Merge pull request #507 from meilisearch/fix-documents-fields-order-inference

Fix the inference of the documents searchable fields

meilisearch-core/Cargo.toml

@@ -17,6 +17,7 @@ env_logger = "0.7.0"
 fst = { version = "0.3.5", default-features = false }
 hashbrown = { version = "0.6.0", features = ["serde"] }
 heed = "0.6.1"
+indexmap = { version = "1.2.0", features = ["serde-1"] }
 intervaltree = "0.2.5"
 itertools = "0.8.2"
 levenshtein_automata = { version = "0.1.1", features = ["fst_automaton"] }
@@ -26,19 +27,18 @@ meilisearch-tokenizer = { path = "../meilisearch-tokenizer", version = "0.8.4" }
 meilisearch-types = { path = "../meilisearch-types", version = "0.8.4" }
 once_cell = "1.2.0"
 ordered-float = { version = "1.0.2", features = ["serde"] }
+regex = "1.3.1"
 sdset = "0.3.6"
 serde = { version = "1.0.101", features = ["derive"] }
 serde_json = "1.0.41"
 siphasher = "0.3.1"
 slice-group-by = "0.2.6"
 zerocopy = "0.2.8"
-regex = "1.3.1"
 
 [dev-dependencies]
 assert_matches = "1.3"
 criterion = "0.3"
 csv = "1.0.7"
-indexmap = { version = "1.2.0", features = ["serde-1"] }
 jemallocator = "0.3.2"
 rustyline = { version = "5.0.0", default-features = false }
 structopt = "0.3.2"

meilisearch-core/src/update/documents_addition.rs

@@ -1,6 +1,7 @@
 use std::collections::HashMap;
 
 use fst::{set::OpBuilder, SetBuilder};
+use indexmap::IndexMap;
 use sdset::{duo::Union, SetOperation};
 use serde::{Deserialize, Serialize};
 
@@ -105,7 +106,7 @@ pub fn push_documents_addition<D: serde::Serialize>(
 pub fn apply_documents_addition<'a, 'b>(
     writer: &'a mut heed::RwTxn<'b, MainT>,
     index: &store::Index,
-    addition: Vec<HashMap<String, serde_json::Value>>,
+    addition: Vec<IndexMap<String, serde_json::Value>>,
 ) -> MResult<()> {
     let mut documents_additions = HashMap::new();
 
@@ -174,7 +175,7 @@ pub fn apply_documents_addition<'a, 'b>(
 pub fn apply_documents_partial_addition<'a, 'b>(
     writer: &'a mut heed::RwTxn<'b, MainT>,
     index: &store::Index,
-    addition: Vec<HashMap<String, serde_json::Value>>,
+    addition: Vec<IndexMap<String, serde_json::Value>>,
 ) -> MResult<()> {
     let mut documents_additions = HashMap::new();
 

meilisearch-core/src/update/mod.rs

@@ -13,15 +13,15 @@ pub use self::documents_deletion::{apply_documents_deletion, DocumentsDeletion};
 pub use self::settings_update::{apply_settings_update, push_settings_update};
 
 use std::cmp;
-use std::collections::HashMap;
 use std::time::Instant;
 
 use chrono::{DateTime, Utc};
-use heed::Result as ZResult;
-use log::debug;
-use serde::{Deserialize, Serialize};
 use fst::{IntoStreamer, Streamer};
+use heed::Result as ZResult;
+use indexmap::IndexMap;
+use log::debug;
 use sdset::Set;
+use serde::{Deserialize, Serialize};
 
 use crate::{store, DocumentId, MResult};
 use crate::database::{MainT, UpdateT};
@@ -48,14 +48,14 @@ impl Update {
         }
     }
 
-    fn documents_addition(data: Vec<HashMap<String, serde_json::Value>>) -> Update {
+    fn documents_addition(data: Vec<IndexMap<String, serde_json::Value>>) -> Update {
         Update {
             data: UpdateData::DocumentsAddition(data),
             enqueued_at: Utc::now(),
         }
     }
 
-    fn documents_partial(data: Vec<HashMap<String, serde_json::Value>>) -> Update {
+    fn documents_partial(data: Vec<IndexMap<String, serde_json::Value>>) -> Update {
         Update {
             data: UpdateData::DocumentsPartial(data),
             enqueued_at: Utc::now(),
@@ -81,8 +81,8 @@ impl Update {
 pub enum UpdateData {
     ClearAll,
     Customs(Vec<u8>),
-    DocumentsAddition(Vec<HashMap<String, serde_json::Value>>),
-    DocumentsPartial(Vec<HashMap<String, serde_json::Value>>),
+    DocumentsAddition(Vec<IndexMap<String, serde_json::Value>>),
+    DocumentsPartial(Vec<IndexMap<String, serde_json::Value>>),
     DocumentsDeletion(Vec<DocumentId>),
     Settings(SettingsUpdate)
 }

meilisearch-http/src/routes/setting.rs

@@ -290,7 +290,7 @@ pub async fn get_searchable(ctx: Request<Data>) -> SResult<Response> {
 
     let schema = index.main.schema(&reader)?;
 
-    let searchable_attributes: Option<HashSet<String>> =
+    let searchable_attributes: Option<Vec<String>> =
         schema.map(|s| s.indexed_name().iter().map(|i| (*i).to_string()).collect());
 
     Ok(tide::Response::new(200)