iOSBuildServer/backend/automation/obfuscation/obfuscation_runtime/discovery.py

144 lines
4.5 KiB
Python

import os
import shutil
import subprocess
from pathlib import Path
def path_matches_dir_names(path: str, dir_names: tuple) -> bool:
if not dir_names:
return False
wanted = {name.lower() for name in dir_names}
return any(part.lower() in wanted for part in Path(path).parts)
def should_exclude_path(path: str, exclude_path_fragments) -> bool:
return any(fragment in path for fragment in exclude_path_fragments)
def _run_command_capture_stdout(args: list) -> str:
try:
completed = subprocess.run(
args,
capture_output=True,
text=True,
check=False,
)
except Exception:
return ""
if completed.returncode not in (0, 1):
return ""
return completed.stdout
def _discover_files_via_os_walk(scan_dirs: list, allowed_exts: tuple, exclude_path_fragments, skip_dir_names: tuple = (), basename_prefixes: tuple = ()) -> list:
files = []
allowed_exts = set(allowed_exts)
for scan_dir in scan_dirs:
for root, dirs, filenames in os.walk(scan_dir):
dirs[:] = [
directory for directory in dirs
if not any(fragment.strip("/") in directory for fragment in exclude_path_fragments)
]
if path_matches_dir_names(root, skip_dir_names):
continue
for filename in filenames:
ext = os.path.splitext(filename)[1]
if ext not in allowed_exts:
continue
name_no_ext = os.path.splitext(filename)[0]
if basename_prefixes and not any(name_no_ext.startswith(prefix) for prefix in basename_prefixes):
continue
full_path = os.path.normpath(os.path.join(root, filename))
if should_exclude_path(full_path, exclude_path_fragments):
continue
if path_matches_dir_names(full_path, skip_dir_names):
continue
files.append(full_path)
return sorted(set(files))
def _discover_files_via_rg(scan_dirs: list, allowed_exts: tuple, exclude_path_fragments, skip_dir_names: tuple = (), basename_prefixes: tuple = ()) -> list:
rg = shutil.which("rg")
if not rg:
return []
cmd = [rg, "--files", "--no-ignore", "--hidden"]
for ext in allowed_exts:
cmd.extend(["-g", f"*{ext}"])
cmd.extend(scan_dirs)
output = _run_command_capture_stdout(cmd)
if not output:
return []
files = []
for line in output.splitlines():
path = os.path.normpath(line.strip())
if not path or should_exclude_path(path, exclude_path_fragments):
continue
if path_matches_dir_names(path, skip_dir_names):
continue
if basename_prefixes:
name_no_ext = os.path.splitext(os.path.basename(path))[0]
if not any(name_no_ext.startswith(prefix) for prefix in basename_prefixes):
continue
files.append(path)
return sorted(set(files))
def discover_files(scan_dirs: list, allowed_exts: tuple, exclude_path_fragments, skip_dir_names: tuple = (), basename_prefixes: tuple = ()) -> list:
files = _discover_files_via_rg(
scan_dirs,
allowed_exts,
exclude_path_fragments,
skip_dir_names=skip_dir_names,
basename_prefixes=basename_prefixes,
)
if files:
return files
return _discover_files_via_os_walk(
scan_dirs,
allowed_exts,
exclude_path_fragments,
skip_dir_names=skip_dir_names,
basename_prefixes=basename_prefixes,
)
def _chunked(items: list, chunk_size: int):
for index in range(0, len(items), chunk_size):
yield items[index:index + chunk_size]
def prefilter_files_with_rg(files: list, pattern: str, chunk_size: int = 256) -> list:
files = sorted(set(files))
if not files:
return []
rg = shutil.which("rg")
if not rg:
return files
matched = set()
for chunk in _chunked(files, chunk_size):
try:
completed = subprocess.run(
[rg, "-l", "--no-messages", "-e", pattern, *chunk],
capture_output=True,
text=True,
check=False,
)
except Exception:
return files
if completed.returncode not in (0, 1):
return files
for line in completed.stdout.splitlines():
path = os.path.normpath(line.strip())
if path:
matched.add(path)
return sorted(matched)