refactor: decouple git.Repository(ctx) from git.Commit & git.Tree (#38464)

1. Storing "ctx" in a long-living object is wrong
2. Make the commit & tree cacheable (for the future performance
optimization)
3. Also fix some bad designs like `// FIXME: bad design, this field can
be nil if the commit is from "last commit cache"`

ref:
* #33893
This commit is contained in:
wxiaoguang
2026-07-16 01:30:01 +08:00
committed by GitHub
parent ed678b9d45
commit 82edc3da01
115 changed files with 732 additions and 864 deletions
@@ -7,6 +7,7 @@ package languagestats
import (
"bytes"
"context"
"io"
"gitea.dev/modules/analyze"
@@ -21,7 +22,7 @@ import (
)
// GetLanguageStats calculates language stats for git repository at specified commit
func GetLanguageStats(repo *git_module.Repository, commitID string) (map[string]int64, error) {
func GetLanguageStats(_ context.Context, repo *git_module.Repository, commitID string) (map[string]int64, error) {
r, err := git.PlainOpen(repo.Path)
if err != nil {
return nil, err
@@ -7,6 +7,7 @@ package languagestats
import (
"bytes"
"context"
"io"
"gitea.dev/modules/analyze"
@@ -19,10 +20,10 @@ import (
)
// GetLanguageStats calculates language stats for git repository at specified commit
func GetLanguageStats(repo *git.Repository, commitID string) (map[string]int64, error) {
func GetLanguageStats(ctx context.Context, repo *git.Repository, commitID string) (map[string]int64, error) {
// We will feed the commit IDs in order into cat-file --batch, followed by blobs as necessary.
// so let's create a batch stdin and stdout
batch, cancel, err := repo.CatFileBatch(repo.Ctx)
batch, cancel, err := repo.CatFileBatch(ctx)
if err != nil {
return nil, err
}
@@ -43,7 +44,7 @@ func GetLanguageStats(repo *git.Repository, commitID string) (map[string]int64,
return nil, git.ErrNotExist{ID: commitID}
}
commit, err := git.CommitFromReader(repo, sha, io.LimitReader(batchReader, commitInfo.Size))
commit, err := git.CommitFromReader(sha, io.LimitReader(batchReader, commitInfo.Size))
if err != nil {
log.Debug("Unable to get commit for: %s. Err: %v", commitID, err)
return nil, err
@@ -52,9 +53,7 @@ func GetLanguageStats(repo *git.Repository, commitID string) (map[string]int64,
return nil, err
}
tree := commit.Tree
entries, err := tree.ListEntriesRecursiveWithSize()
entries, err := commit.Tree().ListEntriesRecursiveWithSize(ctx, repo)
if err != nil {
return nil, err
}
@@ -81,13 +80,15 @@ func GetLanguageStats(repo *git.Repository, commitID string) (map[string]int64,
select {
case <-repo.Ctx.Done():
return sizes, repo.Ctx.Err()
case <-ctx.Done():
return sizes, ctx.Err()
default:
}
contentBuf.Reset()
content = contentBuf.Bytes()
if f.Size() == 0 {
if f.GetSize(ctx, repo) == 0 {
continue
}
@@ -124,7 +125,7 @@ func GetLanguageStats(repo *git.Repository, commitID string) (map[string]int64,
}
// this language will always be added to the size
sizes[language] += f.Size()
sizes[language] += f.GetSize(ctx, repo)
continue
}
}
@@ -138,7 +139,7 @@ func GetLanguageStats(repo *git.Repository, commitID string) (map[string]int64,
// If content can not be read or file is too big just do detection by filename
if f.Size() <= bigFileSize {
if f.GetSize(ctx, repo) <= bigFileSize {
info, _, err := batch.QueryContent(f.ID.String())
if err != nil {
return nil, err
@@ -192,10 +193,10 @@ func GetLanguageStats(repo *git.Repository, commitID string) (map[string]int64,
includedLanguage[language] = included
}
if included || isDetectable.ValueOrDefault(false) {
sizes[language] += f.Size()
sizes[language] += f.GetSize(ctx, repo)
} else if len(sizes) == 0 && (firstExcludedLanguage == "" || firstExcludedLanguage == language) {
firstExcludedLanguage = language
firstExcludedLanguageSize += f.Size()
firstExcludedLanguageSize += f.GetSize(ctx, repo)
}
}
@@ -22,7 +22,7 @@ func TestRepository_GetLanguageStats(t *testing.T) {
require.NoError(t, err)
defer gitRepo.Close()
stats, err := GetLanguageStats(gitRepo, "8fee858da5796dfb37704761701bb8e800ad9ef3")
stats, err := GetLanguageStats(t.Context(), gitRepo, "8fee858da5796dfb37704761701bb8e800ad9ef3")
require.NoError(t, err)
assert.Equal(t, map[string]int64{