From a2f0f953375ac4029aa210d67b21ac448f6ea1e7 Mon Sep 17 00:00:00 2001
From: qdequele <quentin@dequelen.me>
Date: Mon, 2 Mar 2020 14:34:29 +0100
Subject: [PATCH] use distinct on search

---
 meilisearch-core/src/query_builder.rs           |  2 +-
 meilisearch-core/src/store/mod.rs               | 15 +++++++++++++++
 meilisearch-core/src/update/settings_update.rs  |  3 +--
 meilisearch-http/src/helpers/meilisearch.rs     | 17 +++++++++++++++++
 .../tests/settings_ranking_rules.rs             | 12 ++++++------
 meilisearch-schema/src/schema.rs                |  4 ++++
 6 files changed, 44 insertions(+), 9 deletions(-)

diff --git a/meilisearch-core/src/query_builder.rs b/meilisearch-core/src/query_builder.rs
index 5bde93665..6e20d2463 100644
--- a/meilisearch-core/src/query_builder.rs
+++ b/meilisearch-core/src/query_builder.rs
@@ -75,7 +75,7 @@ impl<'c, 'f, 'd> QueryBuilder<'c, 'f, 'd> {
         self.timeout = Some(timeout)
     }
 
-    pub fn with_distinct<F, K>(&mut self, function: F, size: usize)
+    pub fn with_distinct<F>(&mut self, size: usize, function: F)
     where
         F: Fn(DocumentId) -> Option<u64> + 'd,
     {
diff --git a/meilisearch-core/src/store/mod.rs b/meilisearch-core/src/store/mod.rs
index 90c69cc5e..0b12f94dc 100644
--- a/meilisearch-core/src/store/mod.rs
+++ b/meilisearch-core/src/store/mod.rs
@@ -254,6 +254,21 @@ impl Index {
         }
     }
 
+    pub fn document_attribute_bytes<'txn>(
+        &self,
+        reader: &'txn heed::RoTxn<MainT>,
+        document_id: DocumentId,
+        attribute: FieldId,
+    ) -> MResult<Option<&'txn [u8]>> {
+        let bytes = self
+            .documents_fields
+            .document_attribute(reader, document_id, attribute)?;
+        match bytes {
+            Some(bytes) => Ok(Some(bytes)),
+            None => Ok(None),
+        }
+    }
+
     pub fn customs_update(&self, writer: &mut heed::RwTxn<UpdateT>, customs: Vec<u8>) -> ZResult<u64> {
         let _ = self.updates_notifier.send(UpdateEvent::NewUpdate);
         update::push_customs_update(writer, self.updates, self.updates_results, customs)
diff --git a/meilisearch-core/src/update/settings_update.rs b/meilisearch-core/src/update/settings_update.rs
index d0f10eb80..4fb6a8f47 100644
--- a/meilisearch-core/src/update/settings_update.rs
+++ b/meilisearch-core/src/update/settings_update.rs
@@ -50,9 +50,8 @@ pub fn apply_settings_update(
             must_reindex = true;
         },
         UpdateState::Clear => {
-            let clear: Vec<&str> = Vec::new();
-            schema.update_ranked(clear)?;
             index.main.delete_ranking_rules(writer)?;
+            schema.clear_ranked();
             must_reindex = true;
         },
         UpdateState::Nothing => (),
diff --git a/meilisearch-http/src/helpers/meilisearch.rs b/meilisearch-http/src/helpers/meilisearch.rs
index 6e15dcf0b..3d2b2f399 100644
--- a/meilisearch-http/src/helpers/meilisearch.rs
+++ b/meilisearch-http/src/helpers/meilisearch.rs
@@ -3,6 +3,7 @@ use std::collections::{HashMap, HashSet};
 use std::convert::From;
 use std::error;
 use std::fmt;
+use std::hash::{Hash, Hasher};
 use std::time::{Duration, Instant};
 
 use indexmap::IndexMap;
@@ -13,6 +14,7 @@ use meilisearch_core::{Highlight, Index, MainT, RankedMap};
 use meilisearch_schema::{FieldId, Schema};
 use serde::{Deserialize, Serialize};
 use serde_json::Value;
+use siphasher::sip::SipHasher;
 
 #[derive(Debug)]
 pub enum Error {
@@ -205,6 +207,21 @@ impl<'a> SearchBuilder<'a> {
 
         query_builder.with_fetch_timeout(self.timeout);
 
+        if let Some(field) = self.index.main.distinct_attribute(reader)? {
+            if let Some(field_id) = schema.id(&field) {
+                query_builder.with_distinct(1, move |id| {
+                    match self.index.document_attribute_bytes(reader, id, field_id) {
+                        Ok(Some(bytes)) => {
+                            let mut s = SipHasher::new();
+                            bytes.hash(&mut s);
+                            Some(s.finish())
+                        }
+                        _ => None,
+                    }
+                });
+            }
+        }
+
         let start = Instant::now();
         let docs =
             query_builder.query(reader, &self.query, self.offset..(self.offset + self.limit));
diff --git a/meilisearch-http/tests/settings_ranking_rules.rs b/meilisearch-http/tests/settings_ranking_rules.rs
index f221aacfb..786f72dd5 100644
--- a/meilisearch-http/tests/settings_ranking_rules.rs
+++ b/meilisearch-http/tests/settings_ranking_rules.rs
@@ -96,12 +96,12 @@ fn write_all_and_delete() {
     let res_value: Value = serde_json::from_slice(&buf).unwrap();
 
     let json = json!([
-      "typo",
-      "words",
-      "proximity",
-      "attribute",
-      "wordsPosition",
-      "exactness"
+        "typo",
+        "words",
+        "proximity",
+        "attribute",
+        "wordsPosition",
+        "exactness"
     ]);
 
     assert_json_eq!(json, res_value, ordered: false);
diff --git a/meilisearch-schema/src/schema.rs b/meilisearch-schema/src/schema.rs
index 7a4180cb5..4283f203d 100644
--- a/meilisearch-schema/src/schema.rs
+++ b/meilisearch-schema/src/schema.rs
@@ -125,6 +125,10 @@ impl Schema {
         Ok((id, pos.into()))
     }
 
+    pub fn clear_ranked(&mut self) {
+        self.ranked.clear();
+    }
+
     pub fn remove_ranked(&mut self, name: &str) {
         if let Some(id) = self.fields_map.id(name) {
             self.ranked.remove(&id);