Add go wrapper around git diff-tree --raw -r -M

McRaeAlex · McRaeAlex · commit da24f58a5d5d · 2025-01-24T15:40:58.000-08:00
* Implemented calling git diff-tree
 * Ensures wrapper function is called with valid arguments
 * Parses output into go struct, using strong typing when possible
diff --git a/modules/git/parse.go b/modules/git/parse.go
@@ -46,19 +46,9 @@ func parseLsTreeLine(line []byte) (*LsTreeEntry, error) {
 		entry.Size = optional.Some(size)
 	}
 
-	switch string(entryMode) {
-	case "100644":
-		entry.EntryMode = EntryModeBlob
-	case "100755":
-		entry.EntryMode = EntryModeExec
-	case "120000":
-		entry.EntryMode = EntryModeSymlink
-	case "160000":
-		entry.EntryMode = EntryModeCommit
-	case "040000", "040755": // git uses 040000 for tree object, but some users may get 040755 for unknown reasons
-		entry.EntryMode = EntryModeTree
-	default:
-		return nil, fmt.Errorf("unknown type: %v", string(entryMode))
+	entry.EntryMode, err = ParseEntryMode(string(entryMode))
+	if err != nil || entry.EntryMode == EntryModeNoEntry {
+		return nil, fmt.Errorf("invalid ls-tree output (invalid mode): %q, err: %w", line, err)
 	}
 
 	entry.ID, err = NewIDFromString(string(entryObjectID))
diff --git a/modules/git/tree_entry_mode.go b/modules/git/tree_entry_mode.go
@@ -3,14 +3,20 @@
 
 package git
 
-import "strconv"
+import (
+	"fmt"
+	"strconv"
+)
 
