tdsds.com/gitea
1
1
Fork 0
mirror of https://github.com/go-gitea/gitea synced 2025-10-31 11:28:24 +00:00
Code Releases Activity

Patch in exact search for meilisearch (#29671)

Browse Source 
meilisearch does not have an search option to contorl fuzzynes per query
right now:
 - https://github.com/meilisearch/meilisearch/issues/1192
 - https://github.com/orgs/meilisearch/discussions/377
 - https://github.com/meilisearch/meilisearch/discussions/1096

so we have to create a workaround by post-filter the search result in
gitea until this is addressed.

For future works I added an option in backend only atm, to enable
fuzzynes for issue indexer too.
And also refactored the code so the fuzzy option is equal in logic to
code indexer


---
*Sponsored by Kithara Software GmbH*
This commit is contained in:
6543
2024-03-09 02:39:27 +01:00
committed by GitHub
parent baeb251174
commit 7fdc048153
14 changed files with 184 additions and 33 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
modules/indexer/code/bleve/bleve.go
View File

@@ -233,21 +233,21 @@ func (b *Indexer) Delete(_ context.Context, repoID int64) error {
// Search searches for files in the specified repo.
// Returns the matching file-paths
func (b *Indexer) Search(ctx context.Context, repoIDs []int64, language, keyword string, page, pageSize int, isMatch bool) (int64, []*internal.SearchResult, []*internal.SearchResultLanguages, error) {
func (b *Indexer) Search(ctx context.Context, repoIDs []int64, language, keyword string, page, pageSize int, isFuzzy bool) (int64, []*internal.SearchResult, []*internal.SearchResultLanguages, error) {
var (
indexerQuery query.Query
keywordQuery query.Query
)
if isMatch {
prefixQuery := bleve.NewPrefixQuery(keyword)
prefixQuery.FieldVal = "Content"
keywordQuery = prefixQuery
} else {
if isFuzzy {
phraseQuery := bleve.NewMatchPhraseQuery(keyword)
phraseQuery.FieldVal = "Content"
phraseQuery.Analyzer = repoIndexerAnalyzer
keywordQuery = phraseQuery
} else {
prefixQuery := bleve.NewPrefixQuery(keyword)
prefixQuery.FieldVal = "Content"
keywordQuery = prefixQuery
}
if len(repoIDs) > 0 {

8
modules/indexer/code/elasticsearch/elasticsearch.go
View File

@@ -281,10 +281,10 @@ func extractAggs(searchResult *elastic.SearchResult) []*internal.SearchResultLan
}
// Search searches for codes and language stats by given conditions.
func (b *Indexer) Search(ctx context.Context, repoIDs []int64, language, keyword string, page, pageSize int, isMatch bool) (int64, []*internal.SearchResult, []*internal.SearchResultLanguages, error) {
searchType := esMultiMatchTypeBestFields
if isMatch {
searchType = esMultiMatchTypePhrasePrefix
func (b *Indexer) Search(ctx context.Context, repoIDs []int64, language, keyword string, page, pageSize int, isFuzzy bool) (int64, []*internal.SearchResult, []*internal.SearchResultLanguages, error) {
searchType := esMultiMatchTypePhrasePrefix
if isFuzzy {
searchType = esMultiMatchTypeBestFields
}
kwQuery := elastic.NewMultiMatchQuery(keyword, "content").Type(searchType)

2
modules/indexer/code/indexer_test.go
View File

@@ -70,7 +70,7 @@ func testIndexer(name string, t *testing.T, indexer internal.Indexer) {
for _, kw := range keywords {
t.Run(kw.Keyword, func(t *testing.T) {
total, res, langs, err := indexer.Search(context.TODO(), kw.RepoIDs, "", kw.Keyword, 1, 10, false)
total, res, langs, err := indexer.Search(context.TODO(), kw.RepoIDs, "", kw.Keyword, 1, 10, true)
assert.NoError(t, err)
assert.Len(t, kw.IDs, int(total))
assert.Len(t, langs, kw.Langs)

4
modules/indexer/code/internal/indexer.go
View File

@@ -16,7 +16,7 @@ type Indexer interface {
internal.Indexer
Index(ctx context.Context, repo *repo_model.Repository, sha string, changes *RepoChanges) error
Delete(ctx context.Context, repoID int64) error
Search(ctx context.Context, repoIDs []int64, language, keyword string, page, pageSize int, isMatch bool) (int64, []*SearchResult, []*SearchResultLanguages, error)
Search(ctx context.Context, repoIDs []int64, language, keyword string, page, pageSize int, isFuzzy bool) (int64, []*SearchResult, []*SearchResultLanguages, error)
}
// NewDummyIndexer returns a dummy indexer
@@ -38,6 +38,6 @@ func (d *dummyIndexer) Delete(ctx context.Context, repoID int64) error {
return fmt.Errorf("indexer is not ready")
}
func (d *dummyIndexer) Search(ctx context.Context, repoIDs []int64, language, keyword string, page, pageSize int, isMatch bool) (int64, []*SearchResult, []*SearchResultLanguages, error) {
func (d *dummyIndexer) Search(ctx context.Context, repoIDs []int64, language, keyword string, page, pageSize int, isFuzzy bool) (int64, []*SearchResult, []*SearchResultLanguages, error) {
return 0, nil, nil, fmt.Errorf("indexer is not ready")
}

5
modules/indexer/code/search.go
View File

@@ -124,12 +124,13 @@ func searchResult(result *internal.SearchResult, startIndex, endIndex int) (*Res
}
// PerformSearch perform a search on a repository
func PerformSearch(ctx context.Context, repoIDs []int64, language, keyword string, page, pageSize int, isMatch bool) (int, []*Result, []*internal.SearchResultLanguages, error) {
// if isFuzzy is true set the Damerau-Levenshtein distance from 0 to 2
func PerformSearch(ctx context.Context, repoIDs []int64, language, keyword string, page, pageSize int, isFuzzy bool) (int, []*Result, []*internal.SearchResultLanguages, error) {
if len(keyword) == 0 {
return 0, nil, nil, nil
}
total, results, resultLanguages, err := (*globalIndexer.Load()).Search(ctx, repoIDs, language, keyword, page, pageSize, isMatch)
total, results, resultLanguages, err := (*globalIndexer.Load()).Search(ctx, repoIDs, language, keyword, page, pageSize, isFuzzy)
if err != nil {
return 0, nil, nil, err
}