snowykami/meilisearch
1
0
Fork 0
mirror of https://github.com/meilisearch/meilisearch.git synced 2024-11-23 10:37:41 +08:00
Code Issues Actions 3 Packages Projects Releases Wiki Activity

feat: Unique word has multi-word synonyms basically work

Browse Source
This commit is contained in:
Clément Renault 2019-06-17 10:28:43 +02:00
parent 9861c3878e
commit 6cb57aa8a4
No known key found for this signature in database
GPG Key ID: 0151CDAB43460DAE
2 changed files with 49 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
meilidb-core/src/criterion/mod.rs
View File

@ -113,7 +113,7 @@ impl<'a> Default for Criteria<'a> {
}
}
impl<'a> AsRef<[Box<Criterion + 'a>]> for Criteria<'a> {
impl<'a> AsRef<[Box<dyn Criterion + 'a>]> for Criteria<'a> {
fn as_ref(&self) -> &[Box<dyn Criterion + 'a>] {
&self.inner
}

50
meilidb-core/src/query_builder.rs
View File

@ -37,8 +37,10 @@ fn generate_automatons<S: Store>(query: &str, store: &S) -> Result<Vec<(usize, D
let mut stream = words.into_stream();
while let Some(word) = stream.next() {
let word = std::str::from_utf8(word).unwrap();
let lev = if not_prefix_dfa { build_dfa(word) } else { build_prefix_dfa(word) };
automatons.push((index, lev));
for word in split_query_string(word) {
let lev = if not_prefix_dfa { build_dfa(word) } else { build_prefix_dfa(word) };
automatons.push((index, lev));
}
}
}
}
@ -388,6 +390,7 @@ mod tests {
impl InMemorySetStore {
pub fn add_synonym(&mut self, word: &str, new: SetBuf<&str>) {
let word = word.to_lowercase();
let alternatives = self.alternatives.entry(word.as_bytes().to_vec()).or_default();
let new = sdset_into_fstset(&new);
*alternatives = set_from_stream(alternatives.op().add(new.into_stream()).r#union());
@ -641,4 +644,47 @@ mod tests {
});
assert_matches!(iter.next(), None);
}
/// Unique word has multi-word synonyms
#[test]
fn multiword_synonyms() {
let mut store = InMemorySetStore::from_iter(vec![
(&b"new"[..], &[doc_index(0, 0)][..]),
(&b"york"[..], &[doc_index(0, 1)][..]),
(&b"subway"[..], &[doc_index(0, 2)][..]),
]);
store.add_synonym("NY", SetBuf::from_dirty(vec!["NYC", "new york", "new york city"]));
store.add_synonym("NYC", SetBuf::from_dirty(vec!["NY", "new york", "new york city"]));
let builder = QueryBuilder::new(&store);
let results = builder.query("NY subway", 0..20).unwrap();
let mut iter = results.into_iter();
assert_matches!(iter.next(), Some(Document { id: DocumentId(0), matches }) => {
let mut iter = matches.into_iter();
assert_matches!(iter.next(), Some(Match { query_index: 0, word_index: 0, .. })); // new
assert_matches!(iter.next(), Some(Match { query_index: 0, word_index: 0, .. })); // new
assert_matches!(iter.next(), Some(Match { query_index: 0, word_index: 1, .. })); // york
assert_matches!(iter.next(), Some(Match { query_index: 0, word_index: 1, .. })); // york
assert_matches!(iter.next(), Some(Match { query_index: 1, word_index: 2, .. })); // subway
assert_matches!(iter.next(), None);
});
assert_matches!(iter.next(), None);
let builder = QueryBuilder::new(&store);
let results = builder.query("NYC subway", 0..20).unwrap();
let mut iter = results.into_iter();
assert_matches!(iter.next(), Some(Document { id: DocumentId(0), matches }) => {
let mut iter = matches.into_iter();
assert_matches!(iter.next(), Some(Match { query_index: 0, word_index: 0, .. })); // new
assert_matches!(iter.next(), Some(Match { query_index: 0, word_index: 0, .. })); // new
assert_matches!(iter.next(), Some(Match { query_index: 0, word_index: 1, .. })); // york
assert_matches!(iter.next(), Some(Match { query_index: 0, word_index: 1, .. })); // york
assert_matches!(iter.next(), Some(Match { query_index: 1, word_index: 2, .. })); // subway
assert_matches!(iter.next(), None);
});
assert_matches!(iter.next(), None);
}
}