pub-solar/forgejo
5
0
Fork 0
Code Issues Pull requests Packages Projects Releases Wiki Activity

Add .gitattribute assisted language detection to blame, diff and render (#17590)

Browse source 
Use check attribute code to check the assigned language of a file and send that in to
chroma as a hint for the language of the file.

Signed-off-by: Andrew Thornton <art27@cantab.net>
This commit is contained in:
zeripath 2021-11-17 20:37:00 +00:00 committed by GitHub
parent 81a4fc7528
commit 3c4724d70e
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
13 changed files with 223 additions and 97 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

8
docs/content/doc/advanced/config-cheat-sheet.en-us.md
View file

@ -982,6 +982,14 @@ Multiple sanitisation rules can be defined by adding unique subsections, e.g. `[
To apply a sanitisation rules only for a specify external renderer they must use the renderer name, e.g. `[markup.sanitizer.asciidoc.rule-1]`.
If the rule is defined above the renderer ini section or the name does not match a renderer it is applied to every renderer.
## Highlight Mappings (`highlight.mapping`)
- `file_extension e.g. .toml`: **language e.g. ini**. File extension to language mapping overrides.
- Gitea will highlight files using the `linguist-language` or `gitlab-language` attribute from the `.gitattributes` file
if available. If this is not set or the language is unavailable, the file extension will be looked up
in this mapping or the filetype using heuristics.
## Time (`time`)
- `FORMAT`: Time format to display on UI. i.e. RFC1123 or 2006-01-02 15:04:05

17
modules/git/repo_attribute.go
View file

@ -22,6 +22,8 @@ type CheckAttributeOpts struct {
AllAttributes bool
Attributes []string
Filenames []string
IndexFile string
WorkTree string
}
// CheckAttribute return the Blame object of file
@ -31,6 +33,19 @@ func (repo *Repository) CheckAttribute(opts CheckAttributeOpts) (map[string]map[
return nil, fmt.Errorf("git version missing: %v", err)
}
env := []string{}
if len(opts.IndexFile) > 0 && CheckGitVersionAtLeast("1.7.8") == nil {
env = append(env, "GIT_INDEX_FILE="+opts.IndexFile)
}
if len(opts.WorkTree) > 0 && CheckGitVersionAtLeast("1.7.8") == nil {
env = append(env, "GIT_WORK_TREE="+opts.WorkTree)
}
if len(env) > 0 {
env = append(os.Environ(), env...)
}
stdOut := new(bytes.Buffer)
stdErr := new(bytes.Buffer)
@ -61,7 +76,7 @@ func (repo *Repository) CheckAttribute(opts CheckAttributeOpts) (map[string]map[
cmd := NewCommand(cmdArgs...)
if err := cmd.RunInDirPipeline(repo.Path, stdOut, stdErr); err != nil {
if err := cmd.RunInDirTimeoutEnvPipeline(env, -1, repo.Path, stdOut, stdErr); err != nil {
return nil, fmt.Errorf("failed to run check-attr: %v\n%s\n%s", err, stdOut.String(), stdErr.String())
}

12
modules/git/repo_index.go
View file

@ -8,6 +8,7 @@ import (
"bytes"
"context"
"os"
"path/filepath"
"strings"
"code.gitea.io/gitea/modules/log"
@ -45,14 +46,15 @@ func (repo *Repository) readTreeToIndex(id SHA1, indexFilename ...string) error
}
// ReadTreeToTemporaryIndex reads a treeish to a temporary index file
func (repo *Repository) ReadTreeToTemporaryIndex(treeish string) (filename string, cancel context.CancelFunc, err error) {
tmpIndex, err := os.CreateTemp("", "index")
func (repo *Repository) ReadTreeToTemporaryIndex(treeish string) (filename, tmpDir string, cancel context.CancelFunc, err error) {
tmpDir, err = os.MkdirTemp("", "index")
if err != nil {
return
}
filename = tmpIndex.Name()
filename = filepath.Join(tmpDir, ".tmp-index")
cancel = func() {
err := util.Remove(filename)
err := util.RemoveAll(tmpDir)
if err != nil {
log.Error("failed to remove tmp index file: %v", err)
}
@ -60,7 +62,7 @@ func (repo *Repository) ReadTreeToTemporaryIndex(treeish string) (filename strin
err = repo.ReadTreeToIndex(treeish, filename)
if err != nil {
defer cancel()
return "", func() {}, err
return "", "", func() {}, err
}
return
}

