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
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
15
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
18
17
"""Topological sorting routines."""
21
from bzrlib import errors
22
import bzrlib.revision as _mod_revision
19
from __future__ import absolute_import
25
revision as _mod_revision,
25
29
__all__ = ["topo_sort", "TopoSorter", "merge_sort", "MergeSorter"]
31
35
graph -- sequence of pairs of node->parents_list.
33
The result is a list of node names, such that all parents come before
37
The result is a list of node names, such that all parents come before their
36
40
node identifiers can be any hashable object, and are typically strings.
42
This function has the same purpose as the TopoSorter class, but uses a
43
different algorithm to sort the graph. That means that while both return a
44
list with parents before their child nodes, the exact ordering can be
47
topo_sort is faster when the whole list is needed, while when iterating
48
over a part of the list, TopoSorter.iter_topo_order should be used.
38
return TopoSorter(graph).sorted()
50
kg = _mod_graph.KnownGraph(dict(graph))
41
54
class TopoSorter(object):
43
56
def __init__(self, graph):
44
57
"""Topological sorting of a graph.
46
59
:param graph: sequence of pairs of node_name->parent_names_list.
47
60
i.e. [('C', ['B']), ('B', ['A']), ('A', [])]
48
61
For this input the output from the sort or
49
62
iter_topo_order routines will be:
52
65
node identifiers can be any hashable object, and are typically strings.
54
67
If you have a graph like [('a', ['b']), ('a', ['c'])] this will only use
57
70
The graph is sorted lazily: until you iterate or sort the input is
58
71
not processed other than to create an internal representation.
60
iteration or sorting may raise GraphCycleError if a cycle is present
73
iteration or sorting may raise GraphCycleError if a cycle is present
63
# a dict of the graph.
76
# store a dict of the graph.
64
77
self._graph = dict(graph)
65
self._visitable = set(self._graph)
67
# self._original_graph = dict(graph)
69
# this is a stack storing the depth first search into the graph.
70
self._node_name_stack = []
71
# at each level of 'recursion' we have to check each parent. This
72
# stack stores the parents we have not yet checked for the node at the
73
# matching depth in _node_name_stack
74
self._pending_parents_stack = []
75
# this is a set of the completed nodes for fast checking whether a
76
# parent in a node we are processing on the stack has already been
77
# emitted and thus can be skipped.
78
self._completed_node_names = set()
81
80
"""Sort the graph and return as a list.
83
82
After calling this the sorter is empty and you must create a new one.
85
84
return list(self.iter_topo_order())
97
96
def iter_topo_order(self):
98
97
"""Yield the nodes of the graph in a topological order.
100
99
After finishing iteration the sorter is empty and you cannot continue
103
visitable = set(graph)
105
# this is a stack storing the depth first search into the graph.
106
pending_node_stack = []
107
# at each level of 'recursion' we have to check each parent. This
108
# stack stores the parents we have not yet checked for the node at the
109
# matching depth in pending_node_stack
110
pending_parents_stack = []
112
# this is a set of the completed nodes for fast checking whether a
113
# parent in a node we are processing on the stack has already been
114
# emitted and thus can be skipped.
115
completed_node_names = set()
104
118
# now pick a random node in the source graph, and transfer it to the
105
# top of the depth first search stack.
106
node_name, parents = self._graph.popitem()
107
self._push_node(node_name, parents)
108
while self._node_name_stack:
109
# loop until this call completes.
110
parents_to_visit = self._pending_parents_stack[-1]
111
# if all parents are done, the revision is done
119
# top of the depth first search stack of pending nodes.
120
node_name, parents = graph.popitem()
121
pending_node_stack.append(node_name)
122
pending_parents_stack.append(list(parents))
124
# loop until pending_node_stack is empty
125
while pending_node_stack:
126
parents_to_visit = pending_parents_stack[-1]
127
# if there are no parents left, the revision is done
112
128
if not parents_to_visit:
113
129
# append the revision to the topo sorted list
114
# all the nodes parents have been added to the output, now
115
# we can add it to the output.
116
yield self._pop_node()
130
# all the nodes parents have been added to the output,
131
# now we can add it to the output.
132
popped_node = pending_node_stack.pop()
133
pending_parents_stack.pop()
134
completed_node_names.add(popped_node)
118
while self._pending_parents_stack[-1]:
119
# recurse depth first into a single parent
120
next_node_name = self._pending_parents_stack[-1].pop()
121
if next_node_name in self._completed_node_names:
122
# this parent was completed by a child on the
123
# call stack. skip it.
125
if next_node_name not in self._visitable:
127
# otherwise transfer it from the source graph into the
128
# top of the current depth first search stack.
130
parents = self._graph.pop(next_node_name)
132
# if the next node is not in the source graph it has
133
# already been popped from it and placed into the
134
# current search stack (but not completed or we would
135
# have hit the continue 4 lines up.
136
# this indicates a cycle.
137
raise errors.GraphCycleError(self._node_name_stack)
138
self._push_node(next_node_name, parents)
139
# and do not continue processing parents until this 'call'
143
def _push_node(self, node_name, parents):
144
"""Add node_name to the pending node stack.
