Skip to content

feat: add path filtering for commits script #19231

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
just-mitch wants to merge 1 commit into
base: next
Choose a base branch
from
Open

feat: add path filtering for commits script #19231

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
117 changes: 116 additions & 1 deletion scripts/commits
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,6 +7,7 @@ import os
from datetime import datetime
from typing import List, Tuple, Optional
import argparse
from fnmatch import fnmatch

# ANSI color codes
RESET = '\033[0m'
Expand Down Expand Up @@ -147,6 +148,82 @@ def visible_length(text: str) -> int:
text = re.sub(r'\033\]8[^\\]*\\', '', text)
return len(text)

def path_matches(file_path: str, pattern: str) -> bool:
"""Check if a file path matches a pattern (with glob and directory support)."""
# If the pattern contains glob characters, use fnmatch
if '*' in pattern or '?' in pattern or '[' in pattern:
return fnmatch(file_path, pattern)

# Otherwise, treat it as a directory/file prefix
# Match exact file or anything under that directory
if file_path == pattern:
return True
if file_path.startswith(pattern + '/'):
return True

return False

def commit_affects_paths(commit: str, paths: List[str], debug: bool = False) -> bool:
"""Check if a commit affects any of the specified paths (with glob support)."""
if not paths:
return True

# Get list of files affected by this commit
# Use -m to show files for merge commits, --first-parent to compare against first parent
files_cmd = ['git', 'diff-tree', '--no-commit-id', '--name-only', '-r', '-m', '--first-parent', commit]
files_output = run_command(files_cmd)

if not files_output:
return False

affected_files = files_output.split('\n')

# Check if any affected file matches any of the path patterns
for affected_file in affected_files:
if not affected_file:
continue
for path_pattern in paths:
if path_matches(affected_file, path_pattern):
if debug:
print(f"DEBUG: -> INCLUDED: file '{affected_file}' matches include pattern '{path_pattern}'", file=sys.stderr)
return True

if debug:
print(f"DEBUG: -> NOT INCLUDED: no files match include patterns", file=sys.stderr)
return False

def commit_touches_excluded_paths(commit: str, exclude_paths: List[str], debug: bool = False) -> bool:
"""Check if a commit touches any of the excluded paths (with glob support)."""
if not exclude_paths:
return False

# Get list of files affected by this commit
# Use -m to show files for merge commits, --first-parent to compare against first parent
files_cmd = ['git', 'diff-tree', '--no-commit-id', '--name-only', '-r', '-m', '--first-parent', commit]
files_output = run_command(files_cmd)

if not files_output:
return False

affected_files = files_output.split('\n')

if debug and affected_files:
non_empty_files = [f for f in affected_files if f]
if non_empty_files:
print(f"DEBUG: Commit {commit[:7]} touches files: {non_empty_files[:5]}{'...' if len(non_empty_files) > 5 else ''}", file=sys.stderr)

# Check if any affected file matches any of the excluded path patterns
for affected_file in affected_files:
if not affected_file:
continue
for path_pattern in exclude_paths:
if path_matches(affected_file, path_pattern):
if debug:
print(f"DEBUG: -> EXCLUDED: file '{affected_file}' matches pattern '{path_pattern}'", file=sys.stderr)
return True

return False

def get_merge_train_commits(merge_commit: str) -> List[str]:
"""Get commits in a merge train."""
# Get the first parent (the commit this was merged into)
Expand Down Expand Up @@ -256,6 +333,9 @@ def echo_header(text: str) -> None:
print(f"{PURPLE}---{RESET} {BLUE}{BOLD}{text}{RESET} {PURPLE}---{RESET}")

def main():
# Check for debug mode
debug_mode = os.environ.get('DEBUG_COMMITS', '').lower() in ('1', 'true', 'yes')