31
modules/git/repo_language_stats_gogit.go
View file

@ -11,11 +11,10 @@ import (
"bytes"
"context"
"io"
"os"
"strings"
"code.gitea.io/gitea/modules/analyze"
"code.gitea.io/gitea/modules/log"
"code.gitea.io/gitea/modules/util"
"github.com/go-enry/go-enry/v2"
"github.com/go-git/go-git/v5"
@ -48,20 +47,14 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
var checker *CheckAttributeReader
if CheckGitVersionAtLeast("1.7.8") == nil {
indexFilename, deleteTemporaryFile, err := repo.ReadTreeToTemporaryIndex(commitID)
indexFilename, workTree, deleteTemporaryFile, err := repo.ReadTreeToTemporaryIndex(commitID)
if err == nil {
defer deleteTemporaryFile()
tmpWorkTree, err := os.MkdirTemp("", "empty-work-dir")
if err == nil {
defer func() {
_ = util.RemoveAll(tmpWorkTree)
}()
checker = &CheckAttributeReader{
Attributes: []string{"linguist-vendored", "linguist-generated", "linguist-language"},
Attributes: []string{"linguist-vendored", "linguist-generated", "linguist-language", "gitlab-language"},
Repo: repo,
IndexFile: indexFilename,
WorkTree: tmpWorkTree,
WorkTree: workTree,
}
ctx, cancel := context.WithCancel(DefaultContext)
if err := checker.Init(ctx); err != nil {
@ -78,7 +71,6 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
defer cancel()
}
}
}
sizes := make(map[string]int64)
err = tree.Files().ForEach(func(f *object.File) error {
@ -114,6 +106,21 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
sizes[language] += f.Size
return nil
} else if language, has := attrs["gitlab-language"]; has && language != "unspecified" && language != "" {
// strip off a ? if present
if idx := strings.IndexByte(language, '?'); idx >= 0 {
language = language[:idx]
}
if len(language) != 0 {
// group languages, such as Pug -> HTML; SCSS -> CSS
group := enry.GetLanguageGroup(language)
if len(group) != 0 {
language = group
}
sizes[language] += f.Size
return nil
}
}
}
}

32
modules/git/repo_language_stats_nogogit.go
View file

@ -13,11 +13,10 @@ import (
"context"
"io"
"math"
"os"
"strings"
"code.gitea.io/gitea/modules/analyze"
"code.gitea.io/gitea/modules/log"
"code.gitea.io/gitea/modules/util"
"github.com/go-enry/go-enry/v2"
)
@ -68,20 +67,14 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
var checker *CheckAttributeReader
if CheckGitVersionAtLeast("1.7.8") == nil {
indexFilename, deleteTemporaryFile, err := repo.ReadTreeToTemporaryIndex(commitID)
indexFilename, worktree, deleteTemporaryFile, err := repo.ReadTreeToTemporaryIndex(commitID)
if err == nil {
defer deleteTemporaryFile()
tmpWorkTree, err := os.MkdirTemp("", "empty-work-dir")
if err == nil {
defer func() {
_ = util.RemoveAll(tmpWorkTree)
}()
checker = &CheckAttributeReader{
Attributes: []string{"linguist-vendored", "linguist-generated", "linguist-language"},
Attributes: []string{"linguist-vendored", "linguist-generated", "linguist-language", "gitlab-language"},
Repo: repo,
IndexFile: indexFilename,
WorkTree: tmpWorkTree,
WorkTree: worktree,
}
ctx, cancel := context.WithCancel(DefaultContext)
if err := checker.Init(ctx); err != nil {
@ -98,7 +91,6 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
defer cancel()
}
}
}
contentBuf := bytes.Buffer{}
var content []byte
@ -138,7 +130,23 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
sizes[language] += f.Size()
continue
} else if language, has := attrs["gitlab-language"]; has && language != "unspecified" && language != "" {
// strip off a ? if present
if idx := strings.IndexByte(language, '?'); idx >= 0 {
language = language[:idx]
}
if len(language) != 0 {
// group languages, such as Pug -> HTML; SCSS -> CSS
group := enry.GetLanguageGroup(language)
if len(group) != 0 {
language = group
}
sizes[language] += f.Size()
continue
}
}
}
}

