mirror of
https://github.com/daijro/camoufox.git
synced 2026-11-19 00:07:28 +00:00
Values both launchers hard-coded (marker fonts, the Windows 11 fonts, the Apple Silicon and Intel Mac core and panel tables, plausible DPR and core counts, the macOS MacinTalk/Eloquence voices, the cache prefs) now live in python/src/camoufox/launcher-constants.json, which the TypeScript package ships as a data file. gen-fonts-json.py and verify-fonts.py read the same file, so the drifted PingFang marker copy is gone (fonts.json regenerates byte-identical) and verify-fonts no longer ast-parses fingerprints.py. - Drop the duplicate fpgen model pin under browser/scripts/data; the TS test checks the fpgen package's MODEL_PIN against the Python copy. - Move pin-fpgen-model.py and clean-fingerprint-data.py to python/scripts. - Delete the font licence report (font-metrics/licences.py and its saved output, open-fonts.txt): the bundle ships as built, so the report checks nothing. - Delete unreferenced files: cleanfonts.sh, python/examples, typescript/renovate.json. - Delete dead code: classify_installs, remove_mmdb, write_upstream_sh, bump_release, unused test constants and fixture helpers, the _virtual_display attribute writes, package.py's pre-groups font copy (a missing groups.json now fails). - Share require_bundle, zip_entries, gh_headers and parse_shard instead of keeping copies; fold GitHubDownloader into CamoufoxFetcher in both launchers. - TS raises ValueError where Python does, and `camoufox version` prints the fpgen package version as Python does. - Python tests get src/ from pytest's pythonpath instead of sys.path inserts; skips that hid untested paths now assert or test the behaviour. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
120 lines
5.0 KiB
Python
120 lines
5.0 KiB
Python
#!/usr/bin/env python3
|
|
"""Drop identities the shipped data files cannot honestly offer.
|
|
|
|
`camoufox.coherence` checks an identity when one is drawn. This does the same
|
|
to the DATA the identities are drawn FROM, so an impossible row never reaches a
|
|
build in the first place. Two filters for the same rules: the runtime one has to
|
|
produce an identity and so repairs what it can, while this one can simply drop a
|
|
row -- there are plenty of others, and dropping loses nothing but a machine that
|
|
never existed.
|
|
|
|
What it covers:
|
|
|
|
fingerprint-presets.json, fingerprint-presets-v150.json
|
|
Each preset is converted the way a launch converts it and checked, and
|
|
its GPU must be one fpgen has seen Firefox report on that OS: WebGL
|
|
parameters come from fpgen, and a GPU it has never seen has none. A row
|
|
that fails is dropped rather than repaired: repairing would write an
|
|
invented value ("what core count does a 2-core Apple M1 really have?")
|
|
into a file whose entire purpose is being real.
|
|
|
|
Usage:
|
|
python3 python/scripts/clean-fingerprint-data.py # report only
|
|
python3 python/scripts/clean-fingerprint-data.py --write # rewrite the files
|
|
|
|
`--check` exits non-zero when anything would be dropped;
|
|
python/tests/test_shipped_data.py asserts the same thing, so a data refresh that
|
|
reintroduces a bad row fails CI rather than shipping.
|
|
"""
|
|
|
|
import argparse
|
|
import json
|
|
import sys
|
|
from collections import Counter
|
|
from pathlib import Path
|
|
|
|
PYTHON_SRC = Path(__file__).resolve().parents[1] / 'src'
|
|
sys.path.insert(0, str(PYTHON_SRC))
|
|
|
|
from camoufox import coherence # noqa: E402
|
|
from camoufox.fingerprints import from_preset # noqa: E402
|
|
from camoufox.webgl import firefox_gpus # noqa: E402
|
|
|
|
PRESET_FILES = (
|
|
PYTHON_SRC / 'camoufox' / 'fingerprint-presets.json',
|
|
PYTHON_SRC / 'camoufox' / 'fingerprint-presets-v150.json',
|
|
)
|
|
OS_KEY = {'macos': 'mac', 'windows': 'win', 'linux': 'lin'}
|
|
# The Firefox version only decides the UA rewrite, which no rule reads.
|
|
FF_VERSION = '152'
|
|
|
|
|
|
def preset_violations(preset, os_name):
|
|
"""What this preset breaks, as a launch would see it."""
|
|
try:
|
|
config = from_preset(preset, FF_VERSION)
|
|
except Exception as exc: # a row too malformed to convert is itself a defect
|
|
return [coherence.Violation('unconvertible', f'{type(exc).__name__}: {exc}')]
|
|
violations = coherence.validate(config, OS_KEY[os_name])
|
|
gpu = (preset.get('webgl', {}).get('unmaskedVendor'), preset.get('webgl', {}).get('unmaskedRenderer'))
|
|
if gpu not in firefox_gpus(os_name):
|
|
violations.append(coherence.Violation(
|
|
'gpu-without-webgl-data', f'fpgen has never seen Firefox on {os_name} report {gpu[1]!r}'))
|
|
return violations
|
|
|
|
|
|
def clean_presets(path, write):
|
|
data = json.loads(path.read_text())
|
|
dropped = Counter()
|
|
kept_total = dropped_total = 0
|
|
for os_name, presets in data['presets'].items():
|
|
kept = []
|
|
for index, preset in enumerate(presets):
|
|
violations = preset_violations(preset, os_name)
|
|
if violations:
|
|
dropped_total += 1
|
|
for violation in violations:
|
|
dropped[violation.rule] += 1
|
|
print(f' drop {path.name} {os_name}[{index}]: '
|
|
f'{"; ".join(v.detail for v in violations)[:120]}')
|
|
else:
|
|
kept.append(preset)
|
|
kept_total += len(kept)
|
|
data['presets'][os_name] = kept
|
|
if write and dropped_total:
|
|
# Keep the file's own formatting. Rewriting 2-space-indented JSON
|
|
# compactly turns a 38-row deletion into a 22,546-line diff that no
|
|
# reviewer can read, which is how a data change stops being reviewable.
|
|
# ensure_ascii=False as well: the originals hold 'Amélie' as itself,
|
|
# and escaping it would rewrite lines no preset touched.
|
|
# A no-op run must be byte-identical, so the diff of a real run is the
|
|
# dropped rows and nothing else: 2-space indent, non-ASCII kept as
|
|
# itself ('Amélie'), and no trailing newline, which is how these
|
|
# files are written.
|
|
path.write_text(json.dumps(data, indent=2, ensure_ascii=False))
|
|
print(f'{path.name}: {dropped_total} dropped, {kept_total} kept'
|
|
+ (f' {dict(dropped)}' if dropped else ''))
|
|
return dropped_total
|
|
|
|
|
|
def main():
|
|
parser = argparse.ArgumentParser(description=__doc__)
|
|
parser.add_argument('--write', action='store_true', help='rewrite the data files')
|
|
parser.add_argument('--check', action='store_true', help='exit non-zero if anything is unclean')
|
|
args = parser.parse_args()
|
|
|
|
total = sum(clean_presets(path, args.write) for path in PRESET_FILES)
|
|
|
|
if not total:
|
|
print('\nEvery shipped identity is coherent.')
|
|
return 0
|
|
if args.write:
|
|
print(f'\n{total} incoherent entr(ies) removed.')
|
|
return 0
|
|
print(f'\n{total} incoherent entr(ies). Run with --write to remove them.')
|
|
return 1 if args.check else 0
|
|
|
|
|
|
if __name__ == '__main__':
|
|
raise SystemExit(main())
|