +155,11 @@ where
                         let vals = &[Cow::Borrowed(old_val), Cow::Borrowed(v)][..];
                         let val = merge(k, &vals)?;
                         iter.put_current(k, &val)?;
-                    },
+                    }
                     _ => {
                         drop(iter);
                         database.put::<_, ByteSlice, ByteSlice>(wtxn, k, v)?;
-                    },
+                    }
                 }
             }
         }
@@ -181,18 +178,12 @@ pub fn sorter_into_lmdb_database<E>(
 ) -> Result<()>
 where
     Error: From<E>,
-    Error: From<grenad::Error<E>>
+    Error: From<grenad::Error<E>>,
 {
     debug!("Writing MTBL sorter...");
     let before = Instant::now();
 
-    merger_iter_into_lmdb_database(
-        wtxn,
-        database,
-        sorter.into_iter()?,
-        merge,
-        method,
-    )?;
+    merger_iter_into_lmdb_database(wtxn, database, sorter.into_iter()?, merge, method)?;
 
     debug!("MTBL sorter writen in {:.02?}!", before.elapsed());
     Ok(())
@@ -214,7 +205,7 @@ where
             while let Some((k, v)) = sorter.next()? {
                 out_iter.append(k, v)?;
             }
-        },
+        }
         WriteMethod::GetMergePut => {
             while let Some((k, v)) = sorter.next()? {
                 let mut iter = database.prefix_iter_mut::<_, ByteSlice, ByteSlice>(wtxn, k)?;
@@ -226,14 +217,14 @@ where
                             InternalError::IndexingMergingKeys { process: "get-put-merge" }
                         })?;
                         iter.put_current(k, &val)?;
-                    },
+                    }
                     _ => {
                         drop(iter);
                         database.put::<_, ByteSlice, ByteSlice>(wtxn, k, v)?;
-                    },
+                    }
                 }
             }
-        },
+        }
     }
 
     Ok(())
@@ -341,9 +332,7 @@ impl<'t, 'u, 'i, 'a> IndexDocuments<'t, 'u, 'i, 'a> {
 
         // Early return when there is no document to add
         if reader.buffer().is_empty() {
-            return Ok(DocumentAdditionResult {
-                nb_documents: 0,
-            })
+            return Ok(DocumentAdditionResult { nb_documents: 0 });
         }
 
         self.index.set_updated_at(self.wtxn, &Utc::now())?;
@@ -367,7 +356,9 @@ impl<'t, 'u, 'i, 'a> IndexDocuments<'t, 'u, 'i, 'a> {
         let output = match self.update_format {
             UpdateFormat::Csv => transform.output_from_csv(reader, &progress_callback)?,
             UpdateFormat::Json => transform.output_from_json(reader, &progress_callback)?,
-            UpdateFormat::JsonStream => transform.output_from_json_stream(reader, &progress_callback)?,
+            UpdateFormat::JsonStream => {
+                transform.output_from_json_stream(reader, &progress_callback)?
+            }
         };
 
         let nb_documents = output.documents_count;
@@ -380,7 +371,7 @@ impl<'t, 'u, 'i, 'a> IndexDocuments<'t, 'u, 'i, 'a> {
 
     pub fn execute_raw<F>(self, output: TransformOutput, progress_callback: F) -> Result<()>
     where
-        F: Fn(UpdateIndexingStep) + Sync
+        F: Fn(UpdateIndexingStep) + Sync,
     {
         let before_indexing = Instant::now();
 
@@ -457,7 +448,7 @@ impl<'t, 'u, 'i, 'a> IndexDocuments<'t, 'u, 'i, 'a> {
                 // settings if none have already been set.
                 backup_pool = rayon::ThreadPoolBuilder::new().build()?;
                 &backup_pool
-            },
+            }
         };
 
         let readers = pool.install(|| {
@@ -595,11 +586,8 @@ impl<'t, 'u, 'i, 'a> IndexDocuments<'t, 'u, 'i, 'a> {
 
         let mut documents_ids = self.index.documents_ids(self.wtxn)?;
         let contains_documents = !documents_ids.is_empty();
-        let write_method = if contains_documents {
-            WriteMethod::GetMergePut
-        } else {
-            WriteMethod::Append
-        };
+        let write_method =
+            if contains_documents { WriteMethod::GetMergePut } else { WriteMethod::Append };
 
         debug!("Writing using the write method: {:?}", write_method);
 
@@ -634,7 +622,7 @@ impl<'t, 'u, 'i, 'a> IndexDocuments<'t, 'u, 'i, 'a> {
             *self.index.docid_word_positions.as_polymorph(),
             docid_word_positions_readers,
             keep_first,
-            write_method
+            write_method,
         )?;
 
         database_count += 1;
@@ -649,7 +637,7 @@ impl<'t, 'u, 'i, 'a> IndexDocuments<'t, 'u, 'i, 'a> {
             *self.index.documents.as_polymorph(),
             documents_readers,
             keep_first,
-            write_method
+            write_method,
         )?;
 
         database_count += 1;
@@ -730,7 +718,7 @@ impl<'t, 'u, 'i, 'a> IndexDocuments<'t, 'u, 'i, 'a> {
                         fst_merge,
                         WriteMethod::GetMergePut,
                     )?;
-                },
+                }
                 DatabaseType::WordDocids => {
                     debug!("Writing the words docids into LMDB on disk...");
                     let db = *self.index.word_docids.as_polymorph();
@@ -741,7 +729,7 @@ impl<'t, 'u, 'i, 'a> IndexDocuments<'t, 'u, 'i, 'a> {
                         roaring_bitmap_merge,
                         write_method,
                     )?;
-                },
+                }
                 DatabaseType::FacetLevel0NumbersDocids => {
                     debug!("Writing the facet numbers docids into LMDB on disk...");
                     let db = *self.index.facet_id_f64_docids.as_polymorph();
@@ -752,7 +740,7 @@ impl<'t, 'u, 'i, 'a> IndexDocuments<'t, 'u, 'i, 'a> {
                         cbo_roaring_bitmap_merge,
                         write_method,
                     )?;
-                },
+                }
                 DatabaseType::FieldIdWordCountDocids => {
                     debug!("Writing the field id word count docids into LMDB on disk...");
                     let db = *self.index.field_id_word_count_docids.as_polymorph();
@@ -763,7 +751,7 @@ impl<'t, 'u, 'i, 'a> IndexDocuments<'t, 'u, 'i, 'a> {
                         cbo_roaring_bitmap_merge,
                         write_method,
                     )?;
-                },
+                }
                 DatabaseType::WordLevel0PositionDocids => {
                     debug!("Writing the word level 0 positions docids into LMDB on disk...");
                     let db = *self.index.word_level_position_docids.as_polymorph();
@@ -848,9 +836,10 @@ impl<'t, 'u, 'i, 'a> IndexDocuments<'t, 'u, 'i, 'a> {
 
 #[cfg(test)]
 mod tests {
-    use super::*;
     use heed::EnvOpenOptions;
 
+    use super::*;
+
     #[test]
     fn simple_document_replacement() {
         let path = tempfile::tempdir().unwrap();
@@ -1053,9 +1042,8 @@ mod tests {
         assert_eq!(count, 3);
 
         let docs = index.documents(&rtxn, vec![0, 1, 2]).unwrap();
-        let (kevin_id, _) = docs.iter().find(|(_, d)| {
-            d.get(0).unwrap() == br#""updated kevin""#
-        }).unwrap();
+        let (kevin_id, _) =
+            docs.iter().find(|(_, d)| d.get(0).unwrap() == br#""updated kevin""#).unwrap();
         let (id, doc) = docs[*kevin_id as usize];
         assert_eq!(id, *kevin_id);
 
diff --git a/milli/src/update/index_documents/store.rs b/milli/src/update/index_documents/store.rs
index 94ae12108..7318c5bd0 100644
--- a/milli/src/update/index_documents/store.rs
+++ b/milli/src/update/index_documents/store.rs
@@ -8,25 +8,29 @@ use std::{cmp, iter};
 
 use bstr::ByteSlice as _;
 use fst::Set;
-use grenad::{Reader, FileFuse, Writer, Sorter, CompressionType};
+use grenad::{CompressionType, FileFuse, Reader, Sorter, Writer};
 use heed::BytesEncode;
 use linked_hash_map::LinkedHashMap;
 use log::{debug, info};
-use meilisearch_tokenizer::{Analyzer, AnalyzerConfig, Token, TokenKind, token::SeparatorKind};
+use meilisearch_tokenizer::token::SeparatorKind;
+use meilisearch_tokenizer::{Analyzer, AnalyzerConfig, Token, TokenKind};
 use ordered_float::OrderedFloat;
 use roaring::RoaringBitmap;
 use serde_json::Value;
 use tempfile::tempfile;
 
+use super::merge_function::{
+    cbo_roaring_bitmap_merge, fst_merge, keep_first, roaring_bitmap_merge,
+};
+use super::{create_sorter, create_writer, writer_into_reader, MergeFn};
 use crate::error::{Error, InternalError, SerializationError};
-use crate::heed_codec::facet::{FacetValueStringCodec, FacetLevelValueF64Codec};
-use crate::heed_codec::facet::{FieldDocIdFacetStringCodec, FieldDocIdFacetF64Codec};
+use crate::heed_codec::facet::{
+    FacetLevelValueF64Codec, FacetValueStringCodec, FieldDocIdFacetF64Codec,
+    FieldDocIdFacetStringCodec,
+};
 use crate::heed_codec::{BoRoaringBitmapCodec, CboRoaringBitmapCodec};
 use crate::update::UpdateIndexingStep;
-use crate::{json_to_string, SmallVec32, Position, DocumentId, FieldId, Result};
-
-use super::{MergeFn, create_writer, create_sorter, writer_into_reader};
-use super::merge_function::{fst_merge, keep_first, roaring_bitmap_merge, cbo_roaring_bitmap_merge};
+use crate::{json_to_string, DocumentId, FieldId, Position, Result, SmallVec32};
 
 const LMDB_MAX_KEY_LENGTH: usize = 511;
 const ONE_KILOBYTE: usize = 1024 * 1024;
@@ -56,7 +60,8 @@ pub struct Store<'s, A> {
     word_docids: LinkedHashMap<SmallVec32<u8>, RoaringBitmap>,
     word_docids_limit: usize,
     field_id_word_count_docids: HashMap<(FieldId, u8), RoaringBitmap>,
-    words_pairs_proximities_docids: LinkedHashMap<(SmallVec32<u8>, SmallVec32<u8>, u8), RoaringBitmap>,
+    words_pairs_proximities_docids:
+        LinkedHashMap<(SmallVec32<u8>, SmallVec32<u8>, u8), RoaringBitmap>,
     words_pairs_proximities_docids_limit: usize,
     facet_field_number_docids: LinkedHashMap<(FieldId, OrderedFloat<f64>), RoaringBitmap>,
     facet_field_string_docids: LinkedHashMap<(FieldId, String), RoaringBitmap>,
@@ -93,8 +98,7 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
         chunk_compression_level: Option<u32>,
         chunk_fusing_shrink_size: Option<u64>,
         stop_words: Option<&'s Set<A>>,
-    ) -> Result<Self>
-    {
+    ) -> Result<Self> {
         // We divide the max memory by the number of sorter the Store have.
         let max_memory = max_memory.map(|mm| cmp::max(ONE_KILOBYTE, mm / 5));
         let linked_hash_map_size = linked_hash_map_size.unwrap_or(500);
@@ -172,12 +176,10 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
             Some(1024 * 1024 * 1024), // 1MB
         );
 