26
modules/highlight/highlight.go
View file

@ -55,7 +55,7 @@ func NewContext() {
}
// Code returns a HTML version of code string with chroma syntax highlighting classes
func Code(fileName, code string) string {
func Code(fileName, language, code string) string {
NewContext()
// diff view newline will be passed as empty, change to literal \n so it can be copied
@ -69,10 +69,24 @@ func Code(fileName, code string) string {
}
var lexer chroma.Lexer
if len(language) > 0 {
lexer = lexers.Get(language)
if lexer == nil {
// Attempt stripping off the '?'
if idx := strings.IndexByte(language, '?'); idx > 0 {
lexer = lexers.Get(language[:idx])
}
}
}
if lexer == nil {
if val, ok := highlightMapping[filepath.Ext(fileName)]; ok {
//use mapped value to find lexer
lexer = lexers.Get(val)
}
}
if lexer == nil {
if l, ok := cache.Get(fileName); ok {
@ -119,7 +133,7 @@ func CodeFromLexer(lexer chroma.Lexer, code string) string {
}
// File returns a slice of chroma syntax highlighted lines of code
func File(numLines int, fileName string, code []byte) []string {
func File(numLines int, fileName, language string, code []byte) []string {
NewContext()
if len(code) > sizeLimit {
@ -139,9 +153,17 @@ func File(numLines int, fileName string, code []byte) []string {
htmlw := bufio.NewWriter(&htmlbuf)
var lexer chroma.Lexer
// provided language overrides everything
if len(language) > 0 {
lexer = lexers.Get(language)
}
if lexer == nil {
if val, ok := highlightMapping[filepath.Ext(fileName)]; ok {
lexer = lexers.Get(val)
}
}
if lexer == nil {
language := analyze.GetCodeLanguage(fileName, code)

2
modules/highlight/highlight_test.go
View file

@ -96,7 +96,7 @@ steps:
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
if got := File(tt.numLines, tt.fileName, []byte(tt.code)); !reflect.DeepEqual(got, tt.want) {
if got := File(tt.numLines, tt.fileName, "", []byte(tt.code)); !reflect.DeepEqual(got, tt.want) {
t.Errorf("File() = %v, want %v", got, tt.want)
}
})

2
modules/indexer/code/search.go
View file

@ -101,7 +101,7 @@ func searchResult(result *SearchResult, startIndex, endIndex int) (*Result, erro
Language: result.Language,
Color: result.Color,
LineNumbers: lineNumbers,
FormattedLines: highlight.Code(result.Filename, formattedLinesBuffer.String()),
FormattedLines: highlight.Code(result.Filename, "", formattedLinesBuffer.String()),
}, nil
}