# Parse arguments with argparse to support ranges and grep filtering
parser = argparse.ArgumentParser(
description=(
Expand All @@ -274,6 +354,13 @@ Examples:
{parser.prog} --grep '^feat' -m Markdown output of only feats on HEAD
{parser.prog} --no-first-parent main Do not restrict to first-parent history
{parser.prog} v2.0.0..HEAD --grep 'bb|barretenberg' Filter to barretenberg-related changes
{parser.prog} --paths 'src/main.rs' Show commits affecting specific file
{parser.prog} --paths 'src/*.rs' --paths 'Cargo.toml' Multiple paths with globs
{parser.prog} --grep 'feat' --paths 'backend/**' Combine grep and path filters
{parser.prog} v1.0..v2.0 --paths 'docs/**' Commits in range affecting docs
{parser.prog} --exclude-paths 'docs/**' Exclude commits that touch docs
{parser.prog} --exclude-paths '*.md' --exclude-paths 'test/**' Exclude multiple paths
{parser.prog} --paths 'src/**' --exclude-paths 'src/test/**' Include src/ but exclude tests
{parser.prog} -g Group commits by type on HEAD
{parser.prog} -g -m v1.2.0..v2.0.0 Group by type in Markdown for a range
{parser.prog} -g --grep 'feat|fix' main Group by type but only features/fixes
Expand All @@ -282,6 +369,8 @@ Examples:
parser.add_argument('ref', nargs='?', default='HEAD', help='Branch, commit, or range (e.g., v1.2.1..v2.0.1)')
parser.add_argument('limit', nargs='?', type=int, default=50, help='Max commits to inspect (applies to top-level history)')
parser.add_argument('--grep', dest='grep_patterns', action='append', default=[], help='Regex to match commit subjects. Can be repeated.')
parser.add_argument('--paths', dest='file_paths', action='append', default=[], help='File paths or glob patterns to filter commits. Can be repeated.')
parser.add_argument('--exclude-paths', dest='exclude_paths', action='append', default=[], help='File paths or glob patterns to exclude commits. Can be repeated.')
parser.add_argument('--no-first-parent', action='store_true', help='Do not restrict to first-parent history')
parser.add_argument('--markdown', '-m', action='store_true', help='Print output formatted as Markdown')
parser.add_argument('--group-by-type', '-g', action='store_true', help='Group commits by conventional type (feat, fix, refactor, chore, etc.)')
Expand All @@ -290,6 +379,8 @@ Examples:
ref = args.ref
limit = args.limit
grep_patterns: List[str] = args.grep_patterns or []
file_paths: List[str] = args.file_paths or []
exclude_paths: List[str] = args.exclude_paths or []
use_first_parent = not args.no_first_parent

def matches_grep(subject: str) -> bool:
Expand Down Expand Up @@ -322,6 +413,10 @@ Examples:
log_cmd = ['git', 'log', f'--format=%H|%s|%an|%ar', ref, '-n', str(limit)]
if use_first_parent:
log_cmd.insert(2, '--first-parent')
# Add path filtering if specified
if file_paths:
log_cmd.append('--')
log_cmd.extend(file_paths)
log_output = run_command(log_cmd)

if not log_output:
Expand Down Expand Up @@ -362,7 +457,21 @@ Examples:
if len(train_parts) != 3:
continue
train_message, train_author, train_date = train_parts
if matches_grep(train_message):
# Apply grep, path include filters, and path exclude filters (AND logic)
matches_grep_filter = matches_grep(train_message)

if debug_mode:
print(f"DEBUG: Checking merge-train child {train_commit[:7]}: '{train_message[:60]}'", file=sys.stderr)

affects_include_paths = commit_affects_paths(train_commit, file_paths, debug_mode)
touches_exclude_paths = commit_touches_excluded_paths(train_commit, exclude_paths, debug_mode)

if debug_mode:
result = "INCLUDED" if (matches_grep_filter and affects_include_paths and not touches_exclude_paths) else "FILTERED"
print(f"DEBUG: -> Result: {result} (grep={matches_grep_filter}, includes={affects_include_paths}, excludes={touches_exclude_paths})", file=sys.stderr)
print(file=sys.stderr)

if matches_grep_filter and affects_include_paths and not touches_exclude_paths:
filtered_children.append((train_commit, train_message, train_author, train_date))

# Only print the merge-train header if there are matching children
Expand Down Expand Up @@ -402,6 +511,12 @@ Examples:
# Apply grep filter to regular commits
if not matches_grep(message):
continue
# Apply exclude paths filter
touches_exclude = commit_touches_excluded_paths(commit, exclude_paths, debug_mode)
if debug_mode:
print(f"DEBUG: Regular commit {commit[:7]}: excludes={touches_exclude}, message='{message[:60]}'", file=sys.stderr)
if touches_exclude:
continue
if args.group_by_type:
is_break, break_summary = detect_breaking(message, commit)
grouped_entries.append((commit, message, author, date, None, is_break, break_summary))
Expand Down