-        let documents_writer = tempfile().and_then(|f| {
-            create_writer(chunk_compression_type, chunk_compression_level, f)
-        })?;
-        let docid_word_positions_writer = tempfile().and_then(|f| {
-            create_writer(chunk_compression_type, chunk_compression_level, f)
-        })?;
+        let documents_writer = tempfile()
+            .and_then(|f| create_writer(chunk_compression_type, chunk_compression_level, f))?;
+        let docid_word_positions_writer = tempfile()
+            .and_then(|f| create_writer(chunk_compression_type, chunk_compression_level, f))?;
 
         let mut config = AnalyzerConfig::default();
         if let Some(stop_words) = stop_words {
@@ -224,7 +226,9 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
     fn insert_word_docid(&mut self, word: &str, id: DocumentId) -> Result<()> {
         // if get_refresh finds the element it is assured to be at the end of the linked hash map.
         match self.word_docids.get_refresh(word.as_bytes()) {
-            Some(old) => { old.insert(id); },
+            Some(old) => {
+                old.insert(id);
+            }
             None => {
                 let word_vec = SmallVec32::from(word.as_bytes());
                 // A newly inserted element is append at the end of the linked hash map.
@@ -246,15 +250,16 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
         field_id: FieldId,
         value: OrderedFloat<f64>,
         id: DocumentId,
-    ) -> Result<()>
-    {
+    ) -> Result<()> {
         let sorter = &mut self.field_id_docid_facet_numbers_sorter;
         Self::write_field_id_docid_facet_number_value(sorter, field_id, id, value)?;
 
         let key = (field_id, value);
         // if get_refresh finds the element it is assured to be at the end of the linked hash map.
         match self.facet_field_number_docids.get_refresh(&key) {
-            Some(old) => { old.insert(id); },
+            Some(old) => {
+                old.insert(id);
+            }
             None => {
                 // A newly inserted element is append at the end of the linked hash map.
                 self.facet_field_number_docids.insert(key, RoaringBitmap::from_iter(Some(id)));
@@ -279,15 +284,16 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
         field_id: FieldId,
         value: String,
         id: DocumentId,
-    ) -> Result<()>
-    {
+    ) -> Result<()> {
         let sorter = &mut self.field_id_docid_facet_strings_sorter;
         Self::write_field_id_docid_facet_string_value(sorter, field_id, id, &value)?;
 
         let key = (field_id, value);
         // if get_refresh finds the element it is assured to be at the end of the linked hash map.
         match self.facet_field_string_docids.get_refresh(&key) {
-            Some(old) => { old.insert(id); },
+            Some(old) => {
+                old.insert(id);
+            }
             None => {
                 // A newly inserted element is append at the end of the linked hash map.
                 self.facet_field_string_docids.insert(key, RoaringBitmap::from_iter(Some(id)));
@@ -309,10 +315,9 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
     // Save the documents ids under the words pairs proximities that it contains.
     fn insert_words_pairs_proximities_docids<'a>(
         &mut self,
-        words_pairs_proximities: impl IntoIterator<Item=((&'a str, &'a str), u8)>,
+        words_pairs_proximities: impl IntoIterator<Item = ((&'a str, &'a str), u8)>,
         id: DocumentId,
-    ) -> Result<()>
-    {
+    ) -> Result<()> {
         for ((w1, w2), prox) in words_pairs_proximities {
             let w1 = SmallVec32::from(w1.as_bytes());
             let w2 = SmallVec32::from(w2.as_bytes());
@@ -320,7 +325,9 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
             // if get_refresh finds the element it is assured
             // to be at the end of the linked hash map.
             match self.words_pairs_proximities_docids.get_refresh(&key) {
-                Some(old) => { old.insert(id); },
+                Some(old) => {
+                    old.insert(id);
+                }
                 None => {
                     // A newly inserted element is append at the end of the linked hash map.
                     let ids = RoaringBitmap::from_iter(Some(id));
@@ -337,7 +344,10 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
             // Removing front elements is equivalent to removing the LRUs.
             let iter = iter::from_fn(|| self.words_pairs_proximities_docids.pop_front());
             iter.take(overflow).for_each(|x| lrus.push(x));
-            Self::write_words_pairs_proximities(&mut self.words_pairs_proximities_docids_sorter, lrus)?;
+            Self::write_words_pairs_proximities(
+                &mut self.words_pairs_proximities_docids_sorter,
+                lrus,
+            )?;
         }
 
         Ok(())
@@ -350,8 +360,7 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
         facet_numbers_values: &mut HashMap<FieldId, Vec<f64>>,
         facet_strings_values: &mut HashMap<FieldId, Vec<String>>,
         record: &[u8],
-    ) -> Result<()>
-    {
+    ) -> Result<()> {
         // We compute the list of words pairs proximities (self-join) and write it directly to disk.
         let words_pair_proximities = compute_words_pair_proximities(&words_positions);
         self.insert_words_pairs_proximities_docids(words_pair_proximities, document_id)?;
@@ -362,8 +371,16 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
         }
 
         self.documents_writer.insert(document_id.to_be_bytes(), record)?;
-        Self::write_docid_word_positions(&mut self.docid_word_positions_writer, document_id, words_positions)?;
-        Self::write_word_position_docids(&mut self.word_level_position_docids_sorter, document_id, words_positions)?;
+        Self::write_docid_word_positions(
+            &mut self.docid_word_positions_writer,
+            document_id,
+            words_positions,
+        )?;
+        Self::write_word_position_docids(
+            &mut self.word_level_position_docids_sorter,
+            document_id,
+            words_positions,
+        )?;
 
         words_positions.clear();
 
@@ -387,7 +404,7 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
 
     fn write_words_pairs_proximities<E>(
         sorter: &mut Sorter<MergeFn<E>>,
-        iter: impl IntoIterator<Item=((SmallVec32<u8>, SmallVec32<u8>, u8), RoaringBitmap)>,
+        iter: impl IntoIterator<Item = ((SmallVec32<u8>, SmallVec32<u8>, u8), RoaringBitmap)>,
     ) -> Result<()>
     where
         Error: From<E>,
@@ -419,8 +436,7 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
         writer: &mut Writer<File>,
         id: DocumentId,
         words_positions: &HashMap<String, SmallVec32<Position>>,
-    ) -> Result<()>
-    {
+    ) -> Result<()> {
         // We prefix the words by the document id.
         let mut key = id.to_be_bytes().to_vec();
         let mut buffer = Vec::new();
@@ -484,12 +500,9 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
         Ok(())
     }
 
-    fn write_facet_field_string_docids<I, E>(
-        sorter: &mut Sorter<MergeFn<E>>,
-        iter: I,
-    ) -> Result<()>
+    fn write_facet_field_string_docids<I, E>(sorter: &mut Sorter<MergeFn<E>>, iter: I) -> Result<()>
     where
-        I: IntoIterator<Item=((FieldId, String), RoaringBitmap)>,
+        I: IntoIterator<Item = ((FieldId, String), RoaringBitmap)>,
         Error: From<E>,
     {
         let mut key_buffer = Vec::new();
@@ -510,12 +523,9 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
         Ok(())
     }
 
-    fn write_facet_field_number_docids<I, E>(
-        sorter: &mut Sorter<MergeFn<E>>,
-        iter: I,
-    ) -> Result<()>
+    fn write_facet_field_number_docids<I, E>(sorter: &mut Sorter<MergeFn<E>>, iter: I) -> Result<()>
     where
-        I: IntoIterator<Item=((FieldId, OrderedFloat<f64>), RoaringBitmap)>,
+        I: IntoIterator<Item = ((FieldId, OrderedFloat<f64>), RoaringBitmap)>,
         Error: From<E>,
     {
         let mut data_buffer = Vec::new();
@@ -579,7 +589,7 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
 
     fn write_word_docids<I, E>(sorter: &mut Sorter<MergeFn<E>>, iter: I) -> Result<()>
     where
-        I: IntoIterator<Item=(SmallVec32<u8>, RoaringBitmap)>,
+        I: IntoIterator<Item = (SmallVec32<u8>, RoaringBitmap)>,
         Error: From<E>,
     {
         let mut key = Vec::new();
@@ -611,7 +621,8 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
         log_every_n: Option<usize>,
         mut progress_callback: F,
     ) -> Result<Readers>
-    where F: FnMut(UpdateIndexingStep),
+    where
+        F: FnMut(UpdateIndexingStep),
     {
         debug!("{:?}: Indexing in a Store...", thread_index);
 
@@ -629,7 +640,11 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
             if count % num_threads == thread_index {
                 // This is a log routine that we do every `log_every_n` documents.
                 if thread_index == 0 && log_every_n.map_or(false, |len| count % len == 0) {
-                    info!("We have seen {} documents so far ({:.02?}).", format_count(count), before.elapsed());
+                    info!(
+                        "We have seen {} documents so far ({:.02?}).",
+                        format_count(count),
+                        before.elapsed()
+                    );
                     progress_callback(UpdateIndexingStep::IndexDocuments {
                         documents_seen: count,
                         total_documents: documents_count,
@@ -638,12 +653,20 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
                 }
 
                 for (attr, content) in document.iter() {
-                    if self.faceted_fields.contains(&attr) || self.searchable_fields.contains(&attr) {
-                        let value = serde_json::from_slice(content).map_err(InternalError::SerdeJson)?;
+                    if self.faceted_fields.contains(&attr) || self.searchable_fields.contains(&attr)
+                    {
+                        let value =
+                            serde_json::from_slice(content).map_err(InternalError::SerdeJson)?;
 
                         let (facet_numbers, facet_strings) = extract_facet_values(&value);
-                        facet_numbers_values.entry(attr).or_insert_with(Vec::new).extend(facet_numbers);
-                        facet_strings_values.entry(attr).or_insert_with(Vec::new).extend(facet_strings);
+                        facet_numbers_values
+                            .entry(attr)
+                            .or_insert_with(Vec::new)
+                            .extend(facet_numbers);
+                        facet_strings_values
+                            .entry(attr)
+                            .or_insert_with(Vec::new)
+                            .extend(facet_strings);
 
                         if self.searchable_fields.contains(&attr) {
                             let content = match json_to_string(&value) {
@@ -658,12 +681,18 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
                             for (pos, token) in tokens.take_while(|(pos, _)| *pos < MAX_POSITION) {
                                 last_pos = Some(pos);
                                 let position = (attr as usize * MAX_POSITION + pos) as u32;
-                                words_positions.entry(token.text().to_string()).or_insert_with(SmallVec32::new).push(position);
+                                words_positions
+                                    .entry(token.text().to_string())
+                                    .or_insert_with(SmallVec32::new)
+                                    .push(position);
                             }
 
