1
# (C) 2005 Canonical Ltd
3
# This program is free software; you can redistribute it and/or modify
4
# it under the terms of the GNU General Public License as published by
5
# the Free Software Foundation; either version 2 of the License, or
6
# (at your option) any later version.
8
# This program is distributed in the hope that it will be useful,
9
# but WITHOUT ANY WARRANTY; without even the implied warranty of
10
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
11
# GNU General Public License for more details.
13
# You should have received a copy of the GNU General Public License
14
# along with this program; if not, write to the Free Software
15
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
17
import stat, os, sha, time
18
from binascii import b2a_qp, a2b_qp
20
from trace import mutter
21
from errors import BzrError
24
"""File stat cache to speed up tree comparisons.
26
This module basically gives a quick way to find the SHA-1 and related
27
information of a file in the working directory, without actually
28
reading and hashing the whole file.
30
This is done by maintaining a cache indexed by a file fingerprint of
31
(path, size, mtime, ctime, ino, dev) pointing to the SHA-1. If the
32
fingerprint has changed, we assume the file content has not changed
33
either and the SHA-1 is therefore the same.
35
If any of the fingerprint fields have changed then the file content
36
*may* have changed, or it may not have. We need to reread the file
37
contents to make sure, but this is not visible to the user or
38
higher-level code (except as a delay of course).
40
The mtime and ctime are stored with nanosecond fields, but not all
41
filesystems give this level of precision. There is therefore a
42
possible race: the file might be modified twice within a second
43
without changing the size or mtime, and a SHA-1 cached from the first
44
version would be wrong. We handle this by not recording a cached hash
45
for any files which were modified in the current second and that
46
therefore have the chance to change again before the second is up.
48
The only known hole in this design is if the system clock jumps
49
backwards crossing invocations of bzr. Please don't do that; use ntp
50
to gradually adjust your clock or don't use bzr over the step.
52
At the moment this is stored in a simple textfile; it might be nice
55
The cache is represented as a map from file_id to a tuple of (file_id,
56
sha1, path, size, mtime, ctime, ino, dev).
71
def fingerprint(abspath):
73
fs = os.lstat(abspath)
75
# might be missing, etc
78
if stat.S_ISDIR(fs.st_mode):
81
return (fs.st_size, fs.st_mtime,
82
fs.st_ctime, fs.st_ino, fs.st_dev)
85
def _write_cache(basedir, entry_iter, dangerfiles):
86
from atomicfile import AtomicFile
88
cachefn = os.path.join(basedir, '.bzr', 'stat-cache')
89
outf = AtomicFile(cachefn, 'wb', 'utf-8')
91
for entry in entry_iter:
92
if entry[0] in dangerfiles:
94
outf.write(entry[0] + ' ' + entry[1] + ' ')
95
outf.write(b2a_qp(entry[2], True))
96
outf.write(' %d %d %d %d %d\n' % entry[3:])
104
def load_cache(basedir):
110
cachefn = os.path.join(basedir, '.bzr', 'stat-cache')
111
cachefile = codecs.open(cachefn, 'r', 'utf-8')
119
raise BzrError("duplicated file_id in cache: {%s}" % file_id)
120
cache[file_id] = (f[0], f[1], a2b_qp(f[2])) + tuple([long(x) for x in f[3:]])
126
def _files_from_inventory(inv):
127
for path, ie in inv.iter_entries():
128
if ie.kind != 'file':
130
yield ie.file_id, path
134
def update_cache(basedir, inv, flush=False):
135
"""Update and return the cache for the branch.
137
The returned cache may contain entries that have not been written
138
to disk for files recently touched.
140
flush -- discard any previous cache and recalculate from scratch.
144
# TODO: It's supposed to be faster to stat the files in order by inum.
145
# We don't directly know the inum of the files of course but we do
146
# know where they were last sighted, so we can sort by that.
148
assert isinstance(flush, bool)
152
cache = load_cache(basedir)
153
return _update_cache_from_list(basedir, cache, _files_from_inventory(inv))
157
def _update_cache_from_list(basedir, cache, to_update):
158
"""Update and return the cache for given files.
160
cache -- Previously cached values to be validated.
162
to_update -- Sequence of (file_id, path) pairs to check.
167
stat_cnt = missing_cnt = hardcheck = change_cnt = 0
169
# files that have been recently touched and can't be
170
# committed to a persistent cache yet.
173
now = int(time.time())
175
## mutter('update statcache under %r' % basedir)
176
for file_id, path in to_update:
177
abspath = os.path.join(basedir, path)
178
fp = fingerprint(abspath)
181
cacheentry = cache.get(file_id)
183
if fp == None: # not here
190
if (fp[FP_MTIME] >= now) or (fp[FP_CTIME] >= now):
191
dangerfiles.add(file_id)
193
if cacheentry and (cacheentry[3:] == fp):
194
continue # all stat fields unchanged
198
dig = sha.new(file(abspath, 'rb').read()).hexdigest()
200
# We update the cache even if the digest has not changed from
201
# last time we looked, so that the fingerprint fields will
203
cacheentry = (file_id, dig, path) + fp
204
cache[file_id] = cacheentry
207
mutter('statcache: statted %d files, read %d files, %d changed, %d dangerous, '
209
% (stat_cnt, hardcheck, change_cnt, len(dangerfiles), len(cache)))
212
mutter('updating on-disk statcache')
213
_write_cache(basedir, cache.itervalues(), dangerfiles)
added
removed
bzrlib/statcache.py
