Ignore Sync errors on pipes when doing CheckAttributeReader.CheckPath, fix the hang of git cat-file

modules/git/repo_attribute.go

@@ -12,6 +12,8 @@ import (
 	"os"
 	"strconv"
 	"strings"
+
+	"code.gitea.io/gitea/modules/log"
 )
 
 // CheckAttributeOpts represents the possible options to CheckAttribute
@@ -112,56 +114,63 @@ func (c *CheckAttributeReader) Init(ctx context.Context) error {
 
 	if len(c.IndexFile) > 0 && CheckGitVersionAtLeast("1.7.8") == nil {
 		cmdArgs = append(cmdArgs, "--cached")
-		c.env = []string{"GIT_INDEX_FILE=" + c.IndexFile}
+		c.env = append(c.env, "GIT_INDEX_FILE="+c.IndexFile)
 	}
 
 	if len(c.WorkTree) > 0 && CheckGitVersionAtLeast("1.7.8") == nil {
-		c.env = []string{"GIT_WORK_TREE=" + c.WorkTree}
+		c.env = append(c.env, "GIT_WORK_TREE="+c.WorkTree)
 	}
 
-	if len(c.Attributes) > 0 {
-		cmdArgs = append(cmdArgs, c.Attributes...)
-		cmdArgs = append(cmdArgs, "--")
-	} else {
+	c.env = append(c.env, "GIT_FLUSH=1")
+
+	if len(c.Attributes) == 0 {
 		lw := new(nulSeparatedAttributeWriter)
 		lw.attributes = make(chan attributeTriple)
+		lw.closed = make(chan struct{})
 
 		c.stdOut = lw
 		c.stdOut.Close()
 		return fmt.Errorf("no provided Attributes to check")
 	}
 
+	cmdArgs = append(cmdArgs, c.Attributes...)
+	cmdArgs = append(cmdArgs, "--")
+
 	c.ctx, c.cancel = context.WithCancel(ctx)
 	c.cmd = NewCommandContext(c.ctx, cmdArgs...)
+
 	var err error
+
 	c.stdinReader, c.stdinWriter, err = os.Pipe()
 	if err != nil {
+		c.cancel()
 		return err
 	}
 
 	if CheckGitVersionAtLeast("1.8.5") == nil {
 		lw := new(nulSeparatedAttributeWriter)
 		lw.attributes = make(chan attributeTriple, 5)
-
+		lw.closed = make(chan struct{})
 		c.stdOut = lw
 	} else {
 		lw := new(lineSeparatedAttributeWriter)
 		lw.attributes = make(chan attributeTriple, 5)
-
+		lw.closed = make(chan struct{})
 		c.stdOut = lw
 	}
 	return nil
 }
 
 // Run run cmd
 func (c *CheckAttributeReader) Run() error {
+	defer func() {
+		_ = c.Close()
+	}()
 	stdErr := new(bytes.Buffer)
 	err := c.cmd.RunInDirTimeoutEnvFullPipelineFunc(c.env, -1, c.Repo.Path, c.stdOut, stdErr, c.stdinReader, func(_ context.Context, _ context.CancelFunc) error {
 		close(c.running)
 		return nil
 	})
-	defer c.cancel()
-	_ = c.stdOut.Close()
 	if err != nil && c.ctx.Err() != nil && err.Error() != "signal: killed" {
 		return fmt.Errorf("failed to run attr-check. Error: %w\nStderr: %s", err, stdErr.String())
 	}
@@ -170,27 +179,31 @@ func (c *CheckAttributeReader) Run() error {
 }
 
 // CheckPath check attr for given path
-func (c *CheckAttributeReader) CheckPath(path string) (map[string]string, error) {
+func (c *CheckAttributeReader) CheckPath(path string) (rs map[string]string, err error) {
+	defer func() {
+		if err != nil {
+			log.Error("CheckPath returns error: %v", err)
+		}
+	}()
+
 	select {
 	case <-c.ctx.Done():
 		return nil, c.ctx.Err()
 	case <-c.running:
 	}
 
-	if _, err := c.stdinWriter.Write([]byte(path + "\x00")); err != nil {
-		defer c.cancel()
+	if _, err = c.stdinWriter.Write([]byte(path + "\x00")); err != nil {
+		defer c.Close()
 		return nil, err
 	}
 
