##// END OF EJS Templates
exchangev2: fetch manifest revisions...
exchangev2: fetch manifest revisions Now that the server has support for retrieving manifest data, we can implement the client bits to call it. We teach the changeset fetching code to capture the manifest revisions that are encountered on incoming changesets. We then feed this into a new function which filters out known manifests and then batches up manifest data requests to the server. This is different from the previous wire protocol in a few notable ways. First, the client fetches manifest data separately and explicitly. Before, we'd ask the server for data pertaining to some changesets (via a "getbundle" command) and manifests (and files) would be sent automatically. Providing an API for looking up just manifest data separately gives clients much more flexibility for manifest management. For example, a client may choose to only fetch manifest data on demand instead of prefetching it (i.e. partial clone). Second, we send N commands to the server for manifest retrieval instead of 1. This property has a few nice side-effects. One is that the deterministic nature of the requests lends itself to server-side caching. For example, say the remote has 50,000 manifests. If the server is configured to cache responses, each time a new commit arrives, you will have a cache miss and need to regenerate all outgoing data. But if you makes N requests requesting 10,000 manifests each, a new commit will still yield cache hits on the initial, unchanged manifest batches/requests. A derived benefit from these properties is that resumable clone is conceptually simpler to implement. When making a monolithic request for all of the repository data, recovering from an interrupted clone is hard because the server was in the driver's seat and was maintaining state about all the data that needed transferred. With the client driving fetching, the client can persist the set of unfetched entities and retry/resume a fetch if something goes wrong. Or we can fetch all data N changesets at a time and slowly build up a repository. This approach is drastically easier to implement when we have server APIs exposing low-level repository primitives (such as manifests and files). We don't yet support tree manifests. But it should be possible to implement that with the existing wire protocol command. Differential Revision: https://phab.mercurial-scm.org/D4489

File last commit:

r34648:dacfcdd8 default
r39674:d292328e default
Show More
osutil.py
102 lines | 3.5 KiB | text/x-python | PythonLexer
# osutil.py - CFFI version of osutil.c
#
# Copyright 2016 Maciej Fijalkowski <fijall@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.
from __future__ import absolute_import
import os
import stat as statmod
from ..pure.osutil import *
from .. import (
pycompat,
)
if pycompat.isdarwin:
from . import _osutil
ffi = _osutil.ffi
lib = _osutil.lib
listdir_batch_size = 4096
# tweakable number, only affects performance, which chunks
# of bytes do we get back from getattrlistbulk
attrkinds = [None] * 20 # we need the max no for enum VXXX, 20 is plenty
attrkinds[lib.VREG] = statmod.S_IFREG
attrkinds[lib.VDIR] = statmod.S_IFDIR
attrkinds[lib.VLNK] = statmod.S_IFLNK
attrkinds[lib.VBLK] = statmod.S_IFBLK
attrkinds[lib.VCHR] = statmod.S_IFCHR
attrkinds[lib.VFIFO] = statmod.S_IFIFO
attrkinds[lib.VSOCK] = statmod.S_IFSOCK
class stat_res(object):
def __init__(self, st_mode, st_mtime, st_size):
self.st_mode = st_mode
self.st_mtime = st_mtime
self.st_size = st_size
tv_sec_ofs = ffi.offsetof("struct timespec", "tv_sec")
buf = ffi.new("char[]", listdir_batch_size)
def listdirinternal(dfd, req, stat, skip):
ret = []
while True:
r = lib.getattrlistbulk(dfd, req, buf, listdir_batch_size, 0)
if r == 0:
break
if r == -1:
raise OSError(ffi.errno, os.strerror(ffi.errno))
cur = ffi.cast("val_attrs_t*", buf)
for i in range(r):
lgt = cur.length
assert lgt == ffi.cast('uint32_t*', cur)[0]
ofs = cur.name_info.attr_dataoffset
str_lgt = cur.name_info.attr_length
base_ofs = ffi.offsetof('val_attrs_t', 'name_info')
name = str(ffi.buffer(ffi.cast("char*", cur) + base_ofs + ofs,
str_lgt - 1))
tp = attrkinds[cur.obj_type]
if name == "." or name == "..":
continue
if skip == name and tp == statmod.S_ISDIR:
return []
if stat:
mtime = cur.mtime.tv_sec
mode = (cur.accessmask & ~lib.S_IFMT)| tp
ret.append((name, tp, stat_res(st_mode=mode, st_mtime=mtime,
st_size=cur.datalength)))
else:
ret.append((name, tp))
cur = ffi.cast("val_attrs_t*", int(ffi.cast("intptr_t", cur))
+ lgt)
return ret
def listdir(path, stat=False, skip=None):
req = ffi.new("struct attrlist*")
req.bitmapcount = lib.ATTR_BIT_MAP_COUNT
req.commonattr = (lib.ATTR_CMN_RETURNED_ATTRS |
lib.ATTR_CMN_NAME |
lib.ATTR_CMN_OBJTYPE |
lib.ATTR_CMN_ACCESSMASK |
lib.ATTR_CMN_MODTIME)
req.fileattr = lib.ATTR_FILE_DATALENGTH
dfd = lib.open(path, lib.O_RDONLY, 0)
if dfd == -1:
raise OSError(ffi.errno, os.strerror(ffi.errno))
try:
ret = listdirinternal(dfd, req, stat, skip)
finally:
try:
lib.close(dfd)
except BaseException:
pass # we ignore all the errors from closing, not
# much we can do about that
return ret