##// END OF EJS Templates
findrenames: Optimise "addremove -s100" by matching files by their SHA1 hashes....
findrenames: Optimise "addremove -s100" by matching files by their SHA1 hashes. We speed up 'findrenames' for the usecase when a user specifies they want a similarity of 100% by matching files by their exact SHA1 hash value. This reduces the number of comparisons required to find exact matches from O(n^2) to O(n). While it would be nice if we could just use mercurial's pre-calculated SHA1 hash for existing files, this hash includes the file's ancestor information making it unsuitable for our purposes. Instead, we calculate the hash of old content from scratch. The following benchmarks were taken on the current head of crew: addremove 100% similarity: rm -rf *; hg up -C; mv tests tests.new hg --time addremove -s100 --dry-run before: real 176.350 secs (user 128.890+0.000 sys 47.430+0.000) after: real 2.130 secs (user 1.890+0.000 sys 0.240+0.000) addremove 75% similarity: rm -rf *; hg up -C; mv tests tests.new; \ for i in tests.new/*; do echo x >> $i; done hg --time addremove -s75 --dry-run before: real 264.560 secs (user 215.130+0.000 sys 49.410+0.000) after: real 218.710 secs (user 172.790+0.000 sys 45.870+0.000)

File last commit:

r10394:4612cded default
r11060:e6df0177 default
Show More
templatekw.py
265 lines | 8.1 KiB | text/x-python | PythonLexer
Patrick Mezard
cmdutil: replace showlist() closure with a function
r10053 # templatekw.py - common changeset template keywords
#
# Copyright 2005-2009 Matt Mackall <mpm@selenic.com>
#
# This software may be used and distributed according to the terms of the
Matt Mackall
Merge with stable
r10264 # GNU General Public License version 2 or any later version.
Patrick Mezard
cmdutil: replace showlist() closure with a function
r10053
Patrick Mezard
cmdutil: extract repo dependent closures in templatekw
r10055 from node import hex
Patrick Mezard
templatekw: change {file_copies} behaviour, add {file_copies_switch}...
r10060 import encoding, patch, util, error
Patrick Mezard
cmdutil: replace showlist() closure with a function
r10053
Patrick Mezard
templatekw: fix extras, manifest and showlist args (issue1989)...
r10260 def showlist(name, values, plural=None, **args):
Patrick Mezard
cmdutil: replace showlist() closure with a function
r10053 '''expand set of values.
name is name of key in template map.
values is list of strings or dicts.
plural is plural of name, if not simply name + 's'.
expansion works like this, given name 'foo'.
if values is empty, expand 'no_foos'.
if 'foo' not in template map, return values as a string,
joined by space.
expand 'start_foos'.
for each value, expand 'foo'. if 'last_foo' in template
map, expand it instead of 'foo' for last key.
expand 'end_foos'.
'''
Patrick Mezard
templatekw: fix extras, manifest and showlist args (issue1989)...
r10260 templ = args['templ']
Matt Mackall
many, many trivial check-code fixups
r10282 if plural:
names = plural
Patrick Mezard
cmdutil: replace showlist() closure with a function
r10053 else: names = name + 's'
if not values:
noname = 'no_' + names
if noname in templ:
yield templ(noname, **args)
return
if name not in templ:
if isinstance(values[0], str):
yield ' '.join(values)
else:
for v in values:
yield dict(v, **args)
return
startname = 'start_' + names
if startname in templ:
yield templ(startname, **args)
vargs = args.copy()
def one(v, tag=name):
try:
vargs.update(v)
except (AttributeError, ValueError):
try:
for a, b in v:
vargs[a] = b
except ValueError:
vargs[name] = v
return templ(tag, **vargs)
lastname = 'last_' + name
if lastname in templ:
last = values.pop()
else:
last = None
for v in values:
yield one(v)
if last is not None:
yield one(last, tag=lastname)
endname = 'end_' + names
if endname in templ:
yield templ(endname, **args)
Patrick Mezard
cmdutil: extract file changes closures into templatekw
r10056 def getfiles(repo, ctx, revcache):
if 'files' not in revcache:
revcache['files'] = repo.status(ctx.parents()[0].node(),
ctx.node())[:3]
return revcache['files']
Patrick Mezard
cmdutil: extract latest tags closures in templatekw
r10057 def getlatesttags(repo, ctx, cache):
'''return date, distance and name for the latest tag of rev'''
if 'latesttags' not in cache:
# Cache mapping from rev to a tuple with tag date, tag
# distance and tag name
cache['latesttags'] = {-1: (0, 0, 'null')}
latesttags = cache['latesttags']
rev = ctx.rev()
todo = [rev]
while todo:
rev = todo.pop()
if rev in latesttags:
continue
ctx = repo[rev]
tags = [t for t in ctx.tags() if repo.tagtype(t) == 'global']
if tags:
latesttags[rev] = ctx.date()[0], 0, ':'.join(sorted(tags))
continue
try:
# The tuples are laid out so the right one can be found by
# comparison.
pdate, pdist, ptag = max(
latesttags[p.rev()] for p in ctx.parents())
except KeyError:
# Cache miss - recurse
todo.append(rev)
todo.extend(p.rev() for p in ctx.parents())
continue
latesttags[rev] = pdate, pdist + 1, ptag
return latesttags[rev]
Patrick Mezard
templatekw: change {file_copies} behaviour, add {file_copies_switch}...
r10060 def getrenamedfn(repo, endrev=None):
rcache = {}
if endrev is None:
endrev = len(repo)
def getrenamed(fn, rev):
'''looks up all renames for a file (up to endrev) the first
time the file is given. It indexes on the changerev and only
parses the manifest if linkrev != changerev.
