Improve issue & code search (#33860)
Some checks are pending
release-nightly / nightly-binary (push) Waiting to run
release-nightly / nightly-docker-rootful (push) Waiting to run
release-nightly / nightly-docker-rootless (push) Waiting to run

Each "indexer" should provide the "search modes" they support by
themselves. And we need to remove the "fuzzy" search for code.
This commit is contained in:
wxiaoguang
2025-03-13 11:07:48 +08:00
committed by GitHub
parent cd10456664
commit 403775e74e
31 changed files with 317 additions and 172 deletions

View File

@ -11,6 +11,7 @@ import (
"code.gitea.io/gitea/models/db"
"code.gitea.io/gitea/models/unittest"
indexer_module "code.gitea.io/gitea/modules/indexer"
"code.gitea.io/gitea/modules/indexer/code/bleve"
"code.gitea.io/gitea/modules/indexer/code/elasticsearch"
"code.gitea.io/gitea/modules/indexer/code/internal"
@ -39,10 +40,11 @@ func testIndexer(name string, t *testing.T, indexer internal.Indexer) {
assert.NoError(t, setupRepositoryIndexes(t.Context(), indexer))
keywords := []struct {
RepoIDs []int64
Keyword string
Langs int
Results []codeSearchResult
RepoIDs []int64
Keyword string
Langs int
SearchMode indexer_module.SearchModeType
Results []codeSearchResult
}{
// Search for an exact match on the contents of a file
// This scenario yields a single result (the file README.md on the repo '1')
@ -183,9 +185,10 @@ func testIndexer(name string, t *testing.T, indexer internal.Indexer) {
},
// Search for matches on the contents of files regardless of case.
{
RepoIDs: nil,
Keyword: "dESCRIPTION",
Langs: 1,
RepoIDs: nil,
Keyword: "dESCRIPTION",
Langs: 1,
SearchMode: indexer_module.SearchModeFuzzy,
Results: []codeSearchResult{
{
Filename: "README.md",
@ -193,7 +196,7 @@ func testIndexer(name string, t *testing.T, indexer internal.Indexer) {
},
},
},
// Search for an exact match on the filename within the repo '62' (case insenstive).
// Search for an exact match on the filename within the repo '62' (case-insensitive).
// This scenario yields a single result (the file avocado.md on the repo '62')
{
RepoIDs: []int64{62},
@ -206,7 +209,7 @@ func testIndexer(name string, t *testing.T, indexer internal.Indexer) {
},
},
},
// Search for matches on the contents of files when the criteria is a expression.
// Search for matches on the contents of files when the criteria are an expression.
{
RepoIDs: []int64{62},
Keyword: "console.log",
@ -218,7 +221,7 @@ func testIndexer(name string, t *testing.T, indexer internal.Indexer) {
},
},
},
// Search for matches on the contents of files when the criteria is part of a expression.
// Search for matches on the contents of files when the criteria are parts of an expression.
{
RepoIDs: []int64{62},
Keyword: "log",
@ -235,16 +238,16 @@ 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(t.Context(), &internal.SearchOptions{
RepoIDs: kw.RepoIDs,
Keyword: kw.Keyword,
RepoIDs: kw.RepoIDs,
Keyword: kw.Keyword,
SearchMode: kw.SearchMode,
Paginator: &db.ListOptions{
Page: 1,
PageSize: 10,
},
IsKeywordFuzzy: true,
})
assert.NoError(t, err)
assert.Len(t, langs, kw.Langs)
require.NoError(t, err)
require.Len(t, langs, kw.Langs)
hits := make([]codeSearchResult, 0, len(res))
@ -289,7 +292,7 @@ func TestBleveIndexAndSearch(t *testing.T) {
_, err := idx.Init(t.Context())
require.NoError(t, err)
testIndexer("beleve", t, idx)
testIndexer("bleve", t, idx)
}
func TestESIndexAndSearch(t *testing.T) {