diff options
| author | wxiaoguang <wxiaoguang@gmail.com> | 2025-03-13 11:07:48 +0800 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2025-03-13 11:07:48 +0800 |
| commit | 403775e74ed11097b30a9a683b2253c0967c0dcd (patch) | |
| tree | 3d5d4c7687428cad146f7650822c1f0a2af72240 /modules/indexer/code/elasticsearch | |
| parent | cd10456664ce763212f6841be7839ab5d4229c18 (diff) | |
| download | gitea-403775e74ed11097b30a9a683b2253c0967c0dcd.tar.gz gitea-403775e74ed11097b30a9a683b2253c0967c0dcd.zip | |
Improve issue & code search (#33860)
Each "indexer" should provide the "search modes" they support by
themselves. And we need to remove the "fuzzy" search for code.
Diffstat (limited to 'modules/indexer/code/elasticsearch')
| -rw-r--r-- | modules/indexer/code/elasticsearch/elasticsearch.go | 19 |
1 files changed, 9 insertions, 10 deletions
diff --git a/modules/indexer/code/elasticsearch/elasticsearch.go b/modules/indexer/code/elasticsearch/elasticsearch.go index 9abece921e..354a8334fb 100644 --- a/modules/indexer/code/elasticsearch/elasticsearch.go +++ b/modules/indexer/code/elasticsearch/elasticsearch.go @@ -16,6 +16,7 @@ import ( "code.gitea.io/gitea/modules/charset" "code.gitea.io/gitea/modules/git" "code.gitea.io/gitea/modules/gitrepo" + "code.gitea.io/gitea/modules/indexer" "code.gitea.io/gitea/modules/indexer/code/internal" indexer_internal "code.gitea.io/gitea/modules/indexer/internal" inner_elasticsearch "code.gitea.io/gitea/modules/indexer/internal/elasticsearch" @@ -24,7 +25,6 @@ import ( "code.gitea.io/gitea/modules/setting" "code.gitea.io/gitea/modules/timeutil" "code.gitea.io/gitea/modules/typesniffer" - "code.gitea.io/gitea/modules/util" "github.com/go-enry/go-enry/v2" "github.com/olivere/elastic/v7" @@ -46,6 +46,10 @@ type Indexer struct { indexer_internal.Indexer // do not composite inner_elasticsearch.Indexer directly to avoid exposing too much } +func (b *Indexer) SupportedSearchModes() []indexer.SearchMode { + return indexer.SearchModesExactWords() +} + // NewIndexer creates a new elasticsearch indexer func NewIndexer(url, indexerName string) *Indexer { inner := inner_elasticsearch.NewIndexer(url, indexerName, esRepoIndexerLatestVersion, defaultMapping) @@ -361,15 +365,10 @@ func extractAggs(searchResult *elastic.SearchResult) []*internal.SearchResultLan // Search searches for codes and language stats by given conditions. func (b *Indexer) Search(ctx context.Context, opts *internal.SearchOptions) (int64, []*internal.SearchResult, []*internal.SearchResultLanguages, error) { var contentQuery elastic.Query - keywordAsPhrase, isPhrase := internal.ParseKeywordAsPhrase(opts.Keyword) - if isPhrase { - contentQuery = elastic.NewMatchPhraseQuery("content", keywordAsPhrase) - } else { - // TODO: this is the old logic, but not really using "fuzziness" - // * IsKeywordFuzzy=true: "best_fields" - // * IsKeywordFuzzy=false: "phrase_prefix" - contentQuery = elastic.NewMultiMatchQuery("content", opts.Keyword). - Type(util.Iif(opts.IsKeywordFuzzy, esMultiMatchTypeBestFields, esMultiMatchTypePhrasePrefix)) + if opts.SearchMode == indexer.SearchModeExact { + contentQuery = elastic.NewMatchPhraseQuery("content", opts.Keyword) + } else /* words */ { + contentQuery = elastic.NewMultiMatchQuery("content", opts.Keyword).Type(esMultiMatchTypeBestFields).Operator("and") } kwQuery := elastic.NewBoolQuery().Should( contentQuery, |