1
# Copyright (C) 2005 Canonical Ltd
1
# Copyright (C) 2005 by Canonical Development Ltd
3
3
# This program is free software; you can redistribute it and/or modify
4
4
# it under the terms of the GNU General Public License as published by
5
5
# the Free Software Foundation; either version 2 of the License, or
6
6
# (at your option) any later version.
8
8
# This program is distributed in the hope that it will be useful,
9
9
# but WITHOUT ANY WARRANTY; without even the implied warranty of
10
10
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
11
11
# GNU General Public License for more details.
13
13
# You should have received a copy of the GNU General Public License
14
14
# along with this program; if not, write to the Free Software
15
15
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
17
"""A store that keeps the full text of every version.
18
A store that keeps the full text of every version.
19
20
This store keeps uncompressed versions of the full text. It does not
20
21
do any sort of delta compression.
24
26
import bzrlib.store
25
27
from bzrlib.trace import mutter
26
from bzrlib.errors import BzrError, NoSuchFile, FileExists
28
from bzrlib.errors import BzrError
29
30
from cStringIO import StringIO
32
class TextStore(bzrlib.store.TransportStore):
31
from stat import ST_SIZE
33
class TextStore(bzrlib.store.Store):
33
34
"""Store that holds files indexed by unique names.
35
36
Files can be added, but not modified once they are in. Typically
39
40
Files are stored uncompressed, with no delta compression.
42
def _add_compressed(self, fn, f):
43
from cStringIO import StringIO
44
from bzrlib.osutils import pumpfile
44
def __init__(self, basedir):
45
super(TextStore, self).__init__(basedir)
47
def _check_fileid(self, fileid):
48
if not isinstance(fileid, basestring):
49
raise TypeError('Fileids should be a string type: %s %r' % (type(fileid), fileid))
50
if '\\' in fileid or '/' in fileid:
51
raise ValueError("invalid store id %r" % fileid)
53
def _relpath(self, fileid):
54
self._check_fileid(fileid)
57
def add(self, f, fileid):
58
"""Add contents of a file into the store.
60
f -- A file-like object, or string
62
mutter("add store entry %r" % (fileid))
64
fn = self._relpath(fileid)
65
if self._transport.has(fn):
66
raise BzrError("store %r already contains id %r" % (self._transport.base, fileid))
68
self._transport.put(fn, f)
70
def _do_copy(self, other, to_copy, pb, permit_failure=False):
71
if isinstance(other, TextStore):
72
return self._copy_multi_text(other, to_copy, pb,
73
permit_failure=permit_failure)
74
return super(TextStore, self)._do_copy(other, to_copy,
75
pb, permit_failure=permit_failure)
77
def _copy_multi_text(self, other, to_copy, pb,
78
permit_failure=False):
79
# Because of _transport, we can no longer assume
80
# that they are on the same filesystem, we can, however
81
# assume that we only need to copy the exact bytes,
82
# we don't need to process the files.
87
for fileid, has in zip(to_copy, other.has(to_copy)):
89
new_to_copy.add(fileid)
93
#mutter('_copy_multi_text copying %s, failed %s' % (to_copy, failed))
95
paths = [self._relpath(fileid) for fileid in to_copy]
96
count = other._transport.copy_to(paths, self._transport, pb=pb)
97
assert count == len(to_copy)
100
def __contains__(self, fileid):
102
fn = self._relpath(fileid)
103
return self._transport.has(fn)
105
def has(self, fileids, pb=None):
106
"""Return True/False for each entry in fileids.
108
:param fileids: A List or generator yielding file ids.
109
:return: A generator or list returning True/False for each entry.
111
relpaths = (self._relpath(fid) for fid in fileids)
112
return self._transport.has_multi(relpaths, pb=pb)
114
def get(self, fileids, permit_failure=False, pb=None):
115
"""Return a set of files, one for each requested entry.
