From f496cd320d39c7219cc1679c3086cdb002561589 Mon Sep 17 00:00:00 2001
From: many <maxime@meilisearch.com>
Date: Thu, 17 Jun 2021 14:24:59 +0200
Subject: [PATCH] Add distinct integration tests

---
 milli/tests/search/distinct.rs       | 76 ++++++++++++++++++++++++++++
 milli/tests/search/mod.rs            |  1 +
 milli/tests/search/query_criteria.rs |  4 +-
 3 files changed, 79 insertions(+), 2 deletions(-)
 create mode 100644 milli/tests/search/distinct.rs

diff --git a/milli/tests/search/distinct.rs b/milli/tests/search/distinct.rs
new file mode 100644
index 000000000..1204a6bfb
--- /dev/null
+++ b/milli/tests/search/distinct.rs
@@ -0,0 +1,76 @@
+use std::collections::HashSet;
+
+use big_s::S;
+use milli::update::Settings;
+use milli::{Criterion, Search, SearchResult};
+use Criterion::*;
+
+use crate::search::{self, EXTERNAL_DOCUMENTS_IDS};
+
+macro_rules! test_distinct {
+    ($func:ident, $distinct:ident, $criteria:expr) => {
+        #[test]
+        fn $func() {
+            let criteria = $criteria;
+            let index = search::setup_search_index_with_criteria(&criteria);
+
+            // update distinct attribute
+            let mut wtxn = index.write_txn().unwrap();
+            let mut builder = Settings::new(&mut wtxn, &index, 0);
+            builder.set_distinct_field(S(stringify!($distinct)));
+            builder.execute(|_, _| ()).unwrap();
+            wtxn.commit().unwrap();
+
+            let mut rtxn = index.read_txn().unwrap();
+
+            let mut search = Search::new(&mut rtxn, &index);
+            search.query(search::TEST_QUERY);
+            search.limit(EXTERNAL_DOCUMENTS_IDS.len());
+            search.authorize_typos(true);
+            search.optional_words(true);
+
+            let SearchResult { documents_ids, .. } = search.execute().unwrap();
+
+            let mut distinct_values = HashSet::new();
+            let expected_external_ids: Vec<_> = search::expected_order(&criteria, true, true)
+                .into_iter()
+                .filter_map(|d| {
+                    if distinct_values.contains(&d.$distinct) {
+                        None
+                    } else {
+                        distinct_values.insert(d.$distinct.to_owned());
+                        Some(d.id)
+                    }
+                })
+                .collect();
+
+            let documents_ids = search::internal_to_external_ids(&index, &documents_ids);
+            assert_eq!(documents_ids, expected_external_ids);
+        }
+    };
+}
+
+#[rustfmt::skip]
+test_distinct!(distinct_string_default_criteria,          tag,           vec![Words, Typo, Proximity, Attribute, Exactness]);
+#[rustfmt::skip]
+test_distinct!(distinct_number_default_criteria,          asc_desc_rank, vec![Words, Typo, Proximity, Attribute, Exactness]);
+#[rustfmt::skip]
+test_distinct!(distinct_string_criterion_words,           tag,           vec![Words]);
+#[rustfmt::skip]
+test_distinct!(distinct_number_criterion_words,           asc_desc_rank, vec![Words]);
+#[rustfmt::skip]
+test_distinct!(distinct_string_criterion_words_typo,      tag,           vec![Words, Typo]);
+#[rustfmt::skip]
+test_distinct!(distinct_number_criterion_words_typo,      asc_desc_rank, vec![Words, Typo]);
+#[rustfmt::skip]
+test_distinct!(distinct_string_criterion_words_proximity, tag,           vec![Words, Proximity]);
+#[rustfmt::skip]
+test_distinct!(distinct_number_criterion_words_proximity, asc_desc_rank, vec![Words, Proximity]);
+#[rustfmt::skip]
+test_distinct!(distinct_string_criterion_words_attribute, tag,           vec![Words, Attribute]);
+#[rustfmt::skip]
+test_distinct!(distinct_number_criterion_words_attribute, asc_desc_rank, vec![Words, Attribute]);
+#[rustfmt::skip]
+test_distinct!(distinct_string_criterion_words_exactness, tag,           vec![Words, Exactness]);
+#[rustfmt::skip]
+test_distinct!(distinct_number_criterion_words_exactness, asc_desc_rank, vec![Words, Exactness]);
diff --git a/milli/tests/search/mod.rs b/milli/tests/search/mod.rs
index e48d7704d..d37904942 100644
--- a/milli/tests/search/mod.rs
+++ b/milli/tests/search/mod.rs
@@ -9,6 +9,7 @@ use milli::{Criterion, DocumentId, Index};
 use serde::Deserialize;
 use slice_group_by::GroupBy;
 
+mod distinct;
 mod filters;
 mod query_criteria;
 
diff --git a/milli/tests/search/query_criteria.rs b/milli/tests/search/query_criteria.rs
index 7ab9897b3..486768228 100644
--- a/milli/tests/search/query_criteria.rs
+++ b/milli/tests/search/query_criteria.rs
@@ -11,10 +11,10 @@ const ALLOW_OPTIONAL_WORDS: bool = true;
 const DISALLOW_OPTIONAL_WORDS: bool = false;
 
 macro_rules! test_criterion {
-    ($func:ident, $optional_word:ident, $authorize_typos:ident , $criterion:expr) => {
+    ($func:ident, $optional_word:ident, $authorize_typos:ident, $criteria:expr) => {
         #[test]
         fn $func() {
-            let criteria = $criterion;
+            let criteria = $criteria;
             let index = search::setup_search_index_with_criteria(&criteria);
             let mut rtxn = index.read_txn().unwrap();