To get this branch, use:
bzr branch
http://gegoxaren.bato24.eu/bzr/b-gtk/fix-viz
« back to all changes in this revision
Viewing changes to branchview/linegraph.py
- browse files at revision 464.2.1
- compare with another revision
- download diff
- download tarball
- view history from revision 464.2.1
- Committer: Adrian Wilkins
- Date: 2008-04-24 15:26:14 UTC
- mto: This revision was merged to the branch mainline in revision 470.
- Revision ID: adrian.wilkins@gmail.com-20080424152614-2rnnljbro6vzqvf7
Detect the reserved null: revision in appropriate places.
This removes a huge shower of stack traces that get dumped to console when
you look at the bottom of a log.
This removes a huge shower of stack traces that get dumped to console when
you look at the bottom of a log.
- files added:
- .bzrignore
- annotate
- branchview
- icons
- olive
- po
- preferences
- tests
- viz
- files removed:
- .bzrignore
- files renamed:
- graphcell.py => branchview/graphcell.py
- graph.py => branchview/linegraph.py
- __init__.py => viz/__init__.py
- branchwin.py => viz/branchwin.py
- files modified:
- AUTHORS
added
removed
branchview/linegraph.py
1
#!/usr/bin/python
2
# -*- coding: UTF-8 -*-
3
"""Directed graph production.
4
5
This module contains the code to produce an ordered directed graph of a
6
bzr branch, such as we display in the tree view at the top of the bzrk
7
window.
8
"""
9
10
__copyright__ = "Copyright © 2005 Canonical Ltd."
11
__author__ = "Scott James Remnant <scott@ubuntu.com>"
12
13
14
from bzrlib.errors import NoSuchRevision
15
16
17
class DummyRevision(object):
18
"""Dummy bzr revision.
19
20
Sometimes, especially in older bzr branches, a revision is referenced
21
as the parent of another but not actually present in the branch's store.
22
When this happens we use an instance of this class instead of the real
23
Revision object (which we can't get).
24
"""
25
26
def __init__(self, revid):
27
self.revision_id = revid
28
self.parent_ids = []
29
self.committer = None
30
self.message = self.revision_id
31
32
33
def graph(branch, start):
34
"""Produce a directed graph of a bzr branch.
35
36
Traverses the branch revision tree starting at start and produces an
37
ordered list of revisions such that a revision always comes after
38
any revision it is the parent of. It also tries to make a reasonably
39
not-too-stupid decision whether a parent revision is on the same
40
logical branch, as that information is not available with bzr.
41
42
For each revision it then yields a tuple of (revision, node, lines).
43
If the revision is only referenced in the branch and not present in the
44
store, revision will be a DummyRevision object, otherwise it is the bzr
45
Revision object with the meta-data for the revision.
46
47
Node is a tuple of (column, colour) with column being a zero-indexed
48
column number of the graph that this revision represents and colour
49
being a zero-indexed colour (which doesn't specify any actual colour
50
in particular) to draw the node in.
51
52
Lines is a list of tuples which represent lines you should draw away
53
from the revision, if you also need to draw lines into the revision
54
you should use the lines list from the previous iteration. Each
55
typle in the list is in the form (start, end, colour) with start and
56
end being zero-indexed column numbers and colour as in node.
57
58
It's up to you how to actually draw the nodes and lines (straight,
59
curved, kinked, etc.) and to pick the actual colours for each index.
60
"""
61
revisions = { start: branch.get_revision(start) }
62
distances = { start: 0 }
63
colours = { start: 0 }
64
last_colour = 0
65
66
# Sort the revisions; the fastest way to do this is to visit each node
67
# as few times as possible (by keeping the todo list in a set) and record
68
# the largest distance to it before queuing up the children if we
69
# increased the distance. This produces the sort order we desire
70
todo = set([ start ])
71
while todo:
72
revid = todo.pop()
73
revision = revisions[revid]
74
distance = distances[revid] + 1
75
colour = colours[revid]
76
77
reused = False
78
for parent_id in revision.parent_ids:
79
# Check whether there's any point re-processing this
80
if parent_id in distances and distances[parent_id] >= distance:
81
continue
82
83
# Get the parent from the cache, or put it in the cache
84
try:
85
parent = revisions[parent_id]
86
except KeyError:
87
try:
88
parent = revisions[parent_id] \
89
= branch.get_revision(parent_id)
90
except NoSuchRevision:
91
parent = revisions[parent_id] = DummyRevision(parent_id)
