fileset: extract language processing part to new module (API)
I'll add a couple more functions that work on parsed tree.
% wc -l mercurial/fileset*.py
559 mercurial/fileset.py
135 mercurial/filesetlang.py
694 total
#!/usr/bin/env python
# Filter output by pyflakes to control which warnings we check
from __future__ import absolute_import, print_function
import re
import sys
lines = []
for line in sys.stdin:
# We blacklist tests that are too noisy for us
pats = [
r"undefined name 'WindowsError'",
r"redefinition of unused '[^']+' from line",
# for cffi, allow re-exports from pure.*
r"cffi/[^:]*:.*\bimport \*' used",
r"cffi/[^:]*:.*\*' imported but unused",
]
keep = True
for pat in pats:
if re.search(pat, line):
keep = False
break # pattern matches
if keep:
fn = line.split(':', 1)[0]
f = open(fn)
data = f.read()
f.close()
if 'no-' 'check-code' in data:
continue
lines.append(line)
for line in lines:
sys.stdout.write(line)
print()