146
Names in this stack will get emitted into the output as they are popped
149
self._node_name_stack.append(node_name)
150
self._pending_parents_stack.append(list(parents))
153
"""Pop the top node off the stack
155
The node is appended to the sorted output.
157
# we are returning from the flattened call frame:
158
# pop off the local variables
159
node_name = self._node_name_stack.pop()
160
self._pending_parents_stack.pop()
162
self._completed_node_names.add(node_name)
137
# recurse depth first into a single parent
138
next_node_name = parents_to_visit.pop()
140
if next_node_name in completed_node_names:
141
# parent was already completed by a child, skip it.
143
if next_node_name not in visitable:
144
# parent is not a node in the original graph, skip it.
147
# transfer it along with its parents from the source graph
148
# into the top of the current depth first search stack.
150
parents = graph.pop(next_node_name)
152
# if the next node is not in the source graph it has
153
# already been popped from it and placed into the
154
# current search stack (but not completed or we would
155
# have hit the continue 6 lines up). this indicates a
157
raise errors.GraphCycleError(pending_node_stack)
158
pending_node_stack.append(next_node_name)
159
pending_parents_stack.append(list(parents))
166
162
def merge_sort(graph, branch_tip, mainline_revisions=None, generate_revno=False):
167
163
"""Topological sort a graph which groups merges.
169
165
:param graph: sequence of pairs of node->parents_list.
170
:param branch_tip: the tip of the branch to graph. Revisions not
166
:param branch_tip: the tip of the branch to graph. Revisions not
171
167
reachable from branch_tip are not included in the
173
169
:param mainline_revisions: If not None this forces a mainline to be
286
283
C is the end of a cluster due to rule 1.
287
D is not the end of a cluster from rule 1, but is from rule 2: E
284
D is not the end of a cluster from rule 1, but is from rule 2: E
288
285
is not its left most ancestor
289
286
E is the end of a cluster due to rule 1
290
287
F might be but we need more data.
292
289
we show connecting lines to a parent when:
293
290
- The parent is the start of a merge within this cluster.
294
That is, the merge was not done to the mainline before this cluster
291
That is, the merge was not done to the mainline before this cluster
295
292
was merged to the mainline.
296
293
This can be detected thus:
297
* The parent has a higher merge depth and is the next revision in
294
* The parent has a higher merge depth and is the next revision in
300
297
The next revision in the list constraint is needed for this case:
302
B 1 [C, F] # we do not want to show a line to F which is depth 2
299
B 1 [C, F] # we do not want to show a line to F which is depth 2
304
C 1 [H] # note that this is a long line to show back to the
301
C 1 [H] # note that this is a long line to show back to the
305
302
ancestor - see the end of merge rules.
405
# the scheduling order is: F, E, D, C, B, A
402
# the scheduling order is: F, E, D, C, B, A
406
403
# that is - 'left subtree, right subtree, node'
407
404
# which would mean that when we schedule A we can emit the entire tree.
408
405
self._scheduled_nodes = []
409
# This records for each node when we have processed its left most
406
# This records for each node when we have processed its left most
410
407
# unmerged subtree. After this subtree is scheduled, all other subtrees
411
408
# have their merge depth increased by one from this nodes merge depth.
412
409
# it contains tuples - name, merge_depth
545
555
if not left_subtree_pushed_stack[-1]:
546
556
# recurse depth first into the primary parent
547
557
next_node_name = pending_parents_stack[-1].pop(0)
558
is_left_subtree = True
559
left_subtree_pushed_stack[-1] = True
549
561
# place any merges in right-to-left order for scheduling
550
562
# which gives us left-to-right order after we reverse
551
# the scheduled queue. XXX: This has the effect of
563
# the scheduled queue. XXX: This has the effect of
552
564
# allocating common-new revisions to the right-most
553
# subtree rather than the left most, which will
565
# subtree rather than the left most, which will
554
566
# display nicely (you get smaller trees at the top
555
567
# of the combined merge).
556
568
next_node_name = pending_parents_stack[-1].pop()
569
is_left_subtree = False
557
570
if next_node_name in completed_node_names:
558
571
# this parent was completed by a child on the
559
572
# call stack. skip it.
added
removed
bzrlib/tsort.py