92
93
# Make a guess as to whether this node represents the same
94
# branch, or a new one. Penalise same branches in the distance
95
# stakes to give new ones a chance to appear first as one set.
96
if len(revision.parent_ids) == 1:
97
colours[parent_id] = colour
98
distances[parent_id] = distance
99
elif revision.committer == parent.committer and not reused:
100
colours[parent_id] = colour
101
distances[parent_id] = distance
102
reused = True
103
else:
104
colours[parent_id] = last_colour = last_colour + 1
105
distances[parent_id] = distance + 10
106
107
todo.add(parent_id)
108
109
# Now iterate the revisions again, but this time in list order rather
110
# than traversing the tree, and build up the graph lines. We do this
111
# by keeping a list of "hanging parents", which can only be removed
112
# once we encounter the revision being hung.
113
hanging = [ start ]
114
for revid in sorted(distances, key=distances.get):
115
lines = []
116
node = None
117
118
new_hanging = []
119
for h_idx, hang in enumerate(hanging):
120
if hang == revid:
121
# We've matched a hanging revision, so need to output a node
122
# at this point
123
node = (h_idx, colours[revid])
124
125
# Now we need to hang its parents, we put them at the point
126
# the old column was so anything to the right of this has
127
# to move outwards to make room. We also try and collapse
128
# hangs to keep the graph small.
129
for parent_id in revisions[revid].parent_ids:
130
try:
131
n_idx = new_hanging.index(parent_id)
132
except ValueError:
133
n_idx = len(new_hanging)
134
new_hanging.append(parent_id)
135
lines.append((h_idx, n_idx, colours[parent_id]))
136
else:
137
# Revision keeps on hanging, adjust for any change in the
138
# graph shape and try to collapse hangs to keep the graph
139
# small.
140
try:
141
n_idx = new_hanging.index(hang)
142
except ValueError:
143
n_idx = len(new_hanging)
144
new_hanging.append(hang)
145
lines.append((h_idx, n_idx, colours[hang]))
146
hanging = new_hanging
147
148
yield (revisions[revid], node, lines)
1
# -*- coding: UTF-8 -*-
2
"""Directed graph production.
3
4
This module contains the code to produce an ordered directed graph of a
5
bzr branch, such as we display in the tree view at the top of the bzrk
6
window.
7
"""
8
9
__copyright__ = "Copyright © 2005 Canonical Ltd."
10
__author__ = "Scott James Remnant <scott@ubuntu.com>"
11
12
from bzrlib.tsort import merge_sort
13
14
def linegraph(repository, start_revs, maxnum, broken_line_length = None,
15
graph_data = True, mainline_only = False):
16
"""Produce a directed graph of a bzr repository.
17
18
Returns a tuple of (line_graph, revid_index, columns_len) where
19
* line_graph is a list of tuples of (revid,
20
node,
21
lines,
22
parents,
23
children,
24
revno_sequence),
25
* revid_index is a dict of each revision with the key being the revid, and
26
the value the row index, and
27
* columns_len is the number of columns need to draw the line graph.
28
29
30
Node is a tuple of (column, colour) with column being a zero-indexed
31
column number of the graph that this revision represents and colour
32
being a zero-indexed colour (which doesn't specify any actual colour
33
in particular) to draw the node in.
34
35
Lines is a list of tuples which represent lines you should draw away
36
from the revision, if you also need to draw lines into the revision
37
you should use the lines list from the previous iteration. Each
38
typle in the list is in the form (start, end, colour) with start and
39
end being zero-indexed column numbers and colour as in node.
40
41
It's up to you how to actually draw the nodes and lines (straight,
42
curved, kinked, etc.) and to pick the actual colours for each index.
43
"""
44
45
graph = repository.get_graph()
46
graph_parents = {}
47
graph_children = {}
48
for (revid, parent_revids) in graph.iter_ancestry(start_revs):
49
graph_parents[revid] = parent_revids
50
graph_children[revid] = []
51
52
graph_parents["top:"] = start_revs
53
54
if len(graph_parents)>0:
55
merge_sorted_revisions = merge_sort(
56
graph_parents,
57
"top:",
58
generate_revno=True)
59
else:
60
merge_sorted_revisions = ()
61
62
if mainline_only:
63
merge_sorted_revisions = [elem for elem in merge_sorted_revisions \
64
if len(elem[3])==1 ]
65
66
assert merge_sorted_revisions[0][1] == "top:"
67
merge_sorted_revisions = merge_sorted_revisions[1:]
