1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
|
"""Implementation of multiparent diffs for versionedfile-like storage
Provides mp-regen and mp-extract commands.
Focus is on comparing size/performance to knits.
"""
from bzrlib.lazy_import import lazy_import
lazy_import(globals(), """
import (
errno,
os,
sys,
time,
)
from bzrlib import (
commands,
urlutils
)
from bzrlib.workingtree import WorkingTree
from bzrlib.tests import TestUtil
from bzrlib.plugins.multiparent.multiparent import (
MultiVersionedFile,
MultiMemoryVersionedFile,
)
""")
class cmd_mp_regen(commands.Command):
"""Generate a multiparent versionedfile"""
hidden = True
takes_args = ['file?']
takes_options = [commands.Option('sync-snapshots',
help='Snapshots follow source.'),
commands.Option('snapshot-interval', type=int,
help='Take snapshots every x revisions.'),
commands.Option('outfile', type=unicode,
help='Write pseudo-knit to this file.'),
commands.Option('memory', help='Use memory, not disk.'),
commands.Option('extract', help='Test extract time.'),
commands.Option('single', help='Use a single parent.'),
commands.Option('verify', help='Verify added texts.'),
commands.Option('cache', help='Aggresively cache.'),
commands.Option('size', help='Aggressive size.'),
commands.Option('build', help='Aggressive build.'),
]
hidden = True
def run(self, file=None, sync_snapshots=False, snapshot_interval=26,
lsprof_timed=False, dump=False, extract=False, single=False,
verify=False, outfile=None, memory=False, cache=False,
size=False, build=False):
file_weave = get_file_weave(file)
url = file_weave.transport.abspath(file_weave.filename)
sys.stderr.write('Importing: %s\n' % \
urlutils.local_path_from_url(url))
if sync_snapshots:
sys.stderr.write('Snapshots follow input\n')
else:
sys.stderr.write('Snapshot interval: %d\n' % snapshot_interval)
if not memory:
if outfile is None:
filename = 'pknit'
else:
filename = outfile
vf = MultiVersionedFile(filename, snapshot_interval)
else:
vf = MultiMemoryVersionedFile(snapshot_interval)
vf.destroy()
old_snapshots = set(r for r in file_weave.versions() if
file_weave._index.get_method(r) == 'fulltext')
if sync_snapshots:
to_sync = old_snapshots
elif size or build:
assert memory
to_sync = set()
else:
to_sync = vf.select_snapshots(file_weave)
sys.stderr.write("%d fulltext(s)\n" % len(old_snapshots))
sys.stderr.write("%d planned snapshots\n" % len(to_sync))
try:
vf.import_versionedfile(file_weave, to_sync, single_parent=single,
verify=verify, no_cache=not cache)
if size:
snapshots = vf.select_by_size(len(old_snapshots))
for version_id in snapshots:
vf.make_snapshot(version_id)
if build:
ranking = vf.get_build_ranking()
snapshots = ranking[:len(old_snapshots) -\
len(vf._snapshots)]
for version_id in snapshots:
vf.make_snapshot(version_id)
except:
vf.destroy()
raise
try:
sys.stderr.write("%d actual snapshots\n" % len(vf._snapshots))
if not cache:
vf.clear_cache()
if memory:
if outfile is not None:
vf_file = MultiVersionedFile(outfile)
vf_file.import_diffs(vf)
else:
vf_file = vf
finally:
if outfile is None:
vf.destroy()
else:
vf_file.save()
class cmd_mp_extract(commands.Command):
"""Test extraction time multiparent knits"""
hidden = True
takes_options = [
commands.Option('lsprof-timed', help='Use lsprof.'),
commands.Option('parallel', help='Extract multiple versions at once.'),
commands.Option('count', help='Number of cycles to do.', type=int),
]
takes_args = ['filename', 'vfile?']
def run(self, filename, vfile=None, lsprof_timed=False, count=1000,
parallel=False):
vf = MultiVersionedFile(filename)
vf.load()
snapshots = [r for r in vf.versions() if vf.get_diff(r).is_snapshot()]
print '%d snapshots' % len(snapshots)
revisions = list(vf.versions())
revisions = revisions[-count:]
print 'Testing extract time of %d revisions' % len(revisions)
if parallel:
revisions_list = [revisions]
else:
revisions_list = [[r] for r in revisions]
start = time.clock()
for revisions in revisions_list:
vf = MultiVersionedFile(filename)
vf.load()
vf.get_line_list(revisions)
sys.stderr.write(time.clock() - start)
sys.stderr.write('\n')
if lsprof_timed:
from bzrlib.lsprof import profile
vf.clear_cache()
ret, stats = profile(vf.get_line_list, revisions_list[-1][-1])
stats.sort()
stats.pprint()
start = time.clock()
for revisions in revisions_list:
file_weave = get_file_weave(vfile)
file_weave.get_line_list(revisions)
sys.stderr.write(time.clock() - start)
sys.stderr.write('\n')
def get_file_weave(filename=None, wt=None):
if filename is None:
wt, path = WorkingTree.open_containing('.')
return wt.branch.repository.get_inventory_weave()
else:
wt, path = WorkingTree.open_containing(filename)
file_id = wt.path2id(path)
bt = wt.branch.repository.revision_tree(wt.last_revision())
return bt.get_weave(file_id)
commands.register_command(cmd_mp_regen)
commands.register_command(cmd_mp_extract)
|