1
# Copyright (C) 2005, 2006 by Canonical Ltd
2
# Written by Robert Collins <robert.collins@canonical.com>
4
# This program is free software; you can redistribute it and/or modify
5
# it under the terms of the GNU General Public License as published by
6
# the Free Software Foundation; either version 2 of the License, or
7
# (at your option) any later version.
9
# This program is distributed in the hope that it will be useful,
10
# but WITHOUT ANY WARRANTY; without even the implied warranty of
11
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
12
# GNU General Public License for more details.
14
# You should have received a copy of the GNU General Public License
15
# along with this program; if not, write to the Free Software
16
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
18
"""Bzrlib specific gzip tunings. We plan to feed these to the upstream gzip."""
20
# make GzipFile faster:
22
from gzip import U32, LOWU32, FEXTRA, FCOMMENT, FNAME, FHCRC
27
# we want a \n preserved, break on \n only splitlines.
30
__all__ = ["GzipFile"]
33
class GzipFile(gzip.GzipFile):
34
"""Knit tuned version of GzipFile.
36
This is based on the following lsprof stats:
37
python 2.4 stock GzipFile write:
38
58971 0 5644.3090 2721.4730 gzip:193(write)
39
+58971 0 1159.5530 1159.5530 +<built-in method compress>
40
+176913 0 987.0320 987.0320 +<len>
41
+58971 0 423.1450 423.1450 +<zlib.crc32>
42
+58971 0 353.1060 353.1060 +<method 'write' of 'cStringIO.
45
58971 0 4477.2590 2103.1120 bzrlib.knit:1250(write)
46
+58971 0 1297.7620 1297.7620 +<built-in method compress>
47
+58971 0 406.2160 406.2160 +<zlib.crc32>
48
+58971 0 341.9020 341.9020 +<method 'write' of 'cStringIO.
50
+58971 0 328.2670 328.2670 +<len>
53
Yes, its only 1.6 seconds, but they add up.
56
def _add_read_data(self, data):
58
# temp var for len(data) and switch to +='s.
61
self.crc = zlib.crc32(data, self.crc)
63
self.extrasize += len_data
66
def _read(self, size=1024):
67
# various optimisations:
68
# reduces lsprof count from 2500 to
69
# 8337 calls in 1272, 365 internal
70
if self.fileobj is None:
71
raise EOFError, "Reached EOF"
74
# If the _new_member flag is set, we have to
75
# jump to the next member, if there is one.
77
# First, check if we're at the end of the file;
78
# if so, it's time to stop; no more members to read.
79
next_header_bytes = self.fileobj.read(10)
80
if next_header_bytes == '':
81
raise EOFError, "Reached EOF"
84
self._read_gzip_header(next_header_bytes)
85
self.decompress = zlib.decompressobj(-zlib.MAX_WBITS)
86
self._new_member = False
88
# Read a chunk of data from the file
89
buf = self.fileobj.read(size)
91
# If the EOF has been reached, flush the decompression object
92
# and mark this object as finished.
95
self._add_read_data(self.decompress.flush())
96
assert len(self.decompress.unused_data) >= 8, "what does flush do?"
97
self._gzip_tail = self.decompress.unused_data[0:8]
99
# tell the driving read() call we have stuffed all the data
101
raise EOFError, 'Reached EOF'
103
self._add_read_data(self.decompress.decompress(buf))
105
if self.decompress.unused_data != "":
106
# Ending case: we've come to the end of a member in the file,
107
# so seek back to the start of the data for the next member which
108
# is the length of the decompress objects unused data - the first
109
# 8 bytes for the end crc and size records.
111
# so seek back to the start of the unused data, finish up
112
# this member, and read a new gzip header.
113
# (The number of bytes to seek back is the length of the unused
114
# data, minus 8 because those 8 bytes are part of this member.
115
seek_length = len (self.decompress.unused_data) - 8
117
# we read too much data
118
self.fileobj.seek(-seek_length, 1)
119
self._gzip_tail = self.decompress.unused_data[0:8]
120
elif seek_length < 0:
121
# we haven't read enough to check the checksum.
122
assert -8 < seek_length, "too great a seek."
123
buf = self.fileobj.read(-seek_length)
124
self._gzip_tail = self.decompress.unused_data + buf
126
self._gzip_tail = self.decompress.unused_data
128
# Check the CRC and file size, and set the flag so we read
129
# a new member on the next call
131
self._new_member = True
134
"""tuned to reduce function calls and eliminate file seeking:
136
reduces lsprof count from 800 to 288
138
avoid U32 call by using struct format L
141
# We've read to the end of the file, so we should have 8 bytes of
142
# unused data in the decompressor. If we don't, there is a corrupt file.
143
# We use these 8 bytes to calculate the CRC and the recorded file size.
144
# We then check the that the computed CRC and size of the
145
# uncompressed data matches the stored values. Note that the size
146
# stored is the true file size mod 2**32.
147
assert len(self._gzip_tail) == 8, "gzip trailer is incorrect length."
