##// END OF EJS Templates
emitrevision: consider ancestors revision to emit as available base...
emitrevision: consider ancestors revision to emit as available base This should make more delta base valid. This notably affects: * case where we skipped some parent with empty delta to directly delta against an ancestors * case where an intermediate snapshots is stored. This change means we could sent largish intermediate snapshots over the wire. However this is actually a sub goal here. Sending snapshots over the wire means the client have a high odd of simply storing the pre-computed delta instead of doing a lengthy process that will… end up doing the same intermediate snapshot. In addition the overall size of snapshot (or any level) is "only" some or the overall delta size. (0.17% for my mercurial clone, 20% for my clone of Mozilla try). So Sending them other the wire is unlikely to change large impact on the bandwidth used. If we decide that minimising the bandwidth is an explicit goal, we should introduce new logic to filter-out snapshot as delta. The current code has no notion explicite of snapshot so far, they just tended to fall into the wobbly filtering options. In some cases, this patch can yield large improvement to the bundling time: ### data-env-vars.name = mozilla-try-2019-02-18-zstd-sparse-revlog # benchmark.name = perf-bundle # benchmark.variants.revs = last-100000 before: 68.787066 seconds after: 47.552677 seconds (-30.87%) That translate to large improvement to the pull time : ### data-env-vars.name = mozilla-try-2019-02-18-zstd-sparse-revlog # benchmark.name = pull # benchmark.variants.issue6528 = disabled # benchmark.variants.revs = last-100000 before: 142.186625 seconds after: 75.897745 seconds (-46.62%) No significant negative impact have been observed.

File last commit:

r49801:642e31cb default
r50567:e92de86c default
Show More
store.py
194 lines | 6.1 KiB | text/x-python | PythonLexer
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version.
# based on bundleheads extension by Gregory Szorc <gps@mozilla.com>
import abc
import os
import subprocess
from mercurial.node import hex
from mercurial.pycompat import open
from mercurial import pycompat
from mercurial.utils import (
hashutil,
procutil,
)
class BundleWriteException(Exception):
pass
class BundleReadException(Exception):
pass
class abstractbundlestore: # pytype: disable=ignored-metaclass
"""Defines the interface for bundle stores.
A bundle store is an entity that stores raw bundle data. It is a simple
key-value store. However, the keys are chosen by the store. The keys can
be any Python object understood by the corresponding bundle index (see
``abstractbundleindex`` below).
"""
__metaclass__ = abc.ABCMeta
@abc.abstractmethod
def write(self, data):
"""Write bundle data to the store.
This function receives the raw data to be written as a str.
Throws BundleWriteException
The key of the written data MUST be returned.
"""
@abc.abstractmethod
def read(self, key):
"""Obtain bundle data for a key.
Returns None if the bundle isn't known.
Throws BundleReadException
The returned object should be a file object supporting read()
and close().
"""
class filebundlestore:
"""bundle store in filesystem
meant for storing bundles somewhere on disk and on network filesystems
"""
def __init__(self, ui, repo):
self.ui = ui
self.repo = repo
self.storepath = ui.configpath(b'scratchbranch', b'storepath')
if not self.storepath:
self.storepath = self.repo.vfs.join(
b"scratchbranches", b"filebundlestore"
)
if not os.path.exists(self.storepath):
os.makedirs(self.storepath)
def _dirpath(self, hashvalue):
"""First two bytes of the hash are the name of the upper
level directory, next two bytes are the name of the
next level directory"""
return os.path.join(self.storepath, hashvalue[0:2], hashvalue[2:4])
def _filepath(self, filename):
return os.path.join(self._dirpath(filename), filename)
def write(self, data):
filename = hex(hashutil.sha1(data).digest())
dirpath = self._dirpath(filename)
if not os.path.exists(dirpath):
os.makedirs(dirpath)
with open(self._filepath(filename), b'wb') as f:
f.write(data)
return filename
def read(self, key):
try:
with open(self._filepath(key), b'rb') as f:
return f.read()
except IOError:
return None
def format_placeholders_args(args, filename=None, handle=None):
"""Formats `args` with Infinitepush replacements.
Hack to get `str.format()`-ed strings working in a BC way with
bytes.
"""
formatted_args = []
for arg in args:
if filename and arg == b'{filename}':
formatted_args.append(filename)
elif handle and arg == b'{handle}':
formatted_args.append(handle)
else:
formatted_args.append(arg)
return formatted_args
class externalbundlestore(abstractbundlestore):
def __init__(self, put_binary, put_args, get_binary, get_args):
"""
`put_binary` - path to binary file which uploads bundle to external
storage and prints key to stdout
`put_args` - format string with additional args to `put_binary`
{filename} replacement field can be used.
`get_binary` - path to binary file which accepts filename and key
(in that order), downloads bundle from store and saves it to file
`get_args` - format string with additional args to `get_binary`.
{filename} and {handle} replacement field can be used.
"""
self.put_args = put_args
self.get_args = get_args
self.put_binary = put_binary
self.get_binary = get_binary
def _call_binary(self, args):
p = subprocess.Popen(
pycompat.rapply(procutil.tonativestr, args),
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
close_fds=True,
)
stdout, stderr = p.communicate()
returncode = p.returncode
return returncode, stdout, stderr
def write(self, data):
# Won't work on windows because you can't open file second time without
# closing it
# TODO: rewrite without str.format() and replace NamedTemporaryFile()
# with pycompat.namedtempfile()
with pycompat.namedtempfile() as temp:
temp.write(data)
temp.flush()
temp.seek(0)
formatted_args = format_placeholders_args(
self.put_args, filename=temp.name
)
returncode, stdout, stderr = self._call_binary(
[self.put_binary] + formatted_args
)
if returncode != 0:
raise BundleWriteException(
b'Failed to upload to external store: %s' % stderr
)
stdout_lines = stdout.splitlines()
if len(stdout_lines) == 1:
return stdout_lines[0]
else:
raise BundleWriteException(
b'Bad output from %s: %s' % (self.put_binary, stdout)
)
def read(self, handle):
# Won't work on windows because you can't open file second time without
# closing it
with pycompat.namedtempfile() as temp:
formatted_args = format_placeholders_args(
self.get_args, filename=temp.name, handle=handle
)
returncode, stdout, stderr = self._call_binary(
[self.get_binary] + formatted_args
)
if returncode != 0:
raise BundleReadException(
b'Failed to download from external store: %s' % stderr
)
return temp.read()