secubox-deb/scripts/license-headers.py
CyberMind-FR b88b8ada95 fix(license): tighten detect_existing + missing-file = repo-wide (ref #81)
Two bug fixes to the CMSD-1.0 license-header tool:

1. detect_existing() previously matched the SPDX token anywhere in the
   first 10 lines, including inside docstrings and prose comments.
   Tightened the matcher to require comment markers (#, //, *, <!--)
   and whitespace before the SPDX token.

   Regression tests:
   - test_detect_existing_no_false_match_in_docstring
   - test_detect_existing_no_false_match_inline_comment_prose

2. _read_enrollment() now returns ["**"] when the allowlist file is
   absent, per spec §5.2 "missing file → repo-wide enforcement".
   Previously it returned [] (nothing enforced), making Phase C closure
   impossible without an additional file.

   Empty allowlist file still returns [] (Phase A initial), so
   test_main_empty_allowlist_passes_check is unaffected.

   New tests:
   - test_read_enrollment_missing_file_means_repo_wide
   - test_main_check_missing_allowlist_enforces_repo_wide

Suite: 49 → 53 passing. --check still exits 0 repo-wide.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-12 11:39:57 +02:00

330 lines
10 KiB
Python

# SPDX-License-Identifier: LicenseRef-CMSD-1.0
# Copyright (c) 2026 CyberMind — Gérald Kerma <devel@cybermind.fr>
# Source-Disclosed License — All rights reserved except as expressly granted.
# See LICENCE-CMSD-1.0.md for terms.
# scripts/license-headers.py
"""CMSD-1.0 license header tool.
Adds and verifies the SPDX-License-Identifier: LicenseRef-CMSD-1.0 header
on every first-party source file. See docs/superpowers/specs/2026-05-12-license-headers-design.md.
"""
from __future__ import annotations
import argparse
import difflib
import fnmatch
import re
import sys
from pathlib import Path
HEADER_LINES = (
"SPDX-License-Identifier: LicenseRef-CMSD-1.0",
"Copyright (c) 2026 CyberMind — Gérald Kerma <devel@cybermind.fr>",
"Source-Disclosed License — All rights reserved except as expressly granted.",
"See LICENCE-CMSD-1.0.md for terms.",
)
_CMSD_ID = "LicenseRef-CMSD-1.0"
# Matches an SPDX line only when preceded by comment markers and/or
# whitespace. Prevents false-matches when a docstring mentions the
# token "SPDX-License-Identifier:" in prose.
_SPDX_LINE_RE = re.compile(
r"^[\s/*#<!\->]*\s*SPDX-License-Identifier:\s*(\S+)"
)
ENROLLMENT_FILE = "scripts/license-headers-enrolled.txt"
def detect_existing(text: str) -> str:
"""Return 'MATCH', 'FOREIGN', or 'NONE' based on the first 10 lines.
Only lines whose non-whitespace content begins with comment markers
(#, //, *, <!--, -->) and then an SPDX identifier count as a license
declaration. Prose mentions inside docstrings are ignored.
"""
for line in text.splitlines()[:10]:
match = _SPDX_LINE_RE.match(line)
if match:
return "MATCH" if match.group(1) == _CMSD_ID else "FOREIGN"
return "NONE"
def render_header(style: str) -> str:
if style == "hash":
return "".join(f"# {line}\n" for line in HEADER_LINES)
if style == "slash":
return "".join(f"// {line}\n" for line in HEADER_LINES)
if style == "block":
body = "".join(f" * {line}\n" for line in HEADER_LINES)
return f"/*\n{body} */\n"
if style == "html":
body = "".join(f" {line}\n" for line in HEADER_LINES)
return f"<!--\n{body}-->\n"
raise ValueError(f"unknown comment style: {style}")
SKIP_DIRS = frozenset({
"kernel-build", "redroid", "Tow-Boot",
"output", "cache", "backups", "apt", "repo",
"node_modules", ".venv", ".git", "__pycache__", "dist", "build",
})
SKIP_GLOBS = (
"*.min.js",
"*.min.css",
"package-lock.json",
"*.lock",
)
LANG_TABLE: dict[str, tuple[str, str]] = {
".py": ("hash", "python"),
".sh": ("hash", "shebang_hash"),
".js": ("slash", "top"),
".mjs": ("slash", "top"),
".ts": ("slash", "top"),
".css": ("block", "top"),
".c": ("block", "top"),
".h": ("block", "top"),
".html": ("html", "html"),
".htm": ("html", "html"),
".md": ("html", "markdown"),
".toml": ("hash", "top"),
".yaml": ("hash", "top"),
".yml": ("hash", "top"),
".conf": ("hash", "top"),
}
def _place_python(header: str, text: str) -> str:
lines = text.splitlines(keepends=True)
insert_at = 0
# Skip shebang
if lines and lines[0].startswith("#!"):
insert_at = 1
# Skip encoding declaration (PEP 263)
if insert_at < len(lines) and re.match(r"^#.*coding[:=]", lines[insert_at]):
insert_at += 1
return "".join(lines[:insert_at]) + header + "\n" + "".join(lines[insert_at:])
def _place_shebang_hash(header: str, text: str) -> str:
"""Hash-comment language with a shebang line on line 1."""
lines = text.splitlines(keepends=True)
insert_at = 1 if lines and lines[0].startswith("#!") else 0
return "".join(lines[:insert_at]) + header + "\n" + "".join(lines[insert_at:])
def _place_top(header: str, text: str) -> str:
"""Insert header at line 1 with one blank line separator."""
return header + "\n" + text
def _place_html(header: str, text: str) -> str:
lines = text.splitlines(keepends=True)
insert_at = 0
if lines and lines[0].lstrip().lower().startswith("<!doctype"):
insert_at = 1
return "".join(lines[:insert_at]) + header + "\n" + "".join(lines[insert_at:])