68
69
revid_index = {}
70
revno_index = {}
71
72
# This will hold an item for each "branch". For a revisions, the revsion
73
# number less the least significant digit is the branch_id, and used as the
74
# key for the dict. Hence revision with the same revsion number less the
75
# least significant digit are considered to be in the same branch line.
76
# e.g.: for revisions 290.12.1 and 290.12.2, the branch_id would be 290.12,
77
# and these two revisions will be in the same branch line. Each value is
78
# a list of rev_indexes in the branch.
79
branch_lines = {}
80
81
linegraph = []
82
83
for (rev_index, (sequence_number,
84
revid,
85
merge_depth,
86
revno_sequence,
87
end_of_merge)) in enumerate(merge_sorted_revisions):
88
if maxnum and rev_index >= maxnum:
89
break
90
revid_index[revid] = rev_index
91
92
parents = graph_parents[revid]
93
for parent_revid in parents:
94
graph_children[parent_revid].append(revid)
95
96
linegraph.append([revid,
97
None,
98
[],
99
parents,
100
None,
101
revno_sequence])
102
103
if graph_data:
104
revno_index[revno_sequence] = rev_index
105
106
branch_id = revno_sequence[0:-1]
107
108
branch_line = None
109
if branch_id not in branch_lines:
110
branch_line = []
111
branch_lines[branch_id] = branch_line
112
else:
113
branch_line = branch_lines[branch_id]
114
115
branch_line.append(rev_index)
116
117
if graph_data:
118
branch_ids = branch_lines.keys()
119
120
def branch_id_cmp(x, y):
121
"""Compaire branch_id's first by the number of digits, then reversed
122
by their value"""
123
len_x = len(x)
124
len_y = len(y)
125
if len_x == len_y:
126
return -cmp(x, y)
127
return cmp(len_x, len_y)
128
129
branch_ids.sort(branch_id_cmp)
130
# This will hold a tuple of (child_index, parent_index, col_index) for each
131
# line that needs to be drawn. If col_index is not none, then the line is
132
# drawn along that column, else the the line can be drawn directly between
133
# the child and parent because either the child and parent are in the same
134
# branch line, or the child and parent are 1 row apart.
135
lines = []
136
empty_column = [False for i in range(len(graph_parents))]
137
# This will hold a bit map for each cell. If the cell is true, then the
138
# cell allready contains a node or line. This use when deciding what column
139
# to place a branch line or line in, without it overlaping something else.
140
columns = [list(empty_column)]
141
142
143
for branch_id in branch_ids:
144
branch_line = branch_lines[branch_id]
145
146
# Find the col_index for the direct parent branch. This will be the
147
# starting point when looking for a free column.
148
parent_col_index = 0
149
parent_index = None
150
if len(branch_id) > 1:
151
parent_revno = branch_id[0:-1]
152
if parent_revno in revno_index:
153
parent_index = revno_index[parent_revno]
154
parent_node = linegraph[parent_index][1]
155
if parent_node:
156
parent_col_index = parent_node[0]
157
158
159
col_search_order = _branch_line_col_search_order(columns,
160
parent_col_index)
161
color = reduce(lambda x, y: x+y, branch_id, 0)
162
cur_cont_line = []
163
164
line_range = []
165
last_rev_index = None
166
for rev_index in branch_line:
167
if last_rev_index:
168
if broken_line_length and \
169
rev_index - last_rev_index > broken_line_length:
170
line_range.append(last_rev_index+1)
171
line_range.append(rev_index-1)
172
else:
173
line_range.extend(range(last_rev_index+1, rev_index))
174
175
line_range.append(rev_index)
176
last_rev_index = rev_index
177
178
if parent_index:
179
if broken_line_length and \
180
parent_index - last_rev_index > broken_line_length:
181
line_range.append(last_rev_index+1)
182
else:
183
line_range.extend(range(last_rev_index+1, parent_index))
184
185
col_index = _find_free_column(columns,
186
empty_column,
187
col_search_order,
188
line_range)
189
node = (col_index, color)
190
for rev_index in branch_line:
191
linegraph[rev_index][1] = node
192
columns[col_index][rev_index] = True
193
194
for rev_index in branch_line:
195
(sequence_number,
196
revid,
197
merge_depth,
198
revno_sequence,
199
end_of_merge) = merge_sorted_revisions[rev_index]
200
201
linegraph[rev_index][4] = graph_children[revid]
202
col_index = linegraph[rev_index][1][0]
203
204
for parent_revid in graph_parents[revid]:
205
if parent_revid in revid_index:
206
207
parent_index = revid_index[parent_revid]
208
parent_node = linegraph[parent_index][1]
209
if parent_node:
210
parent_col_index = parent_node[0]
211
else:
212
parent_col_index = None
213
col_search_order = \
214
_line_col_search_order(columns,
215
parent_col_index,
216
col_index)