-	if err := c.stdinWriter.Sync(); err != nil {
-		defer c.cancel()
-		return nil, err
-	}
-
-	rs := make(map[string]string)
+	rs = make(map[string]string)
 	for range c.Attributes {
 		select {
-		case attr := <-c.stdOut.ReadAttribute():
+		case attr, ok := <-c.stdOut.ReadAttribute():
+			if !ok {
+				return nil, c.ctx.Err()
+			}
 			rs[attr.Attribute] = attr.Value
 		case <-c.ctx.Done():
 			return nil, c.ctx.Err()
@@ -201,13 +214,16 @@ func (c *CheckAttributeReader) CheckPath(path string) (map[string]string, error)
 
 // Close close pip after use
 func (c *CheckAttributeReader) Close() error {
+	err := c.stdinWriter.Close()
+	_ = c.stdinReader.Close()
+	_ = c.stdOut.Close()
+	c.cancel()
 	select {
 	case <-c.running:
 	default:
 		close(c.running)
 	}
-	defer c.cancel()
-	return c.stdinWriter.Close()
+	return err
 }
 
 type attributeWriter interface {
@@ -224,6 +240,7 @@ type attributeTriple struct {
 type nulSeparatedAttributeWriter struct {
 	tmp        []byte
 	attributes chan attributeTriple
+	closed     chan struct{}
 	working    attributeTriple
 	pos        int
 }
@@ -267,13 +284,20 @@ func (wr *nulSeparatedAttributeWriter) ReadAttribute() <-chan attributeTriple {
 }
 
 func (wr *nulSeparatedAttributeWriter) Close() error {
+	select {
+	case <-wr.closed:
+		return nil
+	default:
+	}
 	close(wr.attributes)
+	close(wr.closed)
 	return nil
 }
 
 type lineSeparatedAttributeWriter struct {
 	tmp        []byte
 	attributes chan attributeTriple
+	closed     chan struct{}
 }
 
 func (wr *lineSeparatedAttributeWriter) Write(p []byte) (n int, err error) {
@@ -356,6 +380,12 @@ func (wr *lineSeparatedAttributeWriter) ReadAttribute() <-chan attributeTriple {
 }
 
 func (wr *lineSeparatedAttributeWriter) Close() error {
+	select {
+	case <-wr.closed:
+		return nil
+	default:
+	}
 	close(wr.attributes)
+	close(wr.closed)
 	return nil
 }

modules/git/repo_language_stats_gogit.go

@@ -15,6 +15,7 @@ import (
 
 	"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"
@@ -50,25 +51,32 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
 		indexFilename, deleteTemporaryFile, err := repo.ReadTreeToTemporaryIndex(commitID)
 		if err == nil {
 			defer deleteTemporaryFile()
-
-			checker = &CheckAttributeReader{
-				Attributes: []string{"linguist-vendored", "linguist-generated", "linguist-language"},
-				Repo:       repo,
-				IndexFile:  indexFilename,
-			}
-			ctx, cancel := context.WithCancel(DefaultContext)
-			if err := checker.Init(ctx); err != nil {
-				log.Error("Unable to open checker for %s. Error: %v", commitID, err)
-			} else {
-				go func() {
-					err = checker.Run()
-					if err != nil {
-						log.Error("Unable to open checker for %s. Error: %v", commitID, err)
-						cancel()
-					}
+			tmpWorkTree, err := ioutil.TempDir("", "empty-work-dir")
+			if err == nil {
+				defer func() {
+					_ = util.RemoveAll(tmpWorkTree)
 				}()
+
+				checker = &CheckAttributeReader{
+					Attributes: []string{"linguist-vendored", "linguist-generated", "linguist-language"},
+					Repo:       repo,
+					IndexFile:  indexFilename,
+					WorkTree:   tmpWorkTree,
+				}
+				ctx, cancel := context.WithCancel(DefaultContext)
+				if err := checker.Init(ctx); err != nil {
+					log.Error("Unable to open checker for %s. Error: %v", commitID, err)
+				} else {
+					go func() {
+						err = checker.Run()
+						if err != nil {
+							log.Error("Unable to open checker for %s. Error: %v", commitID, err)
+							cancel()
+						}
+					}()
+				}
+				defer cancel()
 			}
-			defer cancel()
 		}
 	}
 