                             if let Some(last_pos) = last_pos.filter(|p| *p <= 10) {
                                 let key = (attr, last_pos as u8 + 1);
-                                self.field_id_word_count_docids.entry(key).or_insert_with(RoaringBitmap::new).insert(document_id);
+                                self.field_id_word_count_docids
+                                    .entry(key)
+                                    .or_insert_with(RoaringBitmap::new)
+                                    .insert(document_id);
                             }
                         }
                     }
@@ -713,7 +742,8 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
             self.facet_field_string_docids,
         )?;
 
-        let mut word_docids_wtr = tempfile().and_then(|f| create_writer(comp_type, comp_level, f))?;
+        let mut word_docids_wtr =
+            tempfile().and_then(|f| create_writer(comp_type, comp_level, f))?;
         let mut builder = fst::SetBuilder::memory();
 
         let mut iter = self.word_docids_sorter.into_iter()?;
@@ -737,37 +767,55 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
         let mut main_wtr = tempfile().and_then(|f| create_writer(comp_type, comp_level, f))?;
         self.main_sorter.write_into(&mut main_wtr)?;
 
-        let mut words_pairs_proximities_docids_wtr = tempfile().and_then(|f| create_writer(comp_type, comp_level, f))?;
-        self.words_pairs_proximities_docids_sorter.write_into(&mut words_pairs_proximities_docids_wtr)?;
+        let mut words_pairs_proximities_docids_wtr =
+            tempfile().and_then(|f| create_writer(comp_type, comp_level, f))?;
+        self.words_pairs_proximities_docids_sorter
+            .write_into(&mut words_pairs_proximities_docids_wtr)?;
 
-        let mut word_level_position_docids_wtr = tempfile().and_then(|f| create_writer(comp_type, comp_level, f))?;
+        let mut word_level_position_docids_wtr =
+            tempfile().and_then(|f| create_writer(comp_type, comp_level, f))?;
         self.word_level_position_docids_sorter.write_into(&mut word_level_position_docids_wtr)?;
 
-        let mut field_id_word_count_docids_wtr = tempfile().and_then(|f| create_writer(comp_type, comp_level, f))?;
+        let mut field_id_word_count_docids_wtr =
+            tempfile().and_then(|f| create_writer(comp_type, comp_level, f))?;
         self.field_id_word_count_docids_sorter.write_into(&mut field_id_word_count_docids_wtr)?;
 
-        let mut facet_field_numbers_docids_wtr = tempfile().and_then(|f| create_writer(comp_type, comp_level, f))?;
+        let mut facet_field_numbers_docids_wtr =
+            tempfile().and_then(|f| create_writer(comp_type, comp_level, f))?;
         self.facet_field_numbers_docids_sorter.write_into(&mut facet_field_numbers_docids_wtr)?;
 
-        let mut facet_field_strings_docids_wtr = tempfile().and_then(|f| create_writer(comp_type, comp_level, f))?;
+        let mut facet_field_strings_docids_wtr =
+            tempfile().and_then(|f| create_writer(comp_type, comp_level, f))?;
         self.facet_field_strings_docids_sorter.write_into(&mut facet_field_strings_docids_wtr)?;
 
-        let mut field_id_docid_facet_numbers_wtr = tempfile().and_then(|f| create_writer(comp_type, comp_level, f))?;
-        self.field_id_docid_facet_numbers_sorter.write_into(&mut field_id_docid_facet_numbers_wtr)?;
+        let mut field_id_docid_facet_numbers_wtr =
+            tempfile().and_then(|f| create_writer(comp_type, comp_level, f))?;
+        self.field_id_docid_facet_numbers_sorter
+            .write_into(&mut field_id_docid_facet_numbers_wtr)?;
 
-        let mut field_id_docid_facet_strings_wtr = tempfile().and_then(|f| create_writer(comp_type, comp_level, f))?;
-        self.field_id_docid_facet_strings_sorter.write_into(&mut field_id_docid_facet_strings_wtr)?;
+        let mut field_id_docid_facet_strings_wtr =
+            tempfile().and_then(|f| create_writer(comp_type, comp_level, f))?;
+        self.field_id_docid_facet_strings_sorter
+            .write_into(&mut field_id_docid_facet_strings_wtr)?;
 
         let main = writer_into_reader(main_wtr, shrink_size)?;
         let word_docids = writer_into_reader(word_docids_wtr, shrink_size)?;
-        let words_pairs_proximities_docids = writer_into_reader(words_pairs_proximities_docids_wtr, shrink_size)?;
-        let word_level_position_docids = writer_into_reader(word_level_position_docids_wtr, shrink_size)?;
-        let field_id_word_count_docids = writer_into_reader(field_id_word_count_docids_wtr, shrink_size)?;
-        let facet_field_numbers_docids = writer_into_reader(facet_field_numbers_docids_wtr, shrink_size)?;
-        let facet_field_strings_docids = writer_into_reader(facet_field_strings_docids_wtr, shrink_size)?;
-        let field_id_docid_facet_numbers = writer_into_reader(field_id_docid_facet_numbers_wtr, shrink_size)?;
-        let field_id_docid_facet_strings = writer_into_reader(field_id_docid_facet_strings_wtr, shrink_size)?;
-        let docid_word_positions = writer_into_reader(self.docid_word_positions_writer, shrink_size)?;
+        let words_pairs_proximities_docids =
+            writer_into_reader(words_pairs_proximities_docids_wtr, shrink_size)?;
+        let word_level_position_docids =
+            writer_into_reader(word_level_position_docids_wtr, shrink_size)?;
+        let field_id_word_count_docids =
+            writer_into_reader(field_id_word_count_docids_wtr, shrink_size)?;
+        let facet_field_numbers_docids =
+            writer_into_reader(facet_field_numbers_docids_wtr, shrink_size)?;
+        let facet_field_strings_docids =
+            writer_into_reader(facet_field_strings_docids_wtr, shrink_size)?;
+        let field_id_docid_facet_numbers =
+            writer_into_reader(field_id_docid_facet_numbers_wtr, shrink_size)?;
+        let field_id_docid_facet_strings =
+            writer_into_reader(field_id_docid_facet_strings_wtr, shrink_size)?;
+        let docid_word_positions =
+            writer_into_reader(self.docid_word_positions_writer, shrink_size)?;
         let documents = writer_into_reader(self.documents_writer, shrink_size)?;
 
         Ok(Readers {
@@ -792,8 +840,7 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
 /// close to each other.
 fn compute_words_pair_proximities(
     word_positions: &HashMap<String, SmallVec32<Position>>,
-) -> HashMap<(&str, &str), u8>
-{
+) -> HashMap<(&str, &str), u8> {
     use itertools::Itertools;
 
     let mut words_pair_proximities = HashMap::new();
@@ -828,31 +875,34 @@ fn lmdb_key_valid_size(key: &[u8]) -> bool {
 /// take an iterator on tokens and compute their relative position depending on separator kinds
 /// if it's an `Hard` separator we add an additional relative proximity of 8 between words,
 /// else we keep the standart proximity of 1 between words.
-fn process_tokens<'a>(tokens: impl Iterator<Item = Token<'a>>) -> impl Iterator<Item = (usize, Token<'a>)> {
+fn process_tokens<'a>(
+    tokens: impl Iterator<Item = Token<'a>>,
+) -> impl Iterator<Item = (usize, Token<'a>)> {
     tokens
         .skip_while(|token| token.is_separator().is_some())
         .scan((0, None), |(offset, prev_kind), token| {
-                match token.kind {
-                    TokenKind::Word | TokenKind::StopWord | TokenKind::Unknown => {
-                        *offset += match *prev_kind {
-                            Some(TokenKind::Separator(SeparatorKind::Hard)) => 8,
-                            Some(_) => 1,
-                            None => 0,
-                        };
-                        *prev_kind = Some(token.kind)
-                    }
-                    TokenKind::Separator(SeparatorKind::Hard) => {
-                        *prev_kind = Some(token.kind);
-                    }
-                    TokenKind::Separator(SeparatorKind::Soft)
-                        if *prev_kind != Some(TokenKind::Separator(SeparatorKind::Hard)) => {
-                        *prev_kind = Some(token.kind);
-                    }
-                    _ => (),
+            match token.kind {
+                TokenKind::Word | TokenKind::StopWord | TokenKind::Unknown => {
+                    *offset += match *prev_kind {
+                        Some(TokenKind::Separator(SeparatorKind::Hard)) => 8,
+                        Some(_) => 1,
+                        None => 0,
+                    };
+                    *prev_kind = Some(token.kind)
                 }
+                TokenKind::Separator(SeparatorKind::Hard) => {
+                    *prev_kind = Some(token.kind);
+                }
+                TokenKind::Separator(SeparatorKind::Soft)
+                    if *prev_kind != Some(TokenKind::Separator(SeparatorKind::Hard)) =>
+                {
+                    *prev_kind = Some(token.kind);
+                }
+                _ => (),
+            }
             Some((*offset, token))
         })
-    .filter(|(_, t)| t.is_word())
+        .filter(|(_, t)| t.is_word())
 }
 
 fn extract_facet_values(value: &Value) -> (Vec<f64>, Vec<String>) {
@@ -865,18 +915,22 @@ fn extract_facet_values(value: &Value) -> (Vec<f64>, Vec<String>) {
         match value {
             Value::Null => (),
             Value::Bool(b) => output_strings.push(b.to_string()),
-            Value::Number(number) => if let Some(float) = number.as_f64() {
-                output_numbers.push(float);
-            },
+            Value::Number(number) => {
+                if let Some(float) = number.as_f64() {
+                    output_numbers.push(float);
+                }
+            }
             Value::String(string) => {
                 let string = string.trim().to_lowercase();
                 output_strings.push(string);
-            },
-            Value::Array(values) => if can_recurse {
-                for value in values {
-                    inner_extract_facet_values(value, false, output_numbers, output_strings);
+            }
+            Value::Array(values) => {
+                if can_recurse {
+                    for value in values {
+                        inner_extract_facet_values(value, false, output_numbers, output_strings);
+                    }
                 }
-            },
+            }
             Value::Object(_) => (),
         }
     }
diff --git a/milli/src/update/index_documents/transform.rs b/milli/src/update/index_documents/transform.rs
index 9e88559d0..756ff492e 100644
--- a/milli/src/update/index_documents/transform.rs
+++ b/milli/src/update/index_documents/transform.rs
@@ -10,14 +10,15 @@ use log::info;
 use roaring::RoaringBitmap;
 use serde_json::{Map, Value};
 