46
if isinstance(f, basestring):
50
gf = gzip.GzipFile(mode='wb', fileobj=sio)
51
# if pumpfile handles files that don't fit in ram,
52
# so will this function
56
self._try_put(fn, sio)
58
def _add(self, fn, f):
60
self._add_compressed(fn, f)
64
def _try_put(self, fn, f):
66
self._transport.put_file(fn, f, mode=self._file_mode)
68
if not self._prefixed:
71
self._transport.mkdir(os.path.dirname(fn), mode=self._dir_mode)
74
self._transport.put_file(fn, f, mode=self._file_mode)
77
if fn.endswith('.gz'):
78
return self._get_compressed(fn)
80
return self._transport.get(fn)
82
def _copy_one(self, fileid, suffix, other, pb):
83
# TODO: Once the copy_to interface is improved to allow a source
84
# and destination targets, then we can always do the copy
85
# as long as other is a TextStore
86
if not (isinstance(other, TextStore)
87
and other._prefixed == self._prefixed):
88
return super(TextStore, self)._copy_one(fileid, suffix, other, pb)
90
mutter('_copy_one: %r, %r', fileid, suffix)
91
path = other._get_name(fileid, suffix)
93
raise KeyError(fileid + '-' + str(suffix))
96
result = other._transport.copy_to([path], self._transport,
99
if not self._prefixed:
102
self._transport.mkdir(self.hash_prefix(fileid)[:-1], mode=self._dir_mode)
105
result = other._transport.copy_to([path], self._transport,
106
mode=self._file_mode)
109
raise BzrError('Unable to copy file: %r' % (path,))
111
def _get_compressed(self, filename):
117
TODO: Write some tests to make sure that permit_failure is
120
TODO: What should the exception be for a missing file?
121
KeyError, or NoSuchFile?
124
# This next code gets a bit hairy because it can allow
125
# to not request a file which doesn't seem to exist.
126
# Also, the same fileid may be requested twice, so we
127
# can't just build up a map.
128
rel_paths = [self._relpath(fid) for fid in fileids]
133
for path, has in zip(rel_paths,
134
self._transport.has_multi(rel_paths)):
136
existing_paths.append(path)
137
is_requested.append(True)
139
is_requested.append(False)
140
#mutter('Retrieving %s out of %s' % (existing_paths, rel_paths))
142
#mutter('Retrieving all %s' % (rel_paths, ))
143
existing_paths = rel_paths
144
is_requested = [True for x in rel_paths]
147
for f in self._transport.get_multi(existing_paths, pb=pb):
148
assert count < len(is_requested)
149
while not is_requested[count]:
155
while count < len(is_requested):
160
# TODO: case-insensitive?
161
for f in self._transport.list_dir('.'):
165
return len([f for f in self._transport.list_dir('.')])
168
def __getitem__(self, fileid):
112
169
"""Returns a file reading from a particular entry."""
113
f = self._transport.get(filename)
114
# gzip.GzipFile.read() requires a tell() function
115
# but some transports return objects that cannot seek
116
# so buffer them in a StringIO instead
117
if getattr(f, 'tell', None) is not None:
118
return gzip.GzipFile(mode='rb', fileobj=f)
120
from cStringIO import StringIO
121
sio = StringIO(f.read())
122
return gzip.GzipFile(mode='rb', fileobj=sio)
170
fn = self._relpath(fileid)
171
# This will throw if the file doesn't exist.
173
f = self._transport.get(fn)
175
raise KeyError('This store (%s) does not contain %s' % (self, fileid))
180
def total_size(self):
181
"""Return (count, bytes)
183
This is the (compressed) size stored on disk, not the size of
187
relpaths = [self._relpath(fid) for fid in self]
188
for st in self._transport.stat_multi(relpaths):
194
class ScratchTextStore(TextStore):
195
"""Self-destructing test subclass of TextStore.
197
The Store only exists for the lifetime of the Python object.
198
Obviously you should not put anything precious in it.
201
from transport import transport
202
super(ScratchTextStore, self).__init__(transport(tempfile.mkdtemp()))
205
self._transport.delete_multi(self._transport.list_dir('.'))
206
os.rmdir(self._transport.base)
207
mutter("%r destroyed" % self)