3
# Copyright (C) 2005 Canonical Ltd
5
# This program is free software; you can redistribute it and/or modify
6
# it under the terms of the GNU General Public License as published by
7
# the Free Software Foundation; either version 2 of the License, or
8
# (at your option) any later version.
10
# This program is distributed in the hope that it will be useful,
11
# but WITHOUT ANY WARRANTY; without even the implied warranty of
12
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13
# GNU General Public License for more details.
15
# You should have received a copy of the GNU General Public License
16
# along with this program; if not, write to the Free Software
17
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
19
"""Experiment in converting existing bzr branches to weaves."""
21
# To make this properly useful
23
# 1. assign text version ids, and put those text versions into
24
# the inventory as they're converted.
26
# 2. keep track of the previous version of each file, rather than
27
# just using the last one imported
29
# 3. assign entry versions when files are added, renamed or moved.
31
# 4. when merged-in versions are observed, walk down through them
32
# to discover everything, then commit bottom-up
34
# 5. track ancestry as things are merged in, and commit that in each
37
# Perhaps it's best to first walk the whole graph and make a plan for
38
# what should be imported in what order? Need a kind of topological
39
# sort of all revisions. (Or do we, can we just before doing a revision
40
# see that all its parents have either been converted or abandoned?)
51
import hotshot, hotshot.stats
56
from bzrlib.branch import Branch, find_branch
57
from bzrlib.revfile import Revfile
58
from bzrlib.weave import Weave
59
from bzrlib.weavefile import read_weave, write_weave
60
from bzrlib.progress import ProgressBar
61
from bzrlib.atomicfile import AtomicFile
62
from bzrlib.xml4 import serializer_v4
63
from bzrlib.xml5 import serializer_v5
64
from bzrlib.trace import mutter, note, warning, enable_default_logging
68
class Convert(object):
71
self.converted_revs = set()
72
self.absent_revisions = set()
79
enable_default_logging()
80
self.pb = ProgressBar()
81
self.inv_weave = Weave('__inventory')
82
self.anc_weave = Weave('__ancestry')
86
# holds in-memory weaves for all files
89
b = self.branch = Branch('.', relax_version_check=True)
92
rev_history = b.revision_history()
96
# todo is a stack holding the revisions we still need to process;
97
# appending to it adds new highest-priority revisions
98
self.todo = rev_history[:]
100
self.total_revs = len(self.todo)
103
self._convert_one_rev(self.todo.pop())
106
print 'upgraded to weaves:'
107
print ' %6d revisions and inventories' % len(self.converted_revs)
108
print ' %6d absent revisions removed' % len(self.absent_revisions)
109
print ' %6d texts' % self.text_count
111
self._write_all_weaves()
114
def _write_all_weaves(self):
116
return ############################################
117
# TODO: commit them all atomically at the end, not one by one
118
write_atomic_weave(self.inv_weave, 'weaves/inventory.weave')
119
write_atomic_weave(self.anc_weave, 'weaves/ancestry.weave')
120
for file_id, file_weave in text_weaves.items():
121
self.pb.update('writing weave', i, len(text_weaves))
122
write_atomic_weave(file_weave, 'weaves/%s.weave' % file_id)
128
def _convert_one_rev(self, rev_id):
129
self._bump_progress()
132
if rev_id not in b.revision_store:
134
note('revision {%s} not present in branch; '
135
'will not be converted',
137
self.absent_revisions.add(rev_id)
140
rev_xml = b.revision_store[rev_id].read()
141
inv_xml = b.inventory_store[rev_id].read()
143
rev = serializer_v4.read_revision_from_string(rev_xml)
144
inv = serializer_v4.read_inventory_from_string(inv_xml)
146
# see if parents need to be done first
147
for parent_id in [x.revision_id for x in rev.parents]:
148
if parent_id not in self.converted_revs:
149
self.todo.append(parent_id)
151
self.converted_revs.add(rev_id)
153
return ##########################################
155
new_idx = self.inv_weave.add(rev_id, inv_parents, inv_xml)
156
inv_parents = [new_idx]
158
tree = b.revision_tree(rev_id)
161
# for each file in the inventory, put it into its own revfile
164
if ie.kind != 'file':
166
if last_text_sha.get(file_id) == ie.text_sha1:
169
last_text_sha[file_id] = ie.text_sha1
171
# new text (though possibly already stored); need to store it
172
text_lines = tree.get_file(file_id).readlines()
174
# if the file's created for the first time in this
175
# revision then make a new weave; else find the old one
176
if file_id not in text_weaves:
177
text_weaves[file_id] = Weave()
179
w = text_weaves[file_id]
181
# base the new text version off whatever was last
182
# (actually it'd be better to track this, to allow for
183
# files that are deleted and then reappear)
190
w.add(rev_id, parents, text_lines)
195
def _bump_progress(self):
196
self.pb.update('converting revisions',
197
len(self.converted_revs),
201
def write_atomic_weave(weave, filename):
202
inv_wf = AtomicFile(filename)
204
write_weave(weave, inv_wf)
212
def profile_convert():
213
prof_f = tempfile.NamedTemporaryFile()
215
prof = hotshot.Profile(prof_f.name)
217
prof.runcall(Convert)
220
stats = hotshot.stats.load(prof_f.name)
222
stats.sort_stats('time')
223
# XXX: Might like to write to stderr or the trace file instead but
224
# print_stats seems hardcoded to stdout
225
stats.print_stats(20)
228
if '-p' in sys.argv[1:]:
added
removed
tools/history2weaves.py