-use crate::error::{Error, UserError, InternalError};
-use crate::index::db_name;
-use crate::update::index_documents::merge_function::{merge_obkvs, keep_latest_obkv};
-use crate::update::{AvailableDocumentsIds, UpdateIndexingStep};
-use crate::{BEU32, MergeFn, FieldsIdsMap, ExternalDocumentsIds, FieldId, FieldsDistribution};
-use crate::{Index, Result};
 use super::merge_function::merge_two_obkvs;
-use super::{create_writer, create_sorter, IndexDocumentsMethod};
+use super::{create_sorter, create_writer, IndexDocumentsMethod};
+use crate::error::{Error, InternalError, UserError};
+use crate::index::db_name;
+use crate::update::index_documents::merge_function::{keep_latest_obkv, merge_obkvs};
+use crate::update::{AvailableDocumentsIds, UpdateIndexingStep};
+use crate::{
+    ExternalDocumentsIds, FieldId, FieldsDistribution, FieldsIdsMap, Index, MergeFn, Result, BEU32,
+};
 
 const DEFAULT_PRIMARY_KEY_NAME: &str = "id";
 
@@ -64,7 +65,11 @@ impl Transform<'_, '_> {
         self.output_from_generic_json(reader, false, progress_callback)
     }
 
-    pub fn output_from_json_stream<R, F>(self, reader: R, progress_callback: F) -> Result<TransformOutput>
+    pub fn output_from_json_stream<R, F>(
+        self,
+        reader: R,
+        progress_callback: F,
+    ) -> Result<TransformOutput>
     where
         R: Read,
         F: Fn(UpdateIndexingStep) + Sync,
@@ -86,14 +91,16 @@ impl Transform<'_, '_> {
         let external_documents_ids = self.index.external_documents_ids(self.rtxn).unwrap();
 
         // Deserialize the whole batch of documents in memory.
-        let mut documents: Peekable<Box<dyn Iterator<Item=serde_json::Result<Map<String, Value>>>>> = if is_stream {
+        let mut documents: Peekable<
+            Box<dyn Iterator<Item = serde_json::Result<Map<String, Value>>>>,
+        > = if is_stream {
             let iter = serde_json::Deserializer::from_reader(reader).into_iter();
-            let iter = Box::new(iter) as Box<dyn Iterator<Item=_>>;
+            let iter = Box::new(iter) as Box<dyn Iterator<Item = _>>;
             iter.peekable()
         } else {
             let vec: Vec<_> = serde_json::from_reader(reader).map_err(UserError::SerdeJson)?;
             let iter = vec.into_iter().map(Ok);
-            let iter = Box::new(iter) as Box<dyn Iterator<Item=_>>;
+            let iter = Box::new(iter) as Box<dyn Iterator<Item = _>>;
             iter.peekable()
         };
 
@@ -104,15 +111,16 @@ impl Transform<'_, '_> {
             Err(_) => {
                 let error = documents.next().unwrap().unwrap_err();
                 return Err(UserError::SerdeJson(error).into());
-            },
+            }
         };
 
-        let alternative_name = first.and_then(|doc| doc.keys().find(|f| is_primary_key(f)).cloned());
+        let alternative_name =
+            first.and_then(|doc| doc.keys().find(|f| is_primary_key(f)).cloned());
         let (primary_key_id, primary_key) = compute_primary_key_pair(
             self.index.primary_key(self.rtxn)?,
             &mut fields_ids_map,
             alternative_name,
-            self.autogenerate_docids
+            self.autogenerate_docids,
         )?;
 
         if documents.peek().is_none() {
@@ -173,9 +181,11 @@ impl Transform<'_, '_> {
                 Some(value) => match value {
                     Value::String(string) => Cow::Borrowed(string.as_str()),
                     Value::Number(number) => Cow::Owned(number.to_string()),
-                    content => return Err(UserError::InvalidDocumentId {
-                        document_id: content.clone(),
-                    }.into()),
+                    content => {
+                        return Err(
+                            UserError::InvalidDocumentId { document_id: content.clone() }.into()
+                        )
+                    }
                 },
                 None => {
                     if !self.autogenerate_docids {
@@ -183,7 +193,7 @@ impl Transform<'_, '_> {
                     }
                     let uuid = uuid::Uuid::new_v4().to_hyphenated().encode_lower(&mut uuid_buffer);
                     Cow::Borrowed(uuid)
-                },
+                }
             };
 
             // We iterate in the fields ids ordered.
@@ -194,7 +204,8 @@ impl Transform<'_, '_> {
                 // and this should be the document id we return the one we generated.
                 if let Some(value) = document.get(name) {
                     // We serialize the attribute values.
-                    serde_json::to_writer(&mut json_buffer, value).map_err(InternalError::SerdeJson)?;
+                    serde_json::to_writer(&mut json_buffer, value)
+                        .map_err(InternalError::SerdeJson)?;
                     writer.insert(field_id, &json_buffer)?;
                 }
 
@@ -202,7 +213,8 @@ impl Transform<'_, '_> {
                 if field_id == primary_key_id && validate_document_id(&external_id).is_none() {
                     return Err(UserError::InvalidDocumentId {
                         document_id: Value::from(external_id),
-                    }.into());
+                    }
+                    .into());
                 }
             }
 
@@ -248,9 +260,9 @@ impl Transform<'_, '_> {
         // Extract the position of the primary key in the current headers, None if not found.
         let primary_key_pos = match self.index.primary_key(self.rtxn)? {
             Some(primary_key) => {
-               // The primary key is known so we must find the position in the CSV headers.
-               headers.iter().position(|h| h == primary_key)
-            },
+                // The primary key is known so we must find the position in the CSV headers.
+                headers.iter().position(|h| h == primary_key)
+            }
             None => headers.iter().position(is_primary_key),
         };
 
@@ -261,7 +273,7 @@ impl Transform<'_, '_> {
             self.index.primary_key(self.rtxn)?,
             &mut fields_ids_map,
             alternative_name,
-            self.autogenerate_docids
+            self.autogenerate_docids,
         )?;
 
         // The primary key field is not present in the header, so we need to create it.
@@ -308,27 +320,30 @@ impl Transform<'_, '_> {
                     // We validate the document id [a-zA-Z0-9\-_].
                     match validate_document_id(&external_id) {
                         Some(valid) => valid,
-                        None => return Err(UserError::InvalidDocumentId {
-                            document_id: Value::from(external_id),
-                        }.into()),
+                        None => {
+                            return Err(UserError::InvalidDocumentId {
+                                document_id: Value::from(external_id),
+                            }
+                            .into())
+                        }
                     }
-                },
+                }
                 None => uuid::Uuid::new_v4().to_hyphenated().encode_lower(&mut uuid_buffer),
             };
 
             // When the primary_key_field_id is found in the fields ids list
             // we return the generated document id instead of the record field.
-            let iter = fields_ids.iter()
-                .map(|(fi, i)| {
-                    let field = if *fi == primary_key_id { external_id } else { &record[*i] };
-                    (fi, field)
-                });
+            let iter = fields_ids.iter().map(|(fi, i)| {
+                let field = if *fi == primary_key_id { external_id } else { &record[*i] };
+                (fi, field)
+            });
 
             // We retrieve the field id based on the fields ids map fields ids order.
             for (field_id, field) in iter {
                 // We serialize the attribute values as JSON strings.
                 json_buffer.clear();
-                serde_json::to_writer(&mut json_buffer, &field).map_err(InternalError::SerdeJson)?;
+                serde_json::to_writer(&mut json_buffer, &field)
+                    .map_err(InternalError::SerdeJson)?;
                 writer.insert(*field_id, &json_buffer)?;
             }
 
@@ -410,26 +425,27 @@ impl Transform<'_, '_> {
                         IndexDocumentsMethod::ReplaceDocuments => (docid, update_obkv),
                         IndexDocumentsMethod::UpdateDocuments => {
                             let key = BEU32::new(docid);
-                            let base_obkv = self.index.documents.get(&self.rtxn, &key)?
-                                .ok_or(InternalError::DatabaseMissingEntry {
+                            let base_obkv = self.index.documents.get(&self.rtxn, &key)?.ok_or(
+                                InternalError::DatabaseMissingEntry {
                                     db_name: db_name::DOCUMENTS,
                                     key: None,
-                                })?;
+                                },
+                            )?;
                             let update_obkv = obkv::KvReader::new(update_obkv);
                             merge_two_obkvs(base_obkv, update_obkv, &mut obkv_buffer);
                             (docid, obkv_buffer.as_slice())
                         }
                     }
-                },
+                }
                 None => {
                     // If this user id is new we add it to the external documents ids map
                     // for new ids and into the list of new documents.
-                    let new_docid = available_documents_ids.next()
-                        .ok_or(UserError::DocumentLimitReached)?;
+                    let new_docid =
+                        available_documents_ids.next().ok_or(UserError::DocumentLimitReached)?;
                     new_external_documents_ids_builder.insert(external_id, new_docid as u64)?;
                     new_documents_ids.insert(new_docid);
                     (new_docid, update_obkv)
-                },
+                }
             };
 
             // We insert the document under the documents ids map into the final file.
@@ -450,7 +466,8 @@ impl Transform<'_, '_> {
 
         // We create a final writer to write the new documents in order from the sorter.
         let file = tempfile::tempfile()?;
-        let mut writer = create_writer(self.chunk_compression_type, self.chunk_compression_level, file)?;
+        let mut writer =
+            create_writer(self.chunk_compression_type, self.chunk_compression_level, file)?;
 
         // Once we have written all the documents into the final sorter, we write the documents
         // into this writer, extract the file and reset the seek to be able to read it again.
@@ -485,8 +502,7 @@ impl Transform<'_, '_> {
         primary_key: String,
         old_fields_ids_map: FieldsIdsMap,
         new_fields_ids_map: FieldsIdsMap,
-    ) -> Result<TransformOutput>
-    {
+    ) -> Result<TransformOutput> {
         let fields_distribution = self.index.fields_distribution(self.rtxn)?;
         let external_documents_ids = self.index.external_documents_ids(self.rtxn)?;
         let documents_ids = self.index.documents_ids(self.rtxn)?;
@@ -494,7 +510,8 @@ impl Transform<'_, '_> {
 
         // We create a final writer to write the new documents in order from the sorter.
         let file = tempfile::tempfile()?;
-        let mut writer = create_writer(self.chunk_compression_type, self.chunk_compression_level, file)?;
+        let mut writer =
+            create_writer(self.chunk_compression_type, self.chunk_compression_level, file)?;
 
         let mut obkv_buffer = Vec::new();
         for result in self.index.documents.iter(self.rtxn)? {
@@ -561,20 +578,19 @@ fn compute_primary_key_pair(
                         return Err(UserError::MissingPrimaryKey.into());
                     }
                     DEFAULT_PRIMARY_KEY_NAME.to_string()
-                },
+                }
             };
             let id = fields_ids_map.insert(&name).ok_or(UserError::AttributeLimitReached)?;
             Ok((id, name))
-        },
+        }
     }
 }
 
 fn validate_document_id(document_id: &str) -> Option<&str> {
     let document_id = document_id.trim();
     Some(document_id).filter(|id| {
-        !id.is_empty() && id.chars().all(|c| {
-            matches!(c, 'a'..='z' | 'A'..='Z' | '0'..='9' | '-' | '_')
-        })
+        !id.is_empty()
+            && id.chars().all(|c| matches!(c, 'a'..='z' | 'A'..='Z' | '0'..='9' | '-' | '_'))
     })
 }
 
