From bbbc4410acaf083a0553eaafae93835cc7236119 Mon Sep 17 00:00:00 2001
From: Kerollmops <clement@meilisearch.com>
Date: Mon, 20 Jan 2025 15:37:08 +0100
Subject: [PATCH] Fix the dump creation process

---
 .../src/scheduler/process_dump_creation.rs                | 8 +++++++-
 crates/milli/src/update/new/reindex.rs                    | 5 +----
 2 files changed, 8 insertions(+), 5 deletions(-)

diff --git a/crates/index-scheduler/src/scheduler/process_dump_creation.rs b/crates/index-scheduler/src/scheduler/process_dump_creation.rs
index 3fd5c795b..92464eac7 100644
--- a/crates/index-scheduler/src/scheduler/process_dump_creation.rs
+++ b/crates/index-scheduler/src/scheduler/process_dump_creation.rs
@@ -128,6 +128,7 @@ impl IndexScheduler {
             let embedding_configs = index
                 .embedding_configs(&rtxn)
                 .map_err(|e| Error::from_milli(e, Some(uid.to_string())))?;
+            let decompression_dictionary = index.document_decompression_dictionary(&rtxn)?;
 
             let nb_documents = index
                 .number_of_documents(&rtxn)
@@ -135,8 +136,9 @@ impl IndexScheduler {
                 as u32;
             let (atomic, update_document_progress) = AtomicDocumentStep::new(nb_documents);
             progress.update_progress(update_document_progress);
+            let doc_alloc = bumpalo::Bump::new();
             let documents = index
-                .all_documents(&rtxn)
+                .all_compressed_documents(&rtxn)
                 .map_err(|e| Error::from_milli(e, Some(uid.to_string())))?;
             // 3.1. Dump the documents
             for ret in documents {
@@ -145,6 +147,10 @@ impl IndexScheduler {
                 }
 
                 let (id, doc) = ret.map_err(|e| Error::from_milli(e, Some(uid.to_string())))?;
+                let doc = match decompression_dictionary.as_ref() {
+                    Some(dict) => doc.decompress_into_bump(&doc_alloc, dict)?,
+                    None => doc.as_non_compressed(),
+                };
 
                 let mut document = milli::obkv_to_json(&all_fields, &fields_ids_map, doc)
                     .map_err(|e| Error::from_milli(e, Some(uid.to_string())))?;
diff --git a/crates/milli/src/update/new/reindex.rs b/crates/milli/src/update/new/reindex.rs
index 036e0c838..8215b46c4 100644
--- a/crates/milli/src/update/new/reindex.rs
+++ b/crates/milli/src/update/new/reindex.rs
@@ -19,10 +19,7 @@ pub fn field_distribution(index: &Index, wtxn: &mut RwTxn<'_>, progress: &Progre
     let mut doc_alloc = bumpalo::Bump::new();
 
     let db_document_decompression_dictionary =
-        match index.document_compression_raw_dictionary(wtxn)? {
-            Some(raw) => Some(DecoderDictionary::copy(raw)),
-            None => None,
-        };
+        index.document_compression_raw_dictionary(wtxn)?.map(|raw| DecoderDictionary::copy(raw));
 
     for docid in docids {
         update_document_count.fetch_add(1, std::sync::atomic::Ordering::Relaxed);