217
218
# If this line is really long, break it.
219
if len(branch_id) > 0 and \
220
broken_line_length and \
221
parent_index - rev_index > broken_line_length:
222
child_line_col_index = \
223
_find_free_column(columns,
224
empty_column,
225
col_search_order,
226
(rev_index + 1,))
227
_mark_column_as_used(columns,
228
child_line_col_index,
229
(rev_index + 1,))
230
231
# Recall _line_col_search_order to reset it back to
232
# the beging.
233
col_search_order = \
234
_line_col_search_order(columns,
235
parent_col_index,
236
col_index)
237
parent_col_line_index = \
238
_find_free_column(columns,
239
empty_column,
240
col_search_order,
241
(parent_index - 1,))
242
_mark_column_as_used(columns,
243
parent_col_line_index,
244
(parent_index - 1,))
245
lines.append((rev_index,
246
parent_index,
247
(child_line_col_index,
248
parent_col_line_index)))
249
else :
250
line_col_index = col_index
251
if parent_index - rev_index >1:
252
line_range = range(rev_index + 1, parent_index)
253
line_col_index = \
254
_find_free_column(columns,
255
empty_column,
256
col_search_order,
257
line_range)
258
_mark_column_as_used(columns,
259
line_col_index,
260
line_range)
261
lines.append((rev_index,
262
parent_index,
263
(line_col_index,)))
264
265
for (child_index, parent_index, line_col_indexes) in lines:
266
(child_col_index, child_color) = linegraph[child_index][1]
267
(parent_col_index, parent_color) = linegraph[parent_index][1]
268
269
if len(line_col_indexes) == 1:
270
if parent_index - child_index == 1:
271
linegraph[child_index][2].append(
272
(child_col_index,
273
parent_col_index,
274
parent_color))
275
else:
276
# line from the child's column to the lines column
277
linegraph[child_index][2].append(
278
(child_col_index,
279
line_col_indexes[0],
280
parent_color))
281
# lines down the line's column
282
for line_part_index in range(child_index+1, parent_index-1):
283
linegraph[line_part_index][2].append(
284
(line_col_indexes[0],
285
line_col_indexes[0],
286
parent_color))
287
# line from the line's column to the parent's column
288
linegraph[parent_index-1][2].append(
289
(line_col_indexes[0],
290
parent_col_index,
291
parent_color))
292
else:
293
# Broken line
294
# line from the child's column to the lines column
295
linegraph[child_index][2].append(
296
(child_col_index,
297
line_col_indexes[0],
298
parent_color))
299
# Broken line end
300
linegraph[child_index+1][2].append(
301
(line_col_indexes[0],
302
None,
303
parent_color))
304
305
# Broken line end
306
linegraph[parent_index-2][2].append(
307
(None,
308
line_col_indexes[1],
309
parent_color))
310
# line from the line's column to the parent's column
311
linegraph[parent_index-1][2].append(
312
(line_col_indexes[1],
313
parent_col_index,
314
parent_color))
315
return (linegraph, revid_index, len(columns))
316
else:
317
return (linegraph, revid_index, 0)
318
319
320
def _branch_line_col_search_order(columns, parent_col_index):
321
for col_index in range(parent_col_index, len(columns)):
322
yield col_index
323
for col_index in range(parent_col_index-1, -1, -1):
324
yield col_index
325
326
def _line_col_search_order(columns, parent_col_index, child_col_index):
327
if parent_col_index is not None:
328
max_index = max(parent_col_index, child_col_index)
329
min_index = min(parent_col_index, child_col_index)
330
for col_index in range(max_index, min_index -1, -1):
331
yield col_index
332
else:
333
max_index = child_col_index
334
min_index = child_col_index
335
yield child_col_index
336
i = 1
337
while max_index + i < len(columns) or \
338
min_index - i > -1:
339
if max_index + i < len(columns):
340
yield max_index + i
341
if min_index - i > -1:
342
yield min_index - i
343
i += 1
344
345
def _find_free_column(columns, empty_column, col_search_order, line_range):
346
for col_index in col_search_order:
347
column = columns[col_index]
348
has_overlaping_line = False
349
for row_index in line_range:
350
if column[row_index]:
351
has_overlaping_line = True
352
break
353
if not has_overlaping_line:
354
break
355
else:
356
col_index = len(columns)
357
column = list(empty_column)
358
columns.append(column)
359
return col_index
360
361
def _mark_column_as_used(columns, col_index, line_range):
362
column = columns[col_index]
363
for row_index in line_range:
364
column[row_index] = True
365
366
def same_branch(a, b):
367
"""Return whether we think revisions a and b are on the same branch."""
368
if len(a.parent_ids) == 1:
369
# Defacto same branch if only parent
370
return True
371
elif a.committer == b.committer:
372
# Same committer so may as well be
373
return True
374
else:
375
return False
Loggerhead is a web-based interface for Breezy
Version: 2.0.2.dev0