@@ -583,8 +599,7 @@ mod test {
     use super::*;
 
     mod compute_primary_key {
-        use super::compute_primary_key_pair;
-        use super::FieldsIdsMap;
+        use super::{compute_primary_key_pair, FieldsIdsMap};
 
         #[test]
         fn should_return_primary_key_if_is_some() {
@@ -594,7 +609,8 @@ mod test {
                 Some("toto"),
                 &mut fields_map,
                 Some("tata".to_string()),
-                false);
+                false,
+            );
             assert_eq!(result.unwrap(), (0u8, "toto".to_string()));
             assert_eq!(fields_map.len(), 1);
         }
@@ -602,11 +618,8 @@ mod test {
         #[test]
         fn should_return_alternative_if_primary_is_none() {
             let mut fields_map = FieldsIdsMap::new();
-            let result = compute_primary_key_pair(
-                None,
-                &mut fields_map,
-                Some("tata".to_string()),
-                false);
+            let result =
+                compute_primary_key_pair(None, &mut fields_map, Some("tata".to_string()), false);
             assert_eq!(result.unwrap(), (0u8, "tata".to_string()));
             assert_eq!(fields_map.len(), 1);
         }
@@ -614,23 +627,15 @@ mod test {
         #[test]
         fn should_return_default_if_both_are_none() {
             let mut fields_map = FieldsIdsMap::new();
-            let result = compute_primary_key_pair(
-                None,
-                &mut fields_map,
-                None,
-                true);
+            let result = compute_primary_key_pair(None, &mut fields_map, None, true);
             assert_eq!(result.unwrap(), (0u8, "id".to_string()));
             assert_eq!(fields_map.len(), 1);
         }
 
         #[test]
-        fn should_return_err_if_both_are_none_and_recompute_is_false(){
+        fn should_return_err_if_both_are_none_and_recompute_is_false() {
             let mut fields_map = FieldsIdsMap::new();
-            let result = compute_primary_key_pair(
-                None,
-                &mut fields_map,
-                None,
-                false);
+            let result = compute_primary_key_pair(None, &mut fields_map, None, false);
             assert!(result.is_err());
             assert_eq!(fields_map.len(), 0);
         }
diff --git a/milli/src/update/mod.rs b/milli/src/update/mod.rs
index 203937e2f..36ed7d8fa 100644
--- a/milli/src/update/mod.rs
+++ b/milli/src/update/mod.rs
@@ -2,7 +2,9 @@ pub use self::available_documents_ids::AvailableDocumentsIds;
 pub use self::clear_documents::ClearDocuments;
 pub use self::delete_documents::DeleteDocuments;
 pub use self::facets::Facets;
-pub use self::index_documents::{DocumentAdditionResult, IndexDocuments, IndexDocumentsMethod, UpdateFormat};
+pub use self::index_documents::{
+    DocumentAdditionResult, IndexDocuments, IndexDocumentsMethod, UpdateFormat,
+};
 pub use self::settings::{Setting, Settings};
 pub use self::update_builder::UpdateBuilder;
 pub use self::update_step::UpdateIndexingStep;
diff --git a/milli/src/update/settings.rs b/milli/src/update/settings.rs
index 8f4fe48c9..c6540b33a 100644
--- a/milli/src/update/settings.rs
+++ b/milli/src/update/settings.rs
@@ -34,17 +34,24 @@ impl<T> Setting<T> {
 }
 
 impl<T: Serialize> Serialize for Setting<T> {
-    fn serialize<S>(&self, serializer: S) -> StdResult<S::Ok, S::Error> where S: Serializer {
+    fn serialize<S>(&self, serializer: S) -> StdResult<S::Ok, S::Error>
+    where
+        S: Serializer,
+    {
         match self {
             Self::Set(value) => Some(value),
             // Usually not_set isn't serialized by setting skip_serializing_if field attribute
             Self::NotSet | Self::Reset => None,
-        }.serialize(serializer)
+        }
+        .serialize(serializer)
     }
 }
 
 impl<'de, T: Deserialize<'de>> Deserialize<'de> for Setting<T> {
-    fn deserialize<D>(deserializer: D) -> StdResult<Self, D::Error> where D: Deserializer<'de> {
+    fn deserialize<D>(deserializer: D) -> StdResult<Self, D::Error>
+    where
+        D: Deserializer<'de>,
+    {
         Deserialize::deserialize(deserializer).map(|x| match x {
             Some(x) => Self::Set(x),
             None => Self::Reset, // Reset is forced by sending null value
@@ -141,11 +148,8 @@ impl<'a, 't, 'u, 'i> Settings<'a, 't, 'u, 'i> {
     }
 
     pub fn set_stop_words(&mut self, stop_words: BTreeSet<String>) {
-        self.stop_words = if stop_words.is_empty() {
-            Setting::Reset
-        } else {
-            Setting::Set(stop_words)
-        }
+        self.stop_words =
+            if stop_words.is_empty() { Setting::Reset } else { Setting::Set(stop_words) }
     }
 
     pub fn reset_distinct_field(&mut self) {
@@ -161,11 +165,7 @@ impl<'a, 't, 'u, 'i> Settings<'a, 't, 'u, 'i> {
     }
 
     pub fn set_synonyms(&mut self, synonyms: HashMap<String, Vec<String>>) {
-        self.synonyms = if synonyms.is_empty() {
-            Setting::Reset
-        } else {
-            Setting::Set(synonyms)
-        }
+        self.synonyms = if synonyms.is_empty() { Setting::Reset } else { Setting::Set(synonyms) }
     }
 
     pub fn reset_primary_key(&mut self) {
@@ -178,7 +178,7 @@ impl<'a, 't, 'u, 'i> Settings<'a, 't, 'u, 'i> {
 
     fn reindex<F>(&mut self, cb: &F, old_fields_ids_map: FieldsIdsMap) -> Result<()>
     where
-        F: Fn(UpdateIndexingStep, u64) + Sync
+        F: Fn(UpdateIndexingStep, u64) + Sync,
     {
         let fields_ids_map = self.index.fields_ids_map(self.wtxn)?;
         let update_id = self.update_id;
@@ -203,7 +203,8 @@ impl<'a, 't, 'u, 'i> Settings<'a, 't, 'u, 'i> {
         };
 
         // There already has been a document addition, the primary key should be set by now.
-        let primary_key = self.index.primary_key(&self.wtxn)?.ok_or(UserError::MissingPrimaryKey)?;
+        let primary_key =
+            self.index.primary_key(&self.wtxn)?.ok_or(UserError::MissingPrimaryKey)?;
 
         // We remap the documents fields based on the new `FieldsIdsMap`.
         let output = transform.remap_index_documents(
@@ -236,21 +237,17 @@ impl<'a, 't, 'u, 'i> Settings<'a, 't, 'u, 'i> {
             Setting::Set(ref fields) => {
                 let mut fields_ids_map = self.index.fields_ids_map(self.wtxn)?;
                 // fields are deduplicated, only the first occurrence is taken into account
-                let names: Vec<_> = fields
-                    .iter()
-                    .unique()
-                    .map(String::as_str)
-                    .collect();
+                let names: Vec<_> = fields.iter().unique().map(String::as_str).collect();
 
                 for name in names.iter() {
-                    fields_ids_map
-                        .insert(name)
-                        .ok_or(UserError::AttributeLimitReached)?;
+                    fields_ids_map.insert(name).ok_or(UserError::AttributeLimitReached)?;
                 }
                 self.index.put_displayed_fields(self.wtxn, &names)?;
                 self.index.put_fields_ids_map(self.wtxn, &fields_ids_map)?;
             }
-            Setting::Reset => { self.index.delete_displayed_fields(self.wtxn)?; }
+            Setting::Reset => {
+                self.index.delete_displayed_fields(self.wtxn)?;
+            }
             Setting::NotSet => return Ok(false),
         }
         Ok(true)
@@ -260,14 +257,14 @@ impl<'a, 't, 'u, 'i> Settings<'a, 't, 'u, 'i> {
         match self.distinct_field {
             Setting::Set(ref attr) => {
                 let mut fields_ids_map = self.index.fields_ids_map(self.wtxn)?;
-                fields_ids_map
-                    .insert(attr)
-                    .ok_or(UserError::AttributeLimitReached)?;
+                fields_ids_map.insert(attr).ok_or(UserError::AttributeLimitReached)?;
 
                 self.index.put_distinct_field(self.wtxn, &attr)?;
                 self.index.put_fields_ids_map(self.wtxn, &fields_ids_map)?;
             }
-            Setting::Reset => { self.index.delete_distinct_field(self.wtxn)?; },
+            Setting::Reset => {
+                self.index.delete_distinct_field(self.wtxn)?;
+            }
             Setting::NotSet => return Ok(false),
         }
         Ok(true)
@@ -285,30 +282,24 @@ impl<'a, 't, 'u, 'i> Settings<'a, 't, 'u, 'i> {
 
                 let mut new_fields_ids_map = FieldsIdsMap::new();
                 // fields are deduplicated, only the first occurrence is taken into account
-                let names = fields
-                    .iter()
-                    .unique()
-                    .map(String::as_str)
-                    .collect::<Vec<_>>();
+                let names = fields.iter().unique().map(String::as_str).collect::<Vec<_>>();
 
                 // Add all the searchable attributes to the field map, and then add the
                 // remaining fields from the old field map to the new one
                 for name in names.iter() {
-                    new_fields_ids_map
-                        .insert(&name)
-                        .ok_or(UserError::AttributeLimitReached)?;
+                    new_fields_ids_map.insert(&name).ok_or(UserError::AttributeLimitReached)?;
                 }
 
                 for (_, name) in old_fields_ids_map.iter() {
-                    new_fields_ids_map
-                        .insert(&name)
-                        .ok_or(UserError::AttributeLimitReached)?;
+                    new_fields_ids_map.insert(&name).ok_or(UserError::AttributeLimitReached)?;
                 }
 
                 self.index.put_searchable_fields(self.wtxn, &names)?;
                 self.index.put_fields_ids_map(self.wtxn, &new_fields_ids_map)?;
             }
-            Setting::Reset => { self.index.delete_searchable_fields(self.wtxn)?; }
+            Setting::Reset => {
+                self.index.delete_searchable_fields(self.wtxn)?;
+            }
             Setting::NotSet => return Ok(false),
         }
         Ok(true)
@@ -323,7 +314,9 @@ impl<'a, 't, 'u, 'i> Settings<'a, 't, 'u, 'i> {
                 let fst = fst::Set::from_iter(stop_words)?;
 
                 // Does the new FST differ from the previous one?
-                if current.map_or(true, |current| current.as_fst().as_bytes() != fst.as_fst().as_bytes()) {
+                if current
+                    .map_or(true, |current| current.as_fst().as_bytes() != fst.as_fst().as_bytes())
+                {
                     // we want to re-create our FST.
                     self.index.put_stop_words(self.wtxn, &fst)?;
                     Ok(true)
@@ -343,9 +336,13 @@ impl<'a, 't, 'u, 'i> Settings<'a, 't, 'u, 'i> {
                     analyzer
                         .analyze(text)
                         .tokens()
-                        .filter_map(|token|
-                            if token.is_word() { Some(token.text().to_string()) } else { None }
-                        )
+                        .filter_map(|token| {
+                            if token.is_word() {
+                                Some(token.text().to_string())
+                            } else {
+                                None
+                            }
+                        })
                         .collect::<Vec<_>>()
                 }
 
