bisect: avoid copying ancestor list for non-merge commits
During a bisection, hg needs to compute a list of all ancestors for every
candidate commit. This is accomplished via a bottom-up traversal of the set of
candidates, during which each revision's ancestor list is populated using the
ancestor list of its parent(s). Previously, this involved copying the entire
list, which could be very long in if the bisection range was large.
To help improve this, we can observe that each candidate commit is visited
exactly once, at which point its ancestor list is copied into its children's
lists and then dropped. In the case of non-merge commits, a commit's ancestor
list consists exactly of its parent's list plus itself. This means that we can
trivially reuse the parent's existing list for one of its non-merge children,
which avoids copying entirely if that commit is the parent's only child. This
makes bisections over linear ranges of commits much faster.
During some informal testing in the large publicly-available `mozilla-central`
repository, this noticeably sped up bisections over large ranges of history:
Setup:
$ cd mozilla-central
$ hg bisect --reset
$ hg bisect --good 0
$ hg log -r tip -T '{rev}\n'
628417
Test:
$ time hg bisect --bad tip --noupdate
Before:
real 3m35.927s
user 3m35.553s
sys 0m0.319s
After:
real 1m41.142s
user 1m40.810s
sys 0m0.285s
#!/usr/bin/env python3
#
# check-py3-compat - check Python 3 compatibility of Mercurial files
#
# Copyright 2015 Gregory Szorc <gregory.szorc@gmail.com>
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version.
import ast
import importlib
import os
import sys
import traceback
import warnings
def check_compat_py3(f):
"""Check Python 3 compatibility of a file with Python 3."""
with open(f, 'rb') as fh:
content = fh.read()
try:
ast.parse(content, filename=f)
except SyntaxError as e:
print('%s: invalid syntax: %s' % (f, e))
return
# Try to import the module.
# For now we only support modules in packages because figuring out module
# paths for things not in a package can be confusing.
if f.startswith(
('hgdemandimport/', 'hgext/', 'mercurial/')
) and not f.endswith('__init__.py'):
assert f.endswith('.py')
name = f.replace('/', '.')[:-3]
try:
importlib.import_module(name)
except Exception as e:
exc_type, exc_value, tb = sys.exc_info()
# We walk the stack and ignore frames from our custom importer,
# import mechanisms, and stdlib modules. This kinda/sorta
# emulates CPython behavior in import.c while also attempting
# to pin blame on a Mercurial file.
for frame in reversed(traceback.extract_tb(tb)):
if frame.name == '_call_with_frames_removed':
continue
if 'importlib' in frame.filename:
continue
if 'mercurial/__init__.py' in frame.filename:
continue
if frame.filename.startswith(sys.prefix):
continue
break
if frame.filename:
filename = os.path.basename(frame.filename)
print(
'%s: error importing: <%s> %s (error at %s:%d)'
% (f, type(e).__name__, e, filename, frame.lineno)
)
else:
print(
'%s: error importing module: <%s> %s (line %d)'
% (f, type(e).__name__, e, frame.lineno)
)
if __name__ == '__main__':
# check_compat_py3 will import every filename we specify as long as it
# starts with one of a few prefixes. It does this by converting
# specified filenames like 'mercurial/foo.py' to 'mercurial.foo' and
# importing that. When running standalone (not as part of a test), this
# means we actually import the installed versions, not the files we just
# specified. When running as test-check-py3-compat.t, we technically
# would import the correct paths, but it's cleaner to have both cases
# use the same import logic.
sys.path.insert(0, os.getcwd())
for f in sys.argv[1:]:
with warnings.catch_warnings(record=True) as warns:
check_compat_py3(f)
for w in warns:
print(
warnings.formatwarning(
w.message, w.category, w.filename, w.lineno
).rstrip()
)
sys.exit(0)