Merge 7882819795 into c1d8ee2a8d

crates/meilisearch/src/search/mod.rs

@@ -1733,51 +1733,46 @@ fn format_fields(
     // select the attributes to retrieve
     let displayable_names =
         displayable_ids.iter().map(|&fid| field_ids_map.name(fid).expect("Missing field name"));
-    permissive_json_pointer::map_leaf_values(
+    permissive_json_pointer::map_leaf_values(&mut document, displayable_names, |key, value| {
-        &mut document,
+        // To get the formatting option of each key we need to see all the rules that applies
-        displayable_names,
+        // to the value and merge them together. eg. If a user said he wanted to highlight `doggo`
-        |key, array_indices, value| {
+        // and crop `doggo.name`. `doggo.name` needs to be highlighted + cropped while `doggo.age` is only
-            // To get the formatting option of each key we need to see all the rules that applies
+        // highlighted.
-            // to the value and merge them together. eg. If a user said he wanted to highlight `doggo`
+        // Warn: The time to compute the format list scales with the number of fields to format;
-            // and crop `doggo.name`. `doggo.name` needs to be highlighted + cropped while `doggo.age` is only
+        // cumulated with map_leaf_values that iterates over all the nested fields, it gives a quadratic complexity:
-            // highlighted.
+        // d*f where d is the total number of fields to display and f is the total number of fields to format.
-            // Warn: The time to compute the format list scales with the number of fields to format;
+        let format = formatting_fields_options
-            // cumulated with map_leaf_values that iterates over all the nested fields, it gives a quadratic complexity:
+            .iter()
-            // d*f where d is the total number of fields to display and f is the total number of fields to format.
+            .filter(|(name, _option)| {
-            let format = formatting_fields_options
+                milli::is_faceted_by(name, key) || milli::is_faceted_by(key, name)
+            })
+            .map(|(_, option)| **option)
+            .reduce(|acc, option| acc.merge(option));
+        let mut infos = Vec::new();
+
+        // if no locales has been provided, we try to find the locales in the localized_attributes.
+        let locales = locales.or_else(|| {
+            localized_attributes
                 .iter()
-                .filter(|(name, _option)| {
+                .find(|rule| rule.match_str(key))
-                    milli::is_faceted_by(name, key) || milli::is_faceted_by(key, name)
+                .map(LocalizedAttributesRule::locales)
-                })
+        });
-                .map(|(_, option)| **option)
-                .reduce(|acc, option| acc.merge(option));
-            let mut infos = Vec::new();
 
-            // if no locales has been provided, we try to find the locales in the localized_attributes.
+        *value = format_value(
-            let locales = locales.or_else(|| {
+            std::mem::take(value),
-                localized_attributes
+            builder,
-                    .iter()
+            format,
-                    .find(|rule| rule.match_str(key))
+            &mut infos,
-                    .map(LocalizedAttributesRule::locales)
+            compute_matches,
-            });
+            locales,
+        );
 
-            *value = format_value(
+        if let Some(matches) = matches_position.as_mut() {
-                std::mem::take(value),
+            if !infos.is_empty() {
-                builder,
+                matches.insert(key.to_owned(), infos);
-                format,
-                &mut infos,
-                compute_matches,
-                array_indices,
-                locales,
-            );
-
-            if let Some(matches) = matches_position.as_mut() {
-                if !infos.is_empty() {
-                    matches.insert(key.to_owned(), infos);
-                }
             }
-        },
+        }
-    );
+    });
 
     let selectors = formatted_options
         .keys()
@@ -1795,14 +1790,13 @@ fn format_value(
     format_options: Option<FormatOptions>,
     infos: &mut Vec<MatchBounds>,
     compute_matches: bool,
-    array_indices: &[usize],
     locales: Option<&[Language]>,
 ) -> Value {
     match value {
         Value::String(old_string) => {
             let mut matcher = builder.build(&old_string, locales);
             if compute_matches {
-                let matches = matcher.matches(array_indices);
+                let matches = matcher.matches();
                 infos.extend_from_slice(&matches[..]);
             }
 