@@ -360,25 +357,20 @@ impl<'a, 't, 'u, 'i> Settings<'a, 't, 'u, 'i> {
                 for (word, synonyms) in synonyms {
                     // Normalize both the word and associated synonyms.
                     let normalized_word = normalize(&analyzer, word);
-                    let normalized_synonyms = synonyms
-                        .iter()
-                        .map(|synonym| normalize(&analyzer, synonym));
+                    let normalized_synonyms =
+                        synonyms.iter().map(|synonym| normalize(&analyzer, synonym));
 
                     // Store the normalized synonyms under the normalized word,
                     // merging the possible duplicate words.
-                    let entry = new_synonyms
-                        .entry(normalized_word)
-                        .or_insert_with(Vec::new);
+                    let entry = new_synonyms.entry(normalized_word).or_insert_with(Vec::new);
                     entry.extend(normalized_synonyms);
                 }
 
                 // Make sure that we don't have duplicate synonyms.
-                new_synonyms
-                    .iter_mut()
-                    .for_each(|(_, synonyms)| {
-                        synonyms.sort_unstable();
-                        synonyms.dedup();
-                    });
+                new_synonyms.iter_mut().for_each(|(_, synonyms)| {
+                    synonyms.sort_unstable();
+                    synonyms.dedup();
+                });
 
                 let old_synonyms = self.index.synonyms(self.wtxn)?;
 
@@ -406,7 +398,9 @@ impl<'a, 't, 'u, 'i> Settings<'a, 't, 'u, 'i> {
                 self.index.put_filterable_fields(self.wtxn, &new_facets)?;
                 self.index.put_fields_ids_map(self.wtxn, &fields_ids_map)?;
             }
-            Setting::Reset => { self.index.delete_filterable_fields(self.wtxn)?; }
+            Setting::Reset => {
+                self.index.delete_filterable_fields(self.wtxn)?;
+            }
             Setting::NotSet => (),
         }
         Ok(())
@@ -427,7 +421,9 @@ impl<'a, 't, 'u, 'i> Settings<'a, 't, 'u, 'i> {
                 self.index.put_criteria(self.wtxn, &new_criteria)?;
                 self.index.put_fields_ids_map(self.wtxn, &fields_ids_map)?;
             }
-            Setting::Reset => { self.index.delete_criteria(self.wtxn)?; }
+            Setting::Reset => {
+                self.index.delete_criteria(self.wtxn)?;
+            }
             Setting::NotSet => (),
         }
         Ok(())
@@ -445,7 +441,7 @@ impl<'a, 't, 'u, 'i> Settings<'a, 't, 'u, 'i> {
                 } else {
                     Err(UserError::PrimaryKeyCannotBeChanged.into())
                 }
-            },
+            }
             Setting::Reset => {
                 if self.index.number_of_documents(&self.wtxn)? == 0 {
                     self.index.delete_primary_key(self.wtxn)?;
@@ -453,14 +449,14 @@ impl<'a, 't, 'u, 'i> Settings<'a, 't, 'u, 'i> {
                 } else {
                     Err(UserError::PrimaryKeyCannotBeReset.into())
                 }
-            },
+            }
             Setting::NotSet => Ok(()),
         }
     }
 
     pub fn execute<F>(mut self, progress_callback: F) -> Result<()>
-        where
-            F: Fn(UpdateIndexingStep, u64) + Sync
+    where
+        F: Fn(UpdateIndexingStep, u64) + Sync,
     {
         self.index.set_updated_at(self.wtxn, &Utc::now())?;
 
@@ -493,17 +489,16 @@ impl<'a, 't, 'u, 'i> Settings<'a, 't, 'u, 'i> {
 
 #[cfg(test)]
 mod tests {
-    use heed::EnvOpenOptions;
-    use heed::types::ByteSlice;
-    use maplit::{btreeset, hashmap, hashset};
     use big_s::S;
+    use heed::types::ByteSlice;
+    use heed::EnvOpenOptions;
+    use maplit::{btreeset, hashmap, hashset};
 
+    use super::*;
     use crate::error::Error;
     use crate::update::{IndexDocuments, UpdateFormat};
     use crate::{Criterion, FilterCondition, SearchResult};
 
-    use super::*;
-
     #[test]
     fn set_and_reset_searchable_fields() {
         let path = tempfile::tempdir().unwrap();
@@ -674,7 +669,7 @@ mod tests {
         // Set the filterable fields to be the age.
         let mut wtxn = index.write_txn().unwrap();
         let mut builder = Settings::new(&mut wtxn, &index, 0);
-        builder.set_filterable_fields(hashset!{ S("age") });
+        builder.set_filterable_fields(hashset! { S("age") });
         builder.execute(|_, _| ()).unwrap();
 
         // Then index some documents.
@@ -692,12 +687,15 @@ mod tests {
         // Check that the displayed fields are correctly set.
         let rtxn = index.read_txn().unwrap();
         let fields_ids = index.filterable_fields(&rtxn).unwrap();
-        assert_eq!(fields_ids, hashset!{ S("age") });
+        assert_eq!(fields_ids, hashset! { S("age") });
         // Only count the field_id 0 and level 0 facet values.
         // TODO we must support typed CSVs for numbers to be understood.
-        let count = index.facet_id_f64_docids
+        let count = index
+            .facet_id_f64_docids
             .remap_key_type::<ByteSlice>()
-            .prefix_iter(&rtxn, &[0, 0]).unwrap().count();
+            .prefix_iter(&rtxn, &[0, 0])
+            .unwrap()
+            .count();
         assert_eq!(count, 3);
         drop(rtxn);
 
@@ -718,9 +716,12 @@ mod tests {
         let rtxn = index.read_txn().unwrap();
         // Only count the field_id 0 and level 0 facet values.
         // TODO we must support typed CSVs for numbers to be understood.
-        let count = index.facet_id_f64_docids
+        let count = index
+            .facet_id_f64_docids
             .remap_key_type::<ByteSlice>()
-            .prefix_iter(&rtxn, &[0, 0]).unwrap().count();
+            .prefix_iter(&rtxn, &[0, 0])
+            .unwrap()
+            .count();
         assert_eq!(count, 4);
     }
 
@@ -969,7 +970,7 @@ mod tests {
         let mut wtxn = index.write_txn().unwrap();
         let mut builder = Settings::new(&mut wtxn, &index, 0);
         builder.set_displayed_fields(vec!["hello".to_string()]);
-        builder.set_filterable_fields(hashset!{ S("age"), S("toto") });
+        builder.set_filterable_fields(hashset! { S("age"), S("toto") });
         builder.set_criteria(vec!["asc(toto)".to_string()]);
         builder.execute(|_, _| ()).unwrap();
         wtxn.commit().unwrap();
diff --git a/milli/src/update/update_builder.rs b/milli/src/update/update_builder.rs
index 1d0e776b1..2816ebca0 100644
--- a/milli/src/update/update_builder.rs
+++ b/milli/src/update/update_builder.rs
@@ -1,8 +1,8 @@
 use grenad::CompressionType;
 use rayon::ThreadPool;
 
+use super::{ClearDocuments, DeleteDocuments, Facets, IndexDocuments, Settings};
 use crate::{Index, Result};
-use super::{ClearDocuments, DeleteDocuments, IndexDocuments, Settings, Facets};
 
 pub struct UpdateBuilder<'a> {
     pub(crate) log_every_n: Option<usize>,
@@ -67,8 +67,7 @@ impl<'a> UpdateBuilder<'a> {
         self,
         wtxn: &'t mut heed::RwTxn<'i, 'u>,
         index: &'i Index,
-    ) -> ClearDocuments<'t, 'u, 'i>
-    {
+    ) -> ClearDocuments<'t, 'u, 'i> {
         ClearDocuments::new(wtxn, index, self.update_id)
     }
 
@@ -76,8 +75,7 @@ impl<'a> UpdateBuilder<'a> {
         self,
         wtxn: &'t mut heed::RwTxn<'i, 'u>,
         index: &'i Index,
-    ) -> Result<DeleteDocuments<'t, 'u, 'i>>
-    {
+    ) -> Result<DeleteDocuments<'t, 'u, 'i>> {
         DeleteDocuments::new(wtxn, index, self.update_id)
     }
 
@@ -85,8 +83,7 @@ impl<'a> UpdateBuilder<'a> {
         self,
         wtxn: &'t mut heed::RwTxn<'i, 'u>,
         index: &'i Index,
-    ) -> IndexDocuments<'t, 'u, 'i, 'a>
-    {
+    ) -> IndexDocuments<'t, 'u, 'i, 'a> {
         let mut builder = IndexDocuments::new(wtxn, index, self.update_id);
 
         builder.log_every_n = self.log_every_n;
@@ -105,8 +102,7 @@ impl<'a> UpdateBuilder<'a> {
         self,
         wtxn: &'t mut heed::RwTxn<'i, 'u>,
         index: &'i Index,
-    ) -> Settings<'a, 't, 'u, 'i>
-    {
+    ) -> Settings<'a, 't, 'u, 'i> {
         let mut builder = Settings::new(wtxn, index, self.update_id);
 
         builder.log_every_n = self.log_every_n;
@@ -125,8 +121,7 @@ impl<'a> UpdateBuilder<'a> {
         self,
         wtxn: &'t mut heed::RwTxn<'i, 'u>,
         index: &'i Index,
-    ) -> Facets<'t, 'u, 'i>
-    {
+    ) -> Facets<'t, 'u, 'i> {
         let mut builder = Facets::new(wtxn, index, self.update_id);
 
         builder.chunk_compression_type = self.chunk_compression_type;
diff --git a/milli/src/update/word_prefix_docids.rs b/milli/src/update/word_prefix_docids.rs
index a2197b28c..ffc359719 100644
--- a/milli/src/update/word_prefix_docids.rs
+++ b/milli/src/update/word_prefix_docids.rs
@@ -1,15 +1,13 @@
 use std::str;
 
-use crate::Index;
 use fst::Streamer;
 use grenad::CompressionType;
 use heed::types::ByteSlice;
 