@@ -99,7 +107,7 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
 				if language, has := attrs["linguist-language"]; has && language != "unspecified" && language != "" {
 					// group languages, such as Pug -> HTML; SCSS -> CSS
 					group := enry.GetLanguageGroup(language)
-					if len(group) == 0 {
+					if len(group) != 0 {
 						language = group
 					}
 

modules/git/repo_language_stats_nogogit.go

@@ -12,10 +12,12 @@ import (
 	"bytes"
 	"context"
 	"io"
+	"io/ioutil"
 	"math"
 
 	"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"
 )
@@ -69,25 +71,32 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
 		indexFilename, deleteTemporaryFile, err := repo.ReadTreeToTemporaryIndex(commitID)
 		if err == nil {
 			defer deleteTemporaryFile()
-
-			checker = &CheckAttributeReader{
-				Attributes: []string{"linguist-vendored", "linguist-generated", "linguist-language"},
-				Repo:       repo,
-				IndexFile:  indexFilename,
-			}
-			ctx, cancel := context.WithCancel(DefaultContext)
-			if err := checker.Init(ctx); err != nil {
-				log.Error("Unable to open checker for %s. Error: %v", commitID, err)
-			} else {
-				go func() {
-					err = checker.Run()
-					if err != nil {
-						log.Error("Unable to open checker for %s. Error: %v", commitID, err)
-						cancel()
-					}
+			tmpWorkTree, err := ioutil.TempDir("", "empty-work-dir")
+			if err == nil {
+				defer func() {
+					_ = util.RemoveAll(tmpWorkTree)
 				}()
+
+				checker = &CheckAttributeReader{
+					Attributes: []string{"linguist-vendored", "linguist-generated", "linguist-language"},
+					Repo:       repo,
+					IndexFile:  indexFilename,
+					WorkTree:   tmpWorkTree,
+				}
+				ctx, cancel := context.WithCancel(DefaultContext)
+				if err := checker.Init(ctx); err != nil {
+					log.Error("Unable to open checker for %s. Error: %v", commitID, err)
+				} else {
+					go func() {
+						err = checker.Run()
+						if err != nil {
+							log.Error("Unable to open checker for %s. Error: %v", commitID, err)
+							cancel()
+						}
+					}()
+				}
+				defer cancel()
 			}
-			defer cancel()
 		}
 	}
 
@@ -123,12 +132,11 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
 				if language, has := attrs["linguist-language"]; has && language != "unspecified" && language != "" {
 					// group languages, such as Pug -> HTML; SCSS -> CSS
 					group := enry.GetLanguageGroup(language)
-					if len(group) == 0 {
+					if len(group) != 0 {
 						language = group
 					}
 
 					sizes[language] += f.Size()
-
 					continue
 				}
 			}
@@ -186,7 +194,6 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
 		}
 
 		sizes[language] += f.Size()
-
 		continue
 	}
 

modules/git/repo_language_stats_test.go

@@ -0,0 +1,34 @@
+// Copyright 2020 The Gitea Authors. All rights reserved.
+// Use of this source code is governed by a MIT-style
+// license that can be found in the LICENSE file.
+
+//go:build !gogit
+// +build !gogit
+
+package git
+
+import (
+	"path/filepath"
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+)
+
+func TestRepository_GetLanguageStats(t *testing.T) {
+	repoPath := filepath.Join(testReposDir, "language_stats_repo")
+	gitRepo, err := OpenRepository(repoPath)
+	if !assert.NoError(t, err) {
+		t.Fatal()
+	}
+	defer gitRepo.Close()
+
+	stats, err := gitRepo.GetLanguageStats("8fee858da5796dfb37704761701bb8e800ad9ef3")
+	if !assert.NoError(t, err) {
+		t.Fatal()
+	}
+
+	assert.EqualValues(t, map[string]int64{
+		"Python": 134,
+		"Java":   112,
+	}, stats)
+}

modules/git/tests/repos/language_stats_repo/COMMIT_EDITMSG

@@ -0,0 +1,3 @@
+Add some test files for GetLanguageStats
+
+Signed-off-by: Andrew Thornton <[email protected]>

modules/git/tests/repos/language_stats_repo/HEAD

@@ -0,0 +1 @@
+ref: refs/heads/master

modules/git/tests/repos/language_stats_repo/config

@@ -0,0 +1,5 @@
+[core]
+	repositoryformatversion = 0
+	filemode = true
+	bare = false
+	logallrefupdates = true

modules/git/tests/repos/language_stats_repo/description

@@ -0,0 +1 @@
+Unnamed repository; edit this file 'description' to name the repository.