def _place_markdown(header: str, text: str) -> str:
"""Place after YAML frontmatter if present, else at top."""
lines = text.splitlines(keepends=True)
insert_at = 0
if lines and lines[0].rstrip() == "---":
for i in range(1, len(lines)):
if lines[i].rstrip() == "---":
insert_at = i + 1
break
return "".join(lines[:insert_at]) + header + "\n" + "".join(lines[insert_at:])
_PLACERS = {
"python": _place_python,
"shebang_hash": _place_shebang_hash,
"top": _place_top,
"html": _place_html,
"markdown": _place_markdown,
}
def apply(text: str, ext: str) -> str:
"""Insert the CMSD header into `text` for files with extension `ext`.
Returns `text` unchanged if a CMSD header is already present or if a
foreign SPDX header is detected.
"""
if ext not in LANG_TABLE:
return text
status = detect_existing(text)
if status in ("MATCH", "FOREIGN"):
return text
style, placer_name = LANG_TABLE[ext]
header = render_header(style)
return _PLACERS[placer_name](header, text)
def _is_in_scope(path: Path) -> bool:
"""True iff `path` has an extension in LANG_TABLE and matches no skip glob."""
if path.suffix not in LANG_TABLE:
return False
name = path.name
for pat in SKIP_GLOBS:
if fnmatch.fnmatch(name, pat):
return False
return True
def _matches_allowlist(rel: Path, enrolled: list[str]) -> bool:
"""True iff `rel` matches any glob in `enrolled`."""
rel_str = rel.as_posix()
for pat in enrolled:
if fnmatch.fnmatch(rel_str, pat):
return True
# Allow "common/**" to match "common/a.py"
if pat.endswith("/**") and (
rel_str == pat[:-3] or rel_str.startswith(pat[:-2])
):
return True
return False
def walk(paths: list[Path], enrolled: list[str], repo_root: Path | None = None):
"""Yield in-scope files under each path, honoring SKIP_DIRS, SKIP_GLOBS, allowlist.
`repo_root` (when provided) is the base for allowlist pattern matching.
Allowlist patterns are repo-relative, so a caller walking a subdirectory
must pass `repo_root` separately for the patterns to match. When omitted,
paths are matched relative to their walk root (backwards-compatible).
"""
allowlist_base = repo_root.resolve() if repo_root is not None else None
for root in paths:
root = Path(root)
if root.is_file():
if not _is_in_scope(root):
continue
rel = (
root.resolve().relative_to(allowlist_base)
if allowlist_base is not None
else root
)
if _matches_allowlist(rel, enrolled):
yield root
continue
for p in root.rglob("*"):
# Prune skip-dirs
if any(part in SKIP_DIRS for part in p.relative_to(root).parts):
continue
if not p.is_file():
continue
if not _is_in_scope(p):
continue
if allowlist_base is not None:
try:
rel = p.resolve().relative_to(allowlist_base)
except ValueError:
continue
else:
rel = p.relative_to(root)
if not _matches_allowlist(rel, enrolled):
continue
yield p
def _find_repo_root(start: Path) -> Path:
cur = start.resolve()
while cur != cur.parent:
if (cur / ".git").exists():
return cur
cur = cur.parent
return start
def _read_enrollment(repo_root: Path) -> list[str]:
"""Return enrollment patterns from scripts/license-headers-enrolled.txt.
Phase semantics (per spec §5.2):
* Missing file → ["**"] — repo-wide enforcement (Phase C final state)
* File exists, empty / only comments → [] — nothing enforced (Phase A initial)
* File with patterns → those patterns
"""
f = repo_root / ENROLLMENT_FILE
if not f.exists():
return ["**"]
patterns: list[str] = []
for raw in f.read_text().splitlines():
line = raw.strip()
if not line or line.startswith("#"):
continue
patterns.append(line)
return patterns
def main(argv: list[str] | None = None) -> int:
parser = argparse.ArgumentParser(description="CMSD-1.0 license header tool")
mode = parser.add_mutually_exclusive_group(required=True)
mode.add_argument("--check", action="store_true")
mode.add_argument("--fix", action="store_true")
mode.add_argument("--list", dest="list_", action="store_true")
mode.add_argument("--diff", action="store_true")
parser.add_argument("paths", nargs="*")
try:
args = parser.parse_args(argv)
except SystemExit as e:
return int(e.code) if isinstance(e.code, int) else 2
repo_root = _find_repo_root(Path.cwd())
enrolled = _read_enrollment(repo_root)
paths = [Path(p) for p in args.paths] if args.paths else [repo_root]
missing: list[Path] = []
for p in walk(paths, enrolled, repo_root=repo_root):
text = p.read_text(encoding="utf-8", errors="replace")
status = detect_existing(text)
if status == "FOREIGN":
print(f"skip (foreign SPDX): {p}", file=sys.stderr)
continue
if args.list_:
if status == "NONE":
print(p)
continue
if args.diff:
if status == "NONE":
new = apply(text, p.suffix)
for line in difflib.unified_diff(
text.splitlines(keepends=True),
new.splitlines(keepends=True),
fromfile=str(p),
tofile=str(p),
):
sys.stdout.write(line)
continue
if args.fix:
if status == "NONE":
p.write_text(apply(text, p.suffix), encoding="utf-8")
continue
if args.check:
if status == "NONE":
missing.append(p)
if args.check and missing:
print("Files missing CMSD-1.0 header:")
for p in missing:
print(f" {p}")
return 1
return 0
if __name__ == "__main__":
sys.exit(main(sys.argv[1:]))