-use crate::Result;
-use crate::update::index_documents::WriteMethod;
 use crate::update::index_documents::{
-    create_sorter, roaring_bitmap_merge, sorter_into_lmdb_database,
+    create_sorter, roaring_bitmap_merge, sorter_into_lmdb_database, WriteMethod,
 };
+use crate::{Index, Result};
 
 pub struct WordPrefixDocids<'t, 'u, 'i> {
     wtxn: &'t mut heed::RwTxn<'i, 'u>,
@@ -22,7 +20,10 @@ pub struct WordPrefixDocids<'t, 'u, 'i> {
 }
 
 impl<'t, 'u, 'i> WordPrefixDocids<'t, 'u, 'i> {
-    pub fn new(wtxn: &'t mut heed::RwTxn<'i, 'u>, index: &'i Index) -> WordPrefixDocids<'t, 'u, 'i> {
+    pub fn new(
+        wtxn: &'t mut heed::RwTxn<'i, 'u>,
+        index: &'i Index,
+    ) -> WordPrefixDocids<'t, 'u, 'i> {
         WordPrefixDocids {
             wtxn,
             index,
diff --git a/milli/src/update/word_prefix_pair_proximity_docids.rs b/milli/src/update/word_prefix_pair_proximity_docids.rs
index 9019b26e5..9b876321e 100644
--- a/milli/src/update/word_prefix_pair_proximity_docids.rs
+++ b/milli/src/update/word_prefix_pair_proximity_docids.rs
@@ -1,18 +1,17 @@
 use std::str;
 
 use fst::automaton::{Automaton, Str};
-use fst::{Streamer, IntoStreamer};
+use fst::{IntoStreamer, Streamer};
 use grenad::CompressionType;
-use heed::BytesEncode;
 use heed::types::ByteSlice;
+use heed::BytesEncode;
 use log::debug;
 
-use crate::{Index, Result};
 use crate::heed_codec::StrStrU8Codec;
 use crate::update::index_documents::{
-    WriteMethod, create_sorter, sorter_into_lmdb_database,
-    cbo_roaring_bitmap_merge,
+    cbo_roaring_bitmap_merge, create_sorter, sorter_into_lmdb_database, WriteMethod,
 };
+use crate::{Index, Result};
 
 pub struct WordPrefixPairProximityDocids<'t, 'u, 'i> {
     wtxn: &'t mut heed::RwTxn<'i, 'u>,
@@ -28,8 +27,7 @@ impl<'t, 'u, 'i> WordPrefixPairProximityDocids<'t, 'u, 'i> {
     pub fn new(
         wtxn: &'t mut heed::RwTxn<'i, 'u>,
         index: &'i Index,
-    ) -> WordPrefixPairProximityDocids<'t, 'u, 'i>
-    {
+    ) -> WordPrefixPairProximityDocids<'t, 'u, 'i> {
         WordPrefixPairProximityDocids {
             wtxn,
             index,
diff --git a/milli/src/update/words_level_positions.rs b/milli/src/update/words_level_positions.rs
index e2e3f7b4c..d43cd19b8 100644
--- a/milli/src/update/words_level_positions.rs
+++ b/milli/src/update/words_level_positions.rs
@@ -1,25 +1,23 @@
-use std::{cmp, str};
 use std::convert::TryFrom;
 use std::fs::File;
 use std::num::NonZeroU32;
+use std::{cmp, str};
 
 use fst::automaton::{self, Automaton};
-use fst::{Streamer, IntoStreamer};
-use grenad::{CompressionType, Reader, Writer, FileFuse};
+use fst::{IntoStreamer, Streamer};
+use grenad::{CompressionType, FileFuse, Reader, Writer};
 use heed::types::{ByteSlice, DecodeIgnore, Str};
 use heed::{BytesEncode, Error};
 use log::debug;
 use roaring::RoaringBitmap;
 
 use crate::error::InternalError;
-use crate::heed_codec::{StrLevelPositionCodec, CboRoaringBitmapCodec};
-use crate::Result;
-use crate::update::index_documents::WriteMethod;
+use crate::heed_codec::{CboRoaringBitmapCodec, StrLevelPositionCodec};
 use crate::update::index_documents::{
-    create_writer, create_sorter, writer_into_reader, write_into_lmdb_database,
-    cbo_roaring_bitmap_merge, sorter_into_lmdb_database
+    cbo_roaring_bitmap_merge, create_sorter, create_writer, sorter_into_lmdb_database,
+    write_into_lmdb_database, writer_into_reader, WriteMethod,
 };
-use crate::{Index, TreeLevel};
+use crate::{Index, Result, TreeLevel};
 
 pub struct WordsLevelPositions<'t, 'u, 'i> {
     wtxn: &'t mut heed::RwTxn<'i, 'u>,
@@ -34,7 +32,10 @@ pub struct WordsLevelPositions<'t, 'u, 'i> {
 }
 
 impl<'t, 'u, 'i> WordsLevelPositions<'t, 'u, 'i> {
-    pub fn new(wtxn: &'t mut heed::RwTxn<'i, 'u>, index: &'i Index) -> WordsLevelPositions<'t, 'u, 'i> {
+    pub fn new(
+        wtxn: &'t mut heed::RwTxn<'i, 'u>,
+        index: &'i Index,
+    ) -> WordsLevelPositions<'t, 'u, 'i> {
         WordsLevelPositions {
             wtxn,
             index,
@@ -144,7 +145,9 @@ impl<'t, 'u, 'i> WordsLevelPositions<'t, 'u, 'i> {
             self.wtxn,
             *self.index.word_prefix_level_position_docids.as_polymorph(),
             entries,
-            |_, _| Err(InternalError::IndexingMergingKeys { process: "word prefix level position" }),
+            |_, _| {
+                Err(InternalError::IndexingMergingKeys { process: "word prefix level position" })
+            },
             WriteMethod::Append,
         )?;
 
@@ -176,13 +179,11 @@ fn compute_positions_levels(
     shrink_size: Option<u64>,
     level_group_size: NonZeroU32,
     min_level_size: NonZeroU32,
-) -> Result<Reader<FileFuse>>
-{
+) -> Result<Reader<FileFuse>> {
     // It is forbidden to keep a cursor and write in a database at the same time with LMDB
     // therefore we write the facet levels entries into a grenad file before transfering them.
-    let mut writer = tempfile::tempfile().and_then(|file| {
-        create_writer(compression_type, compression_level, file)
-    })?;
+    let mut writer = tempfile::tempfile()
+        .and_then(|file| create_writer(compression_type, compression_level, file))?;
 
     for result in words_db.iter(rtxn)? {
         let (word, ()) = result?;
@@ -193,7 +194,8 @@ fn compute_positions_levels(
             left..=right
         };
 
-        let first_level_size = words_positions_db.remap_data_type::<DecodeIgnore>()
+        let first_level_size = words_positions_db
+            .remap_data_type::<DecodeIgnore>()
             .range(rtxn, &level_0_range)?
             .fold(Ok(0u32), |count, result| result.and(count).map(|c| c + 1))?;
 
@@ -253,8 +255,7 @@ fn write_level_entry(
     left: u32,
     right: u32,
     ids: &RoaringBitmap,
-) -> Result<()>
-{
+) -> Result<()> {
     let key = (word, level, left, right);
     let key = StrLevelPositionCodec::bytes_encode(&key).ok_or(Error::Encoding)?;
     let data = CboRoaringBitmapCodec::bytes_encode(&ids).ok_or(Error::Encoding)?;
diff --git a/milli/src/update/words_prefixes_fst.rs b/milli/src/update/words_prefixes_fst.rs
index d1aa267b8..f35dea10d 100644
--- a/milli/src/update/words_prefixes_fst.rs
+++ b/milli/src/update/words_prefixes_fst.rs
@@ -2,7 +2,8 @@ use std::iter::FromIterator;
 use std::str;
 
 use fst::Streamer;
-use crate::{Index, SmallString32, Result};
+
+use crate::{Index, Result, SmallString32};
 
 pub struct WordsPrefixesFst<'t, 'u, 'i> {
     wtxn: &'t mut heed::RwTxn<'i, 'u>,
@@ -17,8 +18,7 @@ impl<'t, 'u, 'i> WordsPrefixesFst<'t, 'u, 'i> {
         wtxn: &'t mut heed::RwTxn<'i, 'u>,
         index: &'i Index,
         update_id: u64,
-    ) -> WordsPrefixesFst<'t, 'u, 'i>
-    {
+    ) -> WordsPrefixesFst<'t, 'u, 'i> {
         WordsPrefixesFst {
             wtxn,
             index,
@@ -55,7 +55,6 @@ impl<'t, 'u, 'i> WordsPrefixesFst<'t, 'u, 'i> {
 
         let mut prefix_fsts = Vec::with_capacity(self.max_prefix_length);
         for n in 1..=self.max_prefix_length {
-
             let mut current_prefix = SmallString32::new();
             let mut current_prefix_count = 0;
             let mut builder = fst::SetBuilder::memory();
diff --git a/milli/tests/search/mod.rs b/milli/tests/search/mod.rs
index 8c63e5e08..7842b6c13 100644
--- a/milli/tests/search/mod.rs
+++ b/milli/tests/search/mod.rs
@@ -1,9 +1,8 @@
-use milli::{Criterion, Index, DocumentId};
-use milli::update::{IndexDocuments, UpdateFormat, Settings};
-
 use big_s::S;
 use heed::EnvOpenOptions;
 use maplit::{hashmap, hashset};
+use milli::update::{IndexDocuments, Settings, UpdateFormat};
+use milli::{Criterion, DocumentId, Index};
 use serde::Deserialize;
 use slice_group_by::GroupBy;
 
@@ -11,7 +10,8 @@ mod query_criteria;
 
 pub const TEST_QUERY: &'static str = "hello world america";
 
-pub const EXTERNAL_DOCUMENTS_IDS: &[&str; 17] = &["A", "B", "C", "D", "E", "F", "G", "H", "I", "J", "K", "L", "M", "N", "O", "P", "Q"];
+pub const EXTERNAL_DOCUMENTS_IDS: &[&str; 17] =
+    &["A", "B", "C", "D", "E", "F", "G", "H", "I", "J", "K", "L", "M", "N", "O", "P", "Q"];
 
 pub const CONTENT: &str = include_str!("../assets/test_set.ndjson");
 
@@ -27,16 +27,16 @@ pub fn setup_search_index_with_criteria(criteria: &[Criterion]) -> Index {
 
     let criteria = criteria.iter().map(|c| c.to_string()).collect();
     builder.set_criteria(criteria);
-    builder.set_filterable_fields(hashset!{
+    builder.set_filterable_fields(hashset! {
         S("tag"),
         S("asc_desc_rank"),
     });
-    builder.set_synonyms(hashmap!{
+    builder.set_synonyms(hashmap! {
         S("hello") => vec![S("good morning")],
         S("world") => vec![S("earth")],
         S("america") => vec![S("the united states")],
     });
-    builder.set_searchable_fields(vec![S("title"),S("description")]);
+    builder.set_searchable_fields(vec![S("title"), S("description")]);
     builder.execute(|_, _| ()).unwrap();
 
     // index documents
@@ -53,12 +53,18 @@ pub fn setup_search_index_with_criteria(criteria: &[Criterion]) -> Index {
 pub fn internal_to_external_ids(index: &Index, internal_ids: &[DocumentId]) -> Vec<String> {
     let mut rtxn = index.read_txn().unwrap();
     let docid_map = index.external_documents_ids(&mut rtxn).unwrap();
-    let docid_map: std::collections::HashMap<_, _> = EXTERNAL_DOCUMENTS_IDS.iter().map(|id| (docid_map.get(id).unwrap(), id)).collect();
+    let docid_map: std::collections::HashMap<_, _> =
+        EXTERNAL_DOCUMENTS_IDS.iter().map(|id| (docid_map.get(id).unwrap(), id)).collect();
     internal_ids.iter().map(|id| docid_map.get(id).unwrap().to_string()).collect()
 }
 