13
modules/repofiles/diff_test.go
View file

@ -9,6 +9,7 @@ import (
"code.gitea.io/gitea/models"
"code.gitea.io/gitea/models/unittest"
"code.gitea.io/gitea/modules/json"
"code.gitea.io/gitea/modules/test"
"code.gitea.io/gitea/services/gitdiff"
@ -118,13 +119,21 @@ func TestGetDiffPreview(t *testing.T) {
t.Run("with given branch", func(t *testing.T) {
diff, err := GetDiffPreview(ctx.Repo.Repository, branch, treePath, content)
assert.NoError(t, err)
assert.EqualValues(t, expectedDiff, diff)
expectedBs, err := json.Marshal(expectedDiff)
assert.NoError(t, err)
bs, err := json.Marshal(diff)
assert.NoError(t, err)
assert.EqualValues(t, expectedBs, bs)
})
t.Run("empty branch, same results", func(t *testing.T) {
diff, err := GetDiffPreview(ctx.Repo.Repository, "", treePath, content)
assert.NoError(t, err)
assert.EqualValues(t, expectedDiff, diff)
expectedBs, err := json.Marshal(expectedDiff)
assert.NoError(t, err)
bs, err := json.Marshal(diff)
assert.NoError(t, err)
assert.EqualValues(t, expectedBs, bs)
})
}

28
routers/web/repo/blame.go
View file

@ -16,6 +16,7 @@ import (
"code.gitea.io/gitea/modules/context"
"code.gitea.io/gitea/modules/git"
"code.gitea.io/gitea/modules/highlight"
"code.gitea.io/gitea/modules/log"
"code.gitea.io/gitea/modules/templates"
"code.gitea.io/gitea/modules/timeutil"
"code.gitea.io/gitea/modules/util"
@ -204,6 +205,31 @@ func processBlameParts(ctx *context.Context, blameParts []git.BlamePart) (map[st
func renderBlame(ctx *context.Context, blameParts []git.BlamePart, commitNames map[string]*models.UserCommit, previousCommits map[string]string) {
repoLink := ctx.Repo.RepoLink
language := ""
indexFilename, worktree, deleteTemporaryFile, err := ctx.Repo.GitRepo.ReadTreeToTemporaryIndex(ctx.Repo.CommitID)
if err == nil {
defer deleteTemporaryFile()
filename2attribute2info, err := ctx.Repo.GitRepo.CheckAttribute(git.CheckAttributeOpts{
CachedOnly: true,
Attributes: []string{"linguist-language", "gitlab-language"},
Filenames: []string{ctx.Repo.TreePath},
IndexFile: indexFilename,
WorkTree: worktree,
})
if err != nil {
log.Error("Unable to load attributes for %-v:%s. Error: %v", ctx.Repo.Repository, ctx.Repo.TreePath, err)
}
language = filename2attribute2info[ctx.Repo.TreePath]["linguist-language"]
if language == "" || language == "unspecified" {
language = filename2attribute2info[ctx.Repo.TreePath]["gitlab-language"]
}
if language == "unspecified" {
language = ""
}
}
var lines = make([]string, 0)
rows := make([]*blameRow, 0)
@ -248,7 +274,7 @@ func renderBlame(ctx *context.Context, blameParts []git.BlamePart, commitNames m
line += "\n"
}
fileName := fmt.Sprintf("%v", ctx.Data["FileName"])
line = highlight.Code(fileName, line)
line = highlight.Code(fileName, language, line)
br.Code = gotemplate.HTML(line)
rows = append(rows, br)

28
routers/web/repo/view.go
View file

@ -502,7 +502,33 @@ func renderFile(ctx *context.Context, entry *git.TreeEntry, treeLink, rawLink st
lineNums := linesBytesCount(buf)
ctx.Data["NumLines"] = strconv.Itoa(lineNums)
ctx.Data["NumLinesSet"] = true
ctx.Data["FileContent"] = highlight.File(lineNums, blob.Name(), buf)
language := ""
indexFilename, worktree, deleteTemporaryFile, err := ctx.Repo.GitRepo.ReadTreeToTemporaryIndex(ctx.Repo.CommitID)
if err == nil {
defer deleteTemporaryFile()
filename2attribute2info, err := ctx.Repo.GitRepo.CheckAttribute(git.CheckAttributeOpts{
CachedOnly: true,
Attributes: []string{"linguist-language", "gitlab-language"},
Filenames: []string{ctx.Repo.TreePath},
IndexFile: indexFilename,
WorkTree: worktree,
})
if err != nil {
log.Error("Unable to load attributes for %-v:%s. Error: %v", ctx.Repo.Repository, ctx.Repo.TreePath, err)
}
language = filename2attribute2info[ctx.Repo.TreePath]["linguist-language"]
if language == "" || language == "unspecified" {
language = filename2attribute2info[ctx.Repo.TreePath]["gitlab-language"]
}
if language == "unspecified" {
language = ""
}
}
ctx.Data["FileContent"] = highlight.File(lineNums, blob.Name(), language, buf)
}
if !isLFSFile {
if ctx.Repo.CanEnableEditor() {