@@ -1814,15 +1808,51 @@ fn format_value(
                 None => Value::String(old_string),
             }
         }
-        // `map_leaf_values` makes sure this is only called for leaf fields
+        Value::Array(values) => Value::Array(
-        Value::Array(_) => unreachable!(),
+            values
-        Value::Object(_) => unreachable!(),
+                .into_iter()
+                .map(|v| {
+                    format_value(
+                        v,
+                        builder,
+                        format_options.map(|format_options| FormatOptions {
+                            highlight: format_options.highlight,
+                            crop: None,
+                        }),
+                        infos,
+                        compute_matches,
+                        locales,
+                    )
+                })
+                .collect(),
+        ),
+        Value::Object(object) => Value::Object(
+            object
+                .into_iter()
+                .map(|(k, v)| {
+                    (
+                        k,
+                        format_value(
+                            v,
+                            builder,
+                            format_options.map(|format_options| FormatOptions {
+                                highlight: format_options.highlight,
+                                crop: None,
+                            }),
+                            infos,
+                            compute_matches,
+                            locales,
+                        ),
+                    )
+                })
+                .collect(),
+        ),
         Value::Number(number) => {
             let s = number.to_string();
 
             let mut matcher = builder.build(&s, locales);
             if compute_matches {
-                let matches = matcher.matches(array_indices);
+                let matches = matcher.matches();
                 infos.extend_from_slice(&matches[..]);
             }
 

crates/meilisearch/tests/search/formatted.rs

@@ -208,10 +208,7 @@ async fn format_nested() {
                         "doggos.name": [
                           {
                             "start": 0,
-                            "length": 5,
+                            "length": 5
-                            "indices": [
-                              0
-                            ]
                           }
                         ]
                       }

crates/milli/src/search/new/matches/mod.rs

@@ -105,8 +105,6 @@ impl FormatOptions {
 pub struct MatchBounds {
     pub start: usize,
     pub length: usize,
-    #[serde(skip_serializing_if = "Option::is_none")]
-    pub indices: Option<Vec<usize>>,
 }
 
 /// Structure used to analyze a string, compute words that match,
@@ -222,20 +220,15 @@ impl<'t, 'tokenizer> Matcher<'t, 'tokenizer, '_, '_> {
     }
 
     /// Returns boundaries of the words that match the query.
-    pub fn matches(&mut self, array_indices: &[usize]) -> Vec<MatchBounds> {
+    pub fn matches(&mut self) -> Vec<MatchBounds> {
         match &self.matches {
-            None => self.compute_matches().matches(array_indices),
+            None => self.compute_matches().matches(),
             Some((tokens, matches)) => matches
                 .iter()
                 .map(|m| MatchBounds {
                     start: tokens[m.get_first_token_pos()].byte_start,
                     // TODO: Why is this in chars, while start is in bytes?
                     length: m.char_count,
-                    indices: if array_indices.is_empty() {
-                        None
-                    } else {
-                        Some(array_indices.to_owned())
-                    },
                 })
                 .collect(),
         }

crates/permissive-json-pointer/src/lib.rs

@@ -45,7 +45,7 @@ fn contained_in(selector: &str, key: &str) -> bool {
 /// map_leaf_values(
 ///     value.as_object_mut().unwrap(),
 ///     ["jean.race.name"],
-///     |key, _array_indices, value| match (value, key) {
+///     |key, value| match (value, key) {
 ///         (Value::String(name), "jean.race.name") => *name = "patou".to_string(),
 ///         _ => unreachable!(),
 ///     },
@@ -66,18 +66,17 @@ fn contained_in(selector: &str, key: &str) -> bool {
 pub fn map_leaf_values<'a>(
     value: &mut Map<String, Value>,
     selectors: impl IntoIterator<Item = &'a str>,
-    mut mapper: impl FnMut(&str, &[usize], &mut Value),
+    mut mapper: impl FnMut(&str, &mut Value),
 ) {
     let selectors: Vec<_> = selectors.into_iter().collect();
-    map_leaf_values_in_object(value, &selectors, "", &[], &mut mapper);
+    map_leaf_values_in_object(value, &selectors, "", &mut mapper);
 }
 