 // EntryMode the type of the object in the git tree
 type EntryMode int
 
 // There are only a few file modes in Git. They look like unix file modes, but they can only be
 // one of these.
 const (
+	// EntryModeNoEntry is possible if the file was added or removed in a commit. In the case of
+	// added the base commit will not have the file in its tree so a mode of 0o000000 is used.
+	EntryModeNoEntry EntryMode = 0o000000
 	// EntryModeBlob
 	EntryModeBlob EntryMode = 0o100644
 	// EntryModeExec
@@ -33,3 +39,22 @@ func ToEntryMode(value string) EntryMode {
 	v, _ := strconv.ParseInt(value, 8, 32)
 	return EntryMode(v)
 }
+
+func ParseEntryMode(mode string) (EntryMode, error) {
+	switch mode {
+	case "000000":
+		return EntryModeNoEntry, nil
+	case "100644":
+		return EntryModeBlob, nil
+	case "100755":
+		return EntryModeExec, nil
+	case "120000":
+		return EntryModeSymlink, nil
+	case "160000":
+		return EntryModeCommit, nil
+	case "040000", "040755": // git uses 040000 for tree object, but some users may get 040755 for unknown reasons
+		return EntryModeTree, nil
+	default:
+		return 0, fmt.Errorf("unparsable entry mode: %s", mode)
+	}
+}
diff --git a/services/gitdiff/git_diff_tree.go b/services/gitdiff/git_diff_tree.go
@@ -0,0 +1,219 @@
+// Copyright 2025 The Gitea Authors. All rights reserved.
+// SPDX-License-Identifier: MIT
+
+package gitdiff
+
+import (
+	"bufio"
+	"context"
+	"fmt"
+	"strings"
+
+	"code.gitea.io/gitea/modules/git"
+	"code.gitea.io/gitea/modules/log"
+)
+
+type DiffTree struct {
+	Files []*DiffTreeRecord
+}
+
+type DiffTreeRecord struct {
+	// Status is one of 'added', 'deleted', 'modified', 'renamed', 'copied'
+	Status string
+
+	HeadPath   string
+	BasePath   string
+	HeadMode   git.EntryMode
+	BaseMode   git.EntryMode
+	HeadBlobID string
+	BaseBlobID string
+}
+
+// GetDiffTree returns the list of path of the files that have changed between the two commits
+func GetDiffTree(ctx context.Context, gitRepo *git.Repository, baseSha, headSha string) (*DiffTree, error) {
+	gitDiffTreeRecords, err := runGitDiffTree(ctx, gitRepo, baseSha, headSha)
+	if err != nil {
+		return nil, err
+	}
+
+	return &DiffTree{
+		Files: gitDiffTreeRecords,
+	}, nil
+}
+
+func runGitDiffTree(ctx context.Context, gitRepo *git.Repository, baseSha, headSha string) ([]*DiffTreeRecord, error) {
+	baseCommitID, headCommitID, err := validateGitDiffTreeArguments(gitRepo, baseSha, headSha)
+	if err != nil {
+		return nil, err
+	}
+
+	cmd := git.NewCommand(ctx, "diff-tree", "--raw", "-r", "--find-renames").AddDynamicArguments(baseCommitID, headCommitID)
+	stdout, _, runErr := cmd.RunStdString(&git.RunOpts{Dir: gitRepo.Path})
+	if runErr != nil {
+		log.Warn("git diff-tree: %v", runErr)
+		return nil, runErr
+	}
+
+	return parseGitDiffTree(stdout)
+}
+
+func validateGitDiffTreeArguments(gitRepo *git.Repository, baseSha, headSha string) (string, string, error) {
+	// if the head is empty its an error
+	if headSha == "" {
+		return "", "", fmt.Errorf("headSha is empty")
+	}
+
+	// if the head commit doesn't exist its and error
+	headCommit, err := gitRepo.GetCommit(headSha)
+	if err != nil {
+		return "", "", fmt.Errorf("failed to get commit headSha: %v", err)
+	}
+	headCommitID := headCommit.ID.String()
+
+	// if the base is empty we should use the parent of the head commit
+	if baseSha == "" {
+		// if the headCommit has no parent we should use an empty commit
+		// this can happen when we are generating a diff against an orphaned commit
+		if headCommit.ParentCount() == 0 {
+			objectFormat, err := gitRepo.GetObjectFormat()
+			if err != nil {
+				return "", "", err
+			}
+
+			return objectFormat.EmptyTree().String(), headCommitID, nil
+		}
+
+		baseCommit, err := headCommit.Parent(0)
+		if err != nil {
+			return "", "", fmt.Errorf("baseSha is '', attempted to use parent of commit %s, got error: %v", headCommit.ID.String(), err)
+		}
+		return baseCommit.ID.String(), headCommitID, nil
+	}
+
+	// try and get the base commit
+	baseCommit, err := gitRepo.GetCommit(baseSha)
+	// propagate the error if we couldn't get the base commit
+	if err != nil {
+		return "", "", fmt.Errorf("failed to get base commit %s: %v", baseSha, err)
+	}
+
+	return baseCommit.ID.String(), headCommit.ID.String(), nil
+}
+
+func parseGitDiffTree(output string) ([]*DiffTreeRecord, error) {
+	/*
+		The output of `git diff-tree --raw -r --find-renames` is of the form:
+
+		:<old_mode> <new_mode> <old_sha> <new_sha> <status>\t<path>
+
+		or for renames:
+
+		:<old_mode> <new_mode> <old_sha> <new_sha> <status>\t<old_path>\t<new_path>
+
+		See: <https://git-scm.com/docs/git-diff-tree#_raw_output_format> for more details
+	*/
+	if output == "" {
+		return []*DiffTreeRecord{}, nil
+	}
+
+	results := make([]*DiffTreeRecord, 0)
+
+	lines := bufio.NewScanner(strings.NewReader(output))
+	for lines.Scan() {
+		line := lines.Text()
+
+		if len(line) == 0 {
+			continue
+		}
+
+		record, err := parseGitDiffTreeLine(line)
+		if err != nil {
+			return nil, err
+		}
+
+		results = append(results, record)
+	}
+
+	if err := lines.Err(); err != nil {
+		return nil, err
+	}
+
+	return results, nil
+}
+
+func parseGitDiffTreeLine(line string) (*DiffTreeRecord, error) {
+	line = strings.TrimPrefix(line, ":")
+	splitSections := strings.SplitN(line, "\t", 2)
+	if len(splitSections) < 2 {
+		return nil, fmt.Errorf("unparsable output for diff --raw: `%s`)", line)
+	}
+
+	fields := strings.Fields(splitSections[0])
+	if len(fields) < 5 {
+		return nil, fmt.Errorf("unparsable output for diff --raw: `%s`, expected 5 space delimited values got %d)", line, len(fields))
+	}
+
+	baseMode, err := git.ParseEntryMode(fields[0])
+	if err != nil {
+		return nil, err
+	}
+
+	headMode, err := git.ParseEntryMode(fields[1])
+	if err != nil {
+		return nil, err
+	}
+
+	baseBlobID := fields[2]
+	headBlobID := fields[3]
+
+	status, err := statusFromLetter(fields[4])
+	if err != nil {
+		return nil, err
+	}
+
+	filePaths := strings.Split(splitSections[1], "\t")
+
+	var headPath, basePath string
+	if status == "renamed" {
+		if len(filePaths) != 2 {
+			return nil, fmt.Errorf("unparsable output for diff --raw: `%s`, expected 2 paths found %d", line, len(filePaths))
+		}
+		basePath = filePaths[0]
+		headPath = filePaths[1]
+	} else {
+		basePath = filePaths[0]
+		headPath = filePaths[0]
+	}
+
+	return &DiffTreeRecord{
+		Status:     status,
+		BaseMode:   baseMode,
+		HeadMode:   headMode,
+		BaseBlobID: baseBlobID,
+		HeadBlobID: headBlobID,
+		BasePath:   basePath,
+		HeadPath:   headPath,
+	}, nil
+}
+
+func statusFromLetter(letter string) (string, error) {
+	if len(letter) < 1 {
+		return "", fmt.Errorf("empty status letter")
+	}
+	switch letter[0] {
+	case 'A':
+		return "added", nil
+	case 'D':
+		return "deleted", nil
+	case 'M':
+		return "modified", nil
+	case 'R':
+		// This is of the form "R<score>" but we are choosing to ignore the score
+		return "renamed", nil
+	case 'C':
+		// This is of the form "C<score>" but we are choosing to ignore the score
+		return "copied", nil
+	default:
+		return "", fmt.Errorf("unknown status letter: '%s'", letter)
+	}
+}
diff --git a/services/gitdiff/git_diff_tree_test.go b/services/gitdiff/git_diff_tree_test.go