-pub fn expected_order(criteria: &[Criterion], authorize_typo: bool, optional_words: bool) -> Vec<TestDocument> {
-    let dataset = serde_json::Deserializer::from_str(CONTENT).into_iter().map(|r| r.unwrap()).collect();
+pub fn expected_order(
+    criteria: &[Criterion],
+    authorize_typo: bool,
+    optional_words: bool,
+) -> Vec<TestDocument> {
+    let dataset =
+        serde_json::Deserializer::from_str(CONTENT).into_iter().map(|r| r.unwrap()).collect();
     let mut groups: Vec<Vec<TestDocument>> = vec![dataset];
 
     for criterion in criteria {
@@ -67,32 +73,36 @@ pub fn expected_order(criteria: &[Criterion], authorize_typo: bool, optional_wor
             match criterion {
                 Criterion::Attribute => {
                     group.sort_by_key(|d| d.attribute_rank);
-                    new_groups.extend(group.linear_group_by_key(|d| d.attribute_rank).map(Vec::from));
-                },
+                    new_groups
+                        .extend(group.linear_group_by_key(|d| d.attribute_rank).map(Vec::from));
+                }
                 Criterion::Exactness => {
                     group.sort_by_key(|d| d.exact_rank);
                     new_groups.extend(group.linear_group_by_key(|d| d.exact_rank).map(Vec::from));
-                },
+                }
                 Criterion::Proximity => {
                     group.sort_by_key(|d| d.proximity_rank);
-                    new_groups.extend(group.linear_group_by_key(|d| d.proximity_rank).map(Vec::from));
-                },
+                    new_groups
+                        .extend(group.linear_group_by_key(|d| d.proximity_rank).map(Vec::from));
+                }
                 Criterion::Typo => {
                     group.sort_by_key(|d| d.typo_rank);
                     new_groups.extend(group.linear_group_by_key(|d| d.typo_rank).map(Vec::from));
-                },
+                }
                 Criterion::Words => {
                     group.sort_by_key(|d| d.word_rank);
                     new_groups.extend(group.linear_group_by_key(|d| d.word_rank).map(Vec::from));
-                },
+                }
                 Criterion::Asc(field_name) if field_name == "asc_desc_rank" => {
                     group.sort_by_key(|d| d.asc_desc_rank);
-                    new_groups.extend(group.linear_group_by_key(|d| d.asc_desc_rank).map(Vec::from));
-                },
-                Criterion::Desc(field_name)  if field_name == "asc_desc_rank" => {
+                    new_groups
+                        .extend(group.linear_group_by_key(|d| d.asc_desc_rank).map(Vec::from));
+                }
+                Criterion::Desc(field_name) if field_name == "asc_desc_rank" => {
                     group.sort_by_key(|d| std::cmp::Reverse(d.asc_desc_rank));
-                    new_groups.extend(group.linear_group_by_key(|d| d.asc_desc_rank).map(Vec::from));
-                },
+                    new_groups
+                        .extend(group.linear_group_by_key(|d| d.asc_desc_rank).map(Vec::from));
+                }
                 Criterion::Asc(_) | Criterion::Desc(_) => new_groups.push(group.clone()),
             }
         }
diff --git a/milli/tests/search/query_criteria.rs b/milli/tests/search/query_criteria.rs
index 2b9c5ae5e..19173bc72 100644
--- a/milli/tests/search/query_criteria.rs
+++ b/milli/tests/search/query_criteria.rs
@@ -1,9 +1,9 @@
 use big_s::S;
 use milli::update::Settings;
-use milli::{Search, SearchResult, Criterion};
+use milli::{Criterion, Search, SearchResult};
+use Criterion::*;
 
 use crate::search::{self, EXTERNAL_DOCUMENTS_IDS};
-use Criterion::*;
 
 const ALLOW_TYPOS: bool = true;
 const DISALLOW_TYPOS: bool = false;
@@ -35,29 +35,54 @@ macro_rules! test_criterion {
     }
 }
 
+#[rustfmt::skip]
 test_criterion!(none_allow_typo,                     ALLOW_OPTIONAL_WORDS,      ALLOW_TYPOS);
+#[rustfmt::skip]
 test_criterion!(none_disallow_typo,                  DISALLOW_OPTIONAL_WORDS,   DISALLOW_TYPOS);
+#[rustfmt::skip]
 test_criterion!(words_allow_typo,                    ALLOW_OPTIONAL_WORDS,      ALLOW_TYPOS,    Words);
+#[rustfmt::skip]
 test_criterion!(attribute_allow_typo,                DISALLOW_OPTIONAL_WORDS,   ALLOW_TYPOS,    Attribute);
+#[rustfmt::skip]
 test_criterion!(attribute_disallow_typo,             DISALLOW_OPTIONAL_WORDS,   DISALLOW_TYPOS, Attribute);
+#[rustfmt::skip]
 test_criterion!(exactness_allow_typo,                DISALLOW_OPTIONAL_WORDS,   ALLOW_TYPOS,    Exactness);
+#[rustfmt::skip]
 test_criterion!(exactness_disallow_typo,             DISALLOW_OPTIONAL_WORDS,   DISALLOW_TYPOS, Exactness);
+#[rustfmt::skip]
 test_criterion!(proximity_allow_typo,                DISALLOW_OPTIONAL_WORDS,   ALLOW_TYPOS,    Proximity);
+#[rustfmt::skip]
 test_criterion!(proximity_disallow_typo,             DISALLOW_OPTIONAL_WORDS,   DISALLOW_TYPOS, Proximity);
+#[rustfmt::skip]
 test_criterion!(asc_allow_typo,                      DISALLOW_OPTIONAL_WORDS,   ALLOW_TYPOS,    Asc(S("asc_desc_rank")));
+#[rustfmt::skip]
 test_criterion!(asc_disallow_typo,                   DISALLOW_OPTIONAL_WORDS,   DISALLOW_TYPOS, Asc(S("asc_desc_rank")));
+#[rustfmt::skip]
 test_criterion!(desc_allow_typo,                     DISALLOW_OPTIONAL_WORDS,   ALLOW_TYPOS,    Desc(S("asc_desc_rank")));
+#[rustfmt::skip]
 test_criterion!(desc_disallow_typo,                  DISALLOW_OPTIONAL_WORDS,   DISALLOW_TYPOS, Desc(S("asc_desc_rank")));
+#[rustfmt::skip]
 test_criterion!(asc_unexisting_field_allow_typo,     DISALLOW_OPTIONAL_WORDS,   ALLOW_TYPOS,    Asc(S("unexisting_field")));
+#[rustfmt::skip]
 test_criterion!(asc_unexisting_field_disallow_typo,  DISALLOW_OPTIONAL_WORDS,   DISALLOW_TYPOS, Asc(S("unexisting_field")));
+#[rustfmt::skip]
 test_criterion!(desc_unexisting_field_allow_typo,    DISALLOW_OPTIONAL_WORDS,   ALLOW_TYPOS,    Desc(S("unexisting_field")));
+#[rustfmt::skip]
 test_criterion!(desc_unexisting_field_disallow_typo, DISALLOW_OPTIONAL_WORDS,   DISALLOW_TYPOS, Desc(S("unexisting_field")));
 
 #[test]
 fn criteria_mixup() {
     use Criterion::*;
-    let index = search::setup_search_index_with_criteria(&vec![Words, Attribute, Desc(S("asc_desc_rank")), Exactness, Proximity, Typo]);
+    let index = search::setup_search_index_with_criteria(&vec![
+        Words,
+        Attribute,
+        Desc(S("asc_desc_rank")),
+        Exactness,
+        Proximity,
+        Typo,
+    ]);
 
+    #[rustfmt::skip]
     let criteria_mix = {
         // Criterion doesn't implement Copy, we create a new Criterion using a closure
         let desc = || Desc(S("asc_desc_rank"));
@@ -205,10 +230,11 @@ fn criteria_mixup() {
 
         let SearchResult { documents_ids, .. } = search.execute().unwrap();
 
-        let expected_external_ids: Vec<_> = search::expected_order(&criteria, ALLOW_OPTIONAL_WORDS, ALLOW_TYPOS)
-            .into_iter()
-            .map(|d| d.id)
-            .collect();
+        let expected_external_ids: Vec<_> =
+            search::expected_order(&criteria, ALLOW_OPTIONAL_WORDS, ALLOW_TYPOS)
+                .into_iter()
+                .map(|d| d.id)
+                .collect();
         let documents_ids = search::internal_to_external_ids(&index, &documents_ids);
 
         assert_eq!(documents_ids, expected_external_ids);
diff --git a/script/pre-commit b/script/pre-commit
new file mode 100755
index 000000000..4819a3b52
--- /dev/null
+++ b/script/pre-commit
@@ -0,0 +1,36 @@
+#!/usr/bin/env bash
+
+cargo check --workspace --all-targets &>/dev/null
+result=$?
+
+if [[ ${result} -ne 0 ]] ; then
+	cat <<\EOF
+The project does not compile. You might want to fix your error before commiting.
+
+If you still want to commit you can do it by appending
+--no-verify
+at the end of your previous command.
+
+If you are running a variant of bash you can directly paste this command in your terminal:
+!! --no-verify
+EOF
+    exit 1
+fi
+
+cargo fmt --all -- --check &>/dev/null
+result=$?
+
+if [[ ${result} -ne 0 ]] ; then
+	cat <<\EOF
+The project is badly formatted. Please run:
+cargo fmt --all
+
+If you want to create your commit without propper formatting you can add
+--no-verify
+at the end of your commit.
+
+If you are running a variant of bash you can directly paste this command in your terminal:
+!! --no-verify
+EOF
+    exit 1
+fi
diff --git a/qc_loop.sh b/script/qc_loop.sh
similarity index 100%
rename from qc_loop.sh
rename to script/qc_loop.sh
diff --git a/search/src/main.rs b/search/src/main.rs
index f7f95b730..fba714dab 100644
--- a/search/src/main.rs
+++ b/search/src/main.rs
@@ -6,10 +6,9 @@ use std::time::Instant;
 use byte_unit::Byte;
 use heed::EnvOpenOptions;
 use log::debug;
+use milli::{obkv_to_json, Index};
 use structopt::StructOpt;
 
-use milli::{Index, obkv_to_json};
-
 #[cfg(target_os = "linux")]
 #[global_allocator]
 static ALLOC: jemallocator::Jemalloc = jemallocator::Jemalloc;
@@ -86,7 +85,8 @@ fn main() -> anyhow::Result<()> {
         }
 
         if opt.print_facet_distribution {
-            let facets = index.facets_distribution(&rtxn).candidates(result.candidates).execute()?;
+            let facets =
+                index.facets_distribution(&rtxn).candidates(result.candidates).execute()?;
             serde_json::to_writer(&mut stdout, &facets)?;
             let _ = writeln!(&mut stdout);
         }