##// END OF EJS Templates
hg: obtain lock when creating share from pooled repo (issue5104)...
hg: obtain lock when creating share from pooled repo (issue5104) There are race conditions between clients performing a shared clone to pooled storage: 1) Clients race to create the new shared repo in the pool directory 2) 1 client is seeding the repo in the pool directory and another goes to share it before it is fully cloned We prevent these race conditions by obtaining a lock in the pool directory that is derived from the name of the repo we will be accessing. To test this, a simple generic "lockdelay" extension has been added. The extension inserts an optional, configurable delay before or after lock acquisition. In the test, we delay 2 seconds after lock acquisition in the first process and 1 second before lock acquisition in the 2nd process. This means the first process has 1s to obtain the lock. There is a race condition here. If we encounter it in the wild, we could change the dummy extension to wait on the lock file to appear instead of relying on timing. But that's more complicated. Let's see what happens first.

File last commit:

r20701:d20817ac default
r28289:d493d647 3.7.2 stable
Show More
fix_bytesmod.py
63 lines | 2.3 KiB | text/x-python | PythonLexer
"""Fixer that changes bytes % whatever to a function that actually formats
it."""
from lib2to3 import fixer_base
from lib2to3.fixer_util import is_tuple, Call, Comma, Name, touch_import
# XXX: Implementing a blacklist in 2to3 turned out to be more troublesome than
# blacklisting some modules inside the fixers. So, this is what I came with.
blacklist = ['mercurial/demandimport.py',
'mercurial/py3kcompat.py',
'mercurial/i18n.py',
]
def isnumberremainder(formatstr, data):
try:
if data.value.isdigit():
return True
except AttributeError:
return False
class FixBytesmod(fixer_base.BaseFix):
# XXX: There's one case (I suppose) I can't handle: when a remainder
# operation like foo % bar is performed, I can't really know what the
# contents of foo and bar are. I believe the best approach is to "correct"
# the to-be-converted code and let bytesformatter handle that case in
# runtime.
PATTERN = '''
term< formatstr=STRING '%' data=STRING > |
term< formatstr=STRING '%' data=atom > |
term< formatstr=NAME '%' data=any > |
term< formatstr=any '%' data=any >
'''
def transform(self, node, results):
for bfn in blacklist:
if self.filename.endswith(bfn):
return
if not self.filename.endswith('mercurial/py3kcompat.py'):
touch_import('mercurial', 'py3kcompat', node=node)
formatstr = results['formatstr'].clone()
data = results['data'].clone()
formatstr.prefix = '' # remove spaces from start
if isnumberremainder(formatstr, data):
return
# We have two possibilities:
# 1- An identifier or name is passed, it is going to be a leaf, thus, we
# just need to copy its value as an argument to the formatter;
# 2- A tuple is explicitly passed. In this case, we're gonna explode it
# to pass to the formatter
# TODO: Check for normal strings. They don't need to be translated
if is_tuple(data):
args = [formatstr, Comma().clone()] + \
[c.clone() for c in data.children[:]]
else:
args = [formatstr, Comma().clone(), data]
call = Call(Name('bytesformatter', prefix=' '), args)
return call