Returns rename info for fn at changerev rev.'''
if fn not in rcache:
rcache[fn] = {}
fl = repo.file(fn)
for i in fl:
lr = fl.linkrev(i)
renamed = fl.renamed(fl.node(i))
rcache[fn][lr] = renamed
if lr >= endrev:
break
if rev in rcache[fn]:
return rcache[fn][rev]
# If linkrev != rev (i.e. rev not found in rcache) fallback to
# filectx logic.
try:
return repo[rev][fn].renamed()
except error.LookupError:
return None
return getrenamed
Patrick Mezard
cmdutil: extract repo dependent closures in templatekw
r10055 def showauthor(repo, ctx, templ, **args):
Patrick Mezard
cmdutil: extract ctx dependent closures into templatekw
r10054 return ctx.user()
Patrick Mezard
templatekw: fix extras, manifest and showlist args (issue1989)...
r10260 def showbranches(**args):
branch = args['ctx'].branch()
Patrick Mezard
cmdutil: extract ctx dependent closures into templatekw
r10054 if branch != 'default':
branch = encoding.tolocal(branch)
Patrick Mezard
templatekw: fix extras, manifest and showlist args (issue1989)...
r10260 return showlist('branch', [branch], plural='branches', **args)
Patrick Mezard
cmdutil: extract ctx dependent closures into templatekw
r10054
Patrick Mezard
cmdutil: extract repo dependent closures in templatekw
r10055 def showdate(repo, ctx, templ, **args):
Patrick Mezard
cmdutil: extract ctx dependent closures into templatekw
r10054 return ctx.date()
Patrick Mezard
cmdutil: extract repo dependent closures in templatekw
r10055 def showdescription(repo, ctx, templ, **args):
Patrick Mezard
cmdutil: extract ctx dependent closures into templatekw
r10054 return ctx.description().strip()
Patrick Mezard
cmdutil: extract repo dependent closures in templatekw
r10055 def showdiffstat(repo, ctx, templ, **args):
diff = patch.diff(repo, ctx.parents()[0].node(), ctx.node())
files, adds, removes = 0, 0, 0
for i in patch.diffstatdata(util.iterlines(diff)):
files += 1
adds += i[1]
removes += i[2]
return '%s: +%s/-%s' % (files, adds, removes)
Patrick Mezard
templatekw: fix extras, manifest and showlist args (issue1989)...
r10260 def showextras(**args):
templ = args['templ']
for key, value in sorted(args['ctx'].extra().items()):
Patrick Mezard
cmdutil: extract ctx dependent closures into templatekw
r10054 args = args.copy()
args.update(dict(key=key, value=value))
yield templ('extra', **args)
Patrick Mezard
templatekw: fix extras, manifest and showlist args (issue1989)...
r10260 def showfileadds(**args):
repo, ctx, revcache = args['repo'], args['ctx'], args['revcache']
return showlist('file_add', getfiles(repo, ctx, revcache)[1], **args)
Patrick Mezard
cmdutil: extract file changes closures into templatekw
r10056
Patrick Mezard
templatekw: fix extras, manifest and showlist args (issue1989)...
r10260 def showfilecopies(**args):
Benoit Boissinot
fix coding style (reported by pylint)
r10394 cache, ctx = args['cache'], args['ctx']
Patrick Mezard
templatekw: fix extras, manifest and showlist args (issue1989)...
r10260 copies = args['revcache'].get('copies')
Patrick Mezard
templatekw: change {file_copies} behaviour, add {file_copies_switch}...
r10060 if copies is None:
if 'getrenamed' not in cache:
Patrick Mezard
templatekw: fix extras, manifest and showlist args (issue1989)...
r10260 cache['getrenamed'] = getrenamedfn(args['repo'])
Patrick Mezard
templatekw: change {file_copies} behaviour, add {file_copies_switch}...
r10060 copies = []
getrenamed = cache['getrenamed']
for fn in ctx.files():
rename = getrenamed(fn, ctx.rev())
if rename:
copies.append((fn, rename[0]))
Matt Mackall
many, many trivial check-code fixups
r10282
Patrick Mezard
templatekw: change {file_copies} behaviour, add {file_copies_switch}...
r10060 c = [{'name': x[0], 'source': x[1]} for x in copies]
Patrick Mezard
templatekw: fix extras, manifest and showlist args (issue1989)...
r10260 return showlist('file_copy', c, plural='file_copies', **args)
Patrick Mezard
templatekw: change {file_copies} behaviour, add {file_copies_switch}...