148
crc32, isize = struct.unpack("<LL", self._gzip_tail)
149
# note that isize is unsigned - it can exceed 2GB
150
if crc32 != U32(self.crc):
151
raise IOError, "CRC check failed %d %d" % (crc32, U32(self.crc))
152
elif isize != LOWU32(self.size):
153
raise IOError, "Incorrect length of data produced"
155
def _read_gzip_header(self, bytes=None):
156
"""Supply bytes if the minimum header size is already read.
158
:param bytes: 10 bytes of header data.
160
"""starting cost: 300 in 3998
161
15998 reads from 3998 calls
165
bytes = self.fileobj.read(10)
167
if magic != '\037\213':
168
raise IOError, 'Not a gzipped file'
169
method = ord(bytes[2:3])
171
raise IOError, 'Unknown compression method'
172
flag = ord(bytes[3:4])
173
# modtime = self.fileobj.read(4) (bytes [4:8])
174
# extraflag = self.fileobj.read(1) (bytes[8:9])
175
# os = self.fileobj.read(1) (bytes[9:10])
176
# self.fileobj.read(6)
179
# Read & discard the extra field, if present
180
xlen = ord(self.fileobj.read(1))
181
xlen = xlen + 256*ord(self.fileobj.read(1))
182
self.fileobj.read(xlen)
184
# Read and discard a null-terminated string containing the filename
186
s = self.fileobj.read(1)
187
if not s or s=='\000':
190
# Read and discard a null-terminated string containing a comment
192
s = self.fileobj.read(1)
193
if not s or s=='\000':
196
self.fileobj.read(2) # Read & discard the 16-bit header CRC
198
def readline(self, size=-1):
199
"""Tuned to remove buffer length calls in _unread and...
201
also removes multiple len(c) calls, inlines _unread,
202
total savings - lsprof 5800 to 5300
205
8176 calls to read() in 1684
206
changing the min chunk size to 200 halved all the cache misses
207
leading to a drop to:
209
4168 call to read() in 1646
210
- i.e. just reduced the function call overhead. May be worth
213
if size < 0: size = sys.maxint
215
readsize = min(200, size) # Read from the file in small chunks
218
return "".join(bufs) # Return resulting line
221
c = self.read(readsize)
222
# number of bytes read
226
# We set i=size to break out of the loop under two
227
# conditions: 1) there's no newline, and the chunk is
228
# larger than size, or 2) there is a newline, but the
229
# resulting line would be longer than 'size'.
230
if i==-1 and len_c > size: i=size-1
231
elif size <= i: i = size -1
233
if i >= 0 or c == '':
234
# if i>= 0 we have a newline or have triggered the above
235
# if size is not None condition.
236
# if c == '' its EOF.
237
bufs.append(c[:i+1]) # Add portion of last chunk
238
# -- inlined self._unread --
239
## self._unread(c[i+1:], len_c - i) # Push back rest of chunk
240
self.extrabuf = c[i+1:] + self.extrabuf
241
self.extrasize = len_c - i + self.extrasize
242
self.offset -= len_c - i
243
# -- end inlined self._unread --
244
return ''.join(bufs) # Return resulting line
246
# Append chunk to list, decrease 'size',
249
readsize = min(size, readsize * 2)
251
def readlines(self, sizehint=0):
252
# optimise to avoid all the buffer manipulation
253
# lsprof changed from:
254
# 4168 calls in 5472 with 32000 calls to readline()
257
# Negative numbers result in reading all the lines
260
content = self.read(sizehint)
261
return bzrlib.osutils.split_lines(content)
263
def _unread(self, buf, len_buf=None):
264
"""tuned to remove unneeded len calls.
266
because this is such an inner routine in readline, and readline is
267
in many inner loops, this has been inlined into readline().
269
The len_buf parameter combined with the reduction in len calls dropped
270
the lsprof ms count for this routine on my test data from 800 to 200 -
275
self.extrabuf = buf + self.extrabuf
276
self.extrasize = len_buf + self.extrasize
277
self.offset -= len_buf
279
def write(self, data):
280
if self.mode != gzip.WRITE:
282
raise IOError(errno.EBADF, "write() on read-only GzipFile object")
284
if self.fileobj is None:
285
raise ValueError, "write() on closed GzipFile object"
288
self.size = self.size + data_len
289
self.crc = zlib.crc32(data, self.crc)
290
self.fileobj.write( self.compress.compress(data) )
291
self.offset += data_len
293
def writelines(self, lines):
294
# profiling indicated a significant overhead
295
# calling write for each line.
296
# this batch call is a lot faster :).
297
# (4 seconds to 1 seconds for the sample upgrades I was testing).
298
self.write(''.join(lines))
added
removed
bzrlib/tuned_gzip.py