 pub fn map_leaf_values_in_object(
     value: &mut Map<String, Value>,
     selectors: &[&str],
     base_key: &str,
-    array_indices: &[usize],
+    mapper: &mut impl FnMut(&str, &mut Value),
-    mapper: &mut impl FnMut(&str, &[usize], &mut Value),
 ) {
     for (key, value) in value.iter_mut() {
         let base_key = if base_key.is_empty() {
@@ -95,12 +94,12 @@ pub fn map_leaf_values_in_object(
         if should_continue {
             match value {
                 Value::Object(object) => {
-                    map_leaf_values_in_object(object, selectors, &base_key, array_indices, mapper)
+                    map_leaf_values_in_object(object, selectors, &base_key, mapper)
                 }
                 Value::Array(array) => {
-                    map_leaf_values_in_array(array, selectors, &base_key, array_indices, mapper)
+                    map_leaf_values_in_array(array, selectors, &base_key, mapper)
                 }
-                value => mapper(&base_key, array_indices, value),
+                value => mapper(&base_key, value),
             }
         }
     }
@@ -110,24 +109,13 @@ pub fn map_leaf_values_in_array(
     values: &mut [Value],
     selectors: &[&str],
     base_key: &str,
-    base_array_indices: &[usize],
+    mapper: &mut impl FnMut(&str, &mut Value),
-    mapper: &mut impl FnMut(&str, &[usize], &mut Value),
 ) {
-    // This avoids allocating twice
+    for value in values.iter_mut() {
-    let mut array_indices = Vec::with_capacity(base_array_indices.len() + 1);
-    array_indices.extend_from_slice(base_array_indices);
-    array_indices.push(0);
-
-    for (i, value) in values.iter_mut().enumerate() {
-        *array_indices.last_mut().unwrap() = i;
         match value {
-            Value::Object(object) => {
+            Value::Object(object) => map_leaf_values_in_object(object, selectors, base_key, mapper),
-                map_leaf_values_in_object(object, selectors, base_key, &array_indices, mapper)
+            Value::Array(array) => map_leaf_values_in_array(array, selectors, base_key, mapper),
-            }
+            value => mapper(base_key, value),
-            Value::Array(array) => {
-                map_leaf_values_in_array(array, selectors, base_key, &array_indices, mapper)
-            }
-            value => mapper(base_key, &array_indices, value),
         }
     }
 }
@@ -755,14 +743,12 @@ mod tests {
             }
         });
 
-        map_leaf_values(
+        map_leaf_values(value.as_object_mut().unwrap(), ["jean.race.name"], |key, value| {
-            value.as_object_mut().unwrap(),
+            match (value, key) {
-            ["jean.race.name"],
-            |key, _, value| match (value, key) {
                 (Value::String(name), "jean.race.name") => *name = S("patou"),
                 _ => unreachable!(),
-            },
+            }
-        );
+        });
 
         assert_eq!(
             value,
@@ -789,7 +775,7 @@ mod tests {
         });
 
         let mut calls = 0;
-        map_leaf_values(value.as_object_mut().unwrap(), ["jean"], |key, _, value| {
+        map_leaf_values(value.as_object_mut().unwrap(), ["jean"], |key, value| {
             calls += 1;
             match (value, key) {
                 (Value::String(name), "jean.race.name") => *name = S("patou"),
@@ -812,52 +798,4 @@ mod tests {
             })
         );
     }
-
-    #[test]
-    fn map_array() {
-        let mut value: Value = json!({
-            "no_array": "peter",
-            "simple": ["foo", "bar"],
-            "nested": [
-                {
-                    "a": [
-                        ["cat", "dog"],
-                        ["fox", "bear"],
-                    ],
-                    "b": "hi",
-                },
-                {
-                    "a": ["green", "blue"],
-                },
-            ],
-        });
-
-        map_leaf_values(
-            value.as_object_mut().unwrap(),
-            ["no_array", "simple", "nested"],
-            |_key, array_indices, value| {
-                *value = format!("{array_indices:?}").into();
-            },
-        );
-
-        assert_eq!(
-            value,
-            json!({
-                "no_array": "[]",
-                "simple": ["[0]", "[1]"],
-                "nested": [
-                    {
-                        "a": [
-                            ["[0, 0, 0]", "[0, 0, 1]"],
-                            ["[0, 1, 0]", "[0, 1, 1]"],
-                        ],
-                        "b": "[0]",
-                    },
-                    {
-                        "a": ["[1, 0]", "[1, 1]"],
-                    },
-                ],
-            })
-        );
-    }
 }