r10060
# showfilecopiesswitch() displays file copies only if copy records are
# provided before calling the templater, usually with a --copies
# command line switch.
Patrick Mezard
templatekw: fix extras, manifest and showlist args (issue1989)...
r10260 def showfilecopiesswitch(**args):
copies = args['revcache'].get('copies') or []
Patrick Mezard
templatekw: change {file_copies} behaviour, add {file_copies_switch}...
r10060 c = [{'name': x[0], 'source': x[1]} for x in copies]
Patrick Mezard
templatekw: fix extras, manifest and showlist args (issue1989)...
r10260 return showlist('file_copy', c, plural='file_copies', **args)
Patrick Mezard
cmdutil: extract file copies closure into templatekw
r10058
Patrick Mezard
templatekw: fix extras, manifest and showlist args (issue1989)...
r10260 def showfiledels(**args):
repo, ctx, revcache = args['repo'], args['ctx'], args['revcache']
return showlist('file_del', getfiles(repo, ctx, revcache)[2], **args)
Patrick Mezard
cmdutil: extract file changes closures into templatekw
r10056
Patrick Mezard
templatekw: fix extras, manifest and showlist args (issue1989)...
r10260 def showfilemods(**args):
repo, ctx, revcache = args['repo'], args['ctx'], args['revcache']
return showlist('file_mod', getfiles(repo, ctx, revcache)[0], **args)
Patrick Mezard
cmdutil: extract file changes closures into templatekw
r10056
Patrick Mezard
templatekw: fix extras, manifest and showlist args (issue1989)...
r10260 def showfiles(**args):
return showlist('file', args['ctx'].files(), **args)
Patrick Mezard
cmdutil: extract ctx dependent closures into templatekw
r10054
Patrick Mezard
cmdutil: extract latest tags closures in templatekw
r10057 def showlatesttag(repo, ctx, templ, cache, **args):
return getlatesttags(repo, ctx, cache)[2]
def showlatesttagdistance(repo, ctx, templ, cache, **args):
return getlatesttags(repo, ctx, cache)[1]
Patrick Mezard
templatekw: fix extras, manifest and showlist args (issue1989)...
r10260 def showmanifest(**args):
repo, ctx, templ = args['repo'], args['ctx'], args['templ']
Patrick Mezard
cmdutil: extract repo dependent closures in templatekw
r10055 args = args.copy()
args.update(dict(rev=repo.manifest.rev(ctx.changeset()[0]),
node=hex(ctx.changeset()[0])))
return templ('manifest', **args)
def shownode(repo, ctx, templ, **args):
Patrick Mezard
cmdutil: extract ctx dependent closures into templatekw
r10054 return ctx.hex()
Patrick Mezard
cmdutil: extract repo dependent closures in templatekw
r10055 def showrev(repo, ctx, templ, **args):
Patrick Mezard
cmdutil: extract ctx dependent closures into templatekw
r10054 return ctx.rev()
Patrick Mezard
templatekw: fix extras, manifest and showlist args (issue1989)...
r10260 def showtags(**args):
return showlist('tag', args['ctx'].tags(), **args)
Patrick Mezard
cmdutil: extract ctx dependent closures into templatekw
r10054
Patrick Mezard
templatekw: fix extras, manifest and showlist args (issue1989)...
r10260 # keywords are callables like:
# fn(repo, ctx, templ, cache, revcache, **args)
# with:
# repo - current repository instance
# ctx - the changectx being displayed
# templ - the templater instance
# cache - a cache dictionary for the whole templater run
# revcache - a cache dictionary for the current revision
Patrick Mezard
cmdutil: extract ctx dependent closures into templatekw
r10054 keywords = {
'author': showauthor,
'branches': showbranches,
'date': showdate,
'desc': showdescription,
Patrick Mezard
cmdutil: extract repo dependent closures in templatekw
r10055 'diffstat': showdiffstat,
Patrick Mezard
cmdutil: extract ctx dependent closures into templatekw
r10054 'extras': showextras,
Patrick Mezard
cmdutil: extract file changes closures into templatekw
r10056 'file_adds': showfileadds,
Patrick Mezard
cmdutil: extract file copies closure into templatekw
r10058 'file_copies': showfilecopies,
Patrick Mezard
templatekw: change {file_copies} behaviour, add {file_copies_switch}...
r10060 'file_copies_switch': showfilecopiesswitch,
Patrick Mezard
cmdutil: extract file changes closures into templatekw
r10056 'file_dels': showfiledels,
'file_mods': showfilemods,
Patrick Mezard
cmdutil: extract ctx dependent closures into templatekw
r10054 'files': showfiles,
Patrick Mezard
cmdutil: extract latest tags closures in templatekw
r10057 'latesttag': showlatesttag,
'latesttagdistance': showlatesttagdistance,
Patrick Mezard
cmdutil: extract repo dependent closures in templatekw
r10055 'manifest': showmanifest,
Patrick Mezard
cmdutil: extract ctx dependent closures into templatekw
r10054 'node': shownode,
'rev': showrev,
'tags': showtags,
}