45
services/gitdiff/gitdiff.go
View file

@ -31,7 +31,6 @@ import (
"code.gitea.io/gitea/modules/log"
"code.gitea.io/gitea/modules/process"
"code.gitea.io/gitea/modules/setting"
"code.gitea.io/gitea/modules/util"
"github.com/sergi/go-diff/diffmatchpatch"
stdcharset "golang.org/x/net/html/charset"
@ -178,6 +177,7 @@ func getLineContent(content string) string {
// DiffSection represents a section of a DiffFile.
type DiffSection struct {
file *DiffFile
FileName string
Name string
Lines []*DiffLine
@ -546,6 +546,11 @@ func (diffSection *DiffSection) GetComputedInlineDiffFor(diffLine *DiffLine) tem
diff2 string
)
language := ""
if diffSection.file != nil {
language = diffSection.file.Language
}
// try to find equivalent diff line. ignore, otherwise
switch diffLine.Type {
case DiffLineSection:
@ -553,25 +558,25 @@ func (diffSection *DiffSection) GetComputedInlineDiffFor(diffLine *DiffLine) tem
case DiffLineAdd:
compareDiffLine = diffSection.GetLine(DiffLineDel, diffLine.RightIdx)
if compareDiffLine == nil {
return template.HTML(highlight.Code(diffSection.FileName, diffLine.Content[1:]))
return template.HTML(highlight.Code(diffSection.FileName, language, diffLine.Content[1:]))
}
diff1 = compareDiffLine.Content
diff2 = diffLine.Content
case DiffLineDel:
compareDiffLine = diffSection.GetLine(DiffLineAdd, diffLine.LeftIdx)
if compareDiffLine == nil {
return template.HTML(highlight.Code(diffSection.FileName, diffLine.Content[1:]))
return template.HTML(highlight.Code(diffSection.FileName, language, diffLine.Content[1:]))
}
diff1 = diffLine.Content
diff2 = compareDiffLine.Content
default:
if strings.IndexByte(" +-", diffLine.Content[0]) > -1 {
return template.HTML(highlight.Code(diffSection.FileName, diffLine.Content[1:]))
return template.HTML(highlight.Code(diffSection.FileName, language, diffLine.Content[1:]))
}
return template.HTML(highlight.Code(diffSection.FileName, diffLine.Content))
return template.HTML(highlight.Code(diffSection.FileName, language, diffLine.Content))
}
diffRecord := diffMatchPatch.DiffMain(highlight.Code(diffSection.FileName, diff1[1:]), highlight.Code(diffSection.FileName, diff2[1:]), true)
diffRecord := diffMatchPatch.DiffMain(highlight.Code(diffSection.FileName, language, diff1[1:]), highlight.Code(diffSection.FileName, language, diff2[1:]), true)
diffRecord = diffMatchPatch.DiffCleanupEfficiency(diffRecord)
return diffToHTML(diffSection.FileName, diffRecord, diffLine.Type)
@ -597,6 +602,7 @@ type DiffFile struct {
IsProtected bool
IsGenerated bool
IsVendored bool
Language string
}
// GetType returns type of diff file.
@ -1008,7 +1014,7 @@ func parseHunks(curFile *DiffFile, maxLines, maxLineCharacters int, input *bufio
line := sb.String()
// Create a new section to represent this hunk
curSection = &DiffSection{}
curSection = &DiffSection{file: curFile}
lastLeftIdx = -1
curFile.Sections = append(curFile.Sections, curSection)
@ -1048,7 +1054,7 @@ func parseHunks(curFile *DiffFile, maxLines, maxLineCharacters int, input *bufio
rightLine++
if curSection == nil {
// Create a new section to represent this hunk
curSection = &DiffSection{}
curSection = &DiffSection{file: curFile}
curFile.Sections = append(curFile.Sections, curSection)
lastLeftIdx = -1
}
@ -1074,7 +1080,7 @@ func parseHunks(curFile *DiffFile, maxLines, maxLineCharacters int, input *bufio
}
if curSection == nil {
// Create a new section to represent this hunk
curSection = &DiffSection{}
curSection = &DiffSection{file: curFile}
curFile.Sections = append(curFile.Sections, curSection)
lastLeftIdx = -1
}
@ -1094,7 +1100,7 @@ func parseHunks(curFile *DiffFile, maxLines, maxLineCharacters int, input *bufio
lastLeftIdx = -1
if curSection == nil {
// Create a new section to represent this hunk
curSection = &DiffSection{}
curSection = &DiffSection{file: curFile}
curFile.Sections = append(curFile.Sections, curSection)
}
curSection.Lines = append(curSection.Lines, diffLine)
@ -1302,23 +1308,15 @@ func GetDiffRangeWithWhitespaceBehavior(gitRepo *git.Repository, beforeCommitID,
var checker *git.CheckAttributeReader
if git.CheckGitVersionAtLeast("1.7.8") == nil {
indexFilename, deleteTemporaryFile, err := gitRepo.ReadTreeToTemporaryIndex(afterCommitID)
indexFilename, worktree, deleteTemporaryFile, err := gitRepo.ReadTreeToTemporaryIndex(afterCommitID)
if err == nil {
defer deleteTemporaryFile()
workdir, err := os.MkdirTemp("", "empty-work-dir")
if err != nil {
log.Error("Unable to create temporary directory: %v", err)
return nil, err
}
defer func() {
_ = util.RemoveAll(workdir)
}()
checker = &git.CheckAttributeReader{
Attributes: []string{"linguist-vendored", "linguist-generated"},
Attributes: []string{"linguist-vendored", "linguist-generated", "linguist-language", "gitlab-language"},
Repo: gitRepo,
IndexFile: indexFilename,
WorkTree: workdir,
WorkTree: worktree,
}
ctx, cancel := context.WithCancel(git.DefaultContext)
if err := checker.Init(ctx); err != nil {
@ -1361,6 +1359,11 @@ func GetDiffRangeWithWhitespaceBehavior(gitRepo *git.Repository, beforeCommitID,
gotGenerated = generated == "false"
}
}
if language, has := attrs["linguist-language"]; has && language != "unspecified" && language != "" {
diffFile.Language = language
} else if language, has := attrs["gitlab-language"]; has && language != "unspecified" && language != "" {
diffFile.Language = language
}
} else {
log.Error("Unexpected error: %v", err)
}

2
services/gitdiff/gitdiff_test.go
View file

@ -533,7 +533,7 @@ func TestGetDiffRangeWithWhitespaceBehavior(t *testing.T) {
func TestDiffToHTML_14231(t *testing.T) {
setting.Cfg = ini.Empty()
diffRecord := diffMatchPatch.DiffMain(highlight.Code("main.v", " run()\n"), highlight.Code("main.v", " run(db)\n"), true)
diffRecord := diffMatchPatch.DiffMain(highlight.Code("main.v", "", " run()\n"), highlight.Code("main.v", "", " run(db)\n"), true)
diffRecord = diffMatchPatch.DiffCleanupEfficiency(diffRecord)
expected := ` <span class="n">run</span><span class="added-code"><span class="o">(</span><span class="n">db</span></span><span class="o">)</span>`