bzr branch
http://gegoxaren.bato24.eu/bzr/brz/remove-bazaar
|
0.9.2
by Aaron Bentley
Get single-parent comparison working |
1 |
from difflib import SequenceMatcher |
2 |
||
|
0.9.3
by Aaron Bentley
Get three-parent comparisions under test |
3 |
|
|
0.9.1
by Aaron Bentley
Get trivial case passing |
4 |
class MultiParent(object): |
5 |
||
|
0.9.2
by Aaron Bentley
Get single-parent comparison working |
6 |
def __init__(self, hunks=None): |
7 |
if hunks is not None: |
|
8 |
self.hunks = hunks |
|
9 |
else: |
|
10 |
self.hunks = [] |
|
11 |
||
12 |
def __repr__(self): |
|
13 |
return "MultiParent(%r)" % self.hunks |
|
14 |
||
15 |
def __eq__(self, other): |
|
16 |
if self.__class__ is not other.__class__: |
|
17 |
return False |
|
18 |
return (self.hunks == other.hunks) |
|
|
0.9.1
by Aaron Bentley
Get trivial case passing |
19 |
|
20 |
@staticmethod
|
|
21 |
def from_lines(text, parents=()): |
|
|
0.9.2
by Aaron Bentley
Get single-parent comparison working |
22 |
def compare(parent): |
23 |
return SequenceMatcher(None, parent, text).get_matching_blocks() |
|
24 |
parent_comparisons = [compare(p) for p in parents] |
|
25 |
cur_line = 0 |
|
26 |
new_text = NewText([]) |
|
27 |
parent_text = [] |
|
28 |
block_iter = [iter(i) for i in parent_comparisons] |
|
29 |
diff = MultiParent([]) |
|
30 |
def next_block(p): |
|
31 |
try: |
|
32 |
return block_iter[p].next() |
|
33 |
except StopIteration: |
|
34 |
return None |
|
35 |
cur_block = [next_block(p) for p, i in enumerate(block_iter)] |
|
36 |
while cur_line < len(text): |
|
37 |
best_match = None |
|
38 |
for p, block in enumerate(cur_block): |
|
39 |
if block is None: |
|
40 |
continue
|
|
41 |
i, j, n = block |
|
42 |
while j + n < cur_line: |
|
43 |
block = cur_block[p] = next_block(p) |
|
44 |
if block is None: |
|
45 |
break
|
|
46 |
i, j, n = block |
|
47 |
if block is None: |
|
48 |
continue
|
|
49 |
if j > cur_line: |
|
50 |
continue
|
|
51 |
offset = cur_line - j |
|
52 |
i += offset |
|
53 |
j = cur_line |
|
54 |
n -= offset |
|
55 |
if n == 0: |
|
56 |
continue
|
|
57 |
if best_match is None or n > best_match.num_lines: |
|
58 |
best_match = ParentText(p, i, j, n) |
|
59 |
if best_match is None: |
|
60 |
new_text.lines.append(text[cur_line]) |
|
61 |
cur_line += 1 |
|
62 |
else: |
|
63 |
if len(new_text.lines) > 0: |
|
64 |
diff.hunks.append(new_text) |
|
65 |
new_text = NewText([]) |
|
66 |
diff.hunks.append(best_match) |
|
67 |
cur_line += best_match.num_lines |
|
68 |
if len(new_text.lines) > 0: |
|
69 |
diff.hunks.append(new_text) |
|
|
0.9.1
by Aaron Bentley
Get trivial case passing |
70 |
return diff |
71 |
||
72 |
@classmethod
|
|
73 |
def from_texts(cls, text, parents=()): |
|
74 |
return cls.from_lines(text.splitlines(True), |
|
75 |
[p.splitlines(True) for p in parents]) |
|
76 |
||
|
0.9.4
by Aaron Bentley
Start supporting serialization |
77 |
def to_patch(self): |
78 |
for hunk in self.hunks: |
|
79 |
for line in hunk.to_patch(): |
|
80 |
yield line |
|
81 |
||
|
0.9.9
by Aaron Bentley
Much progress on non-naive text reconstruction |
82 |
def range_iterator(self): |
83 |
start = 0 |
|
84 |
for hunk in self.hunks: |
|
85 |
if isinstance(hunk, NewText): |
|
86 |
kind = 'new' |
|
87 |
end = start + len(hunk.lines) |
|
88 |
data = hunk.lines |
|
89 |
else: |
|
90 |
kind = 'parent' |
|
91 |
start = hunk.child_pos |
|
92 |
end = start + hunk.num_lines |
|
93 |
data = (hunk.parent, hunk.parent_pos, hunk.parent_pos + |
|
94 |
hunk.num_lines) |
|
95 |
yield start, end, kind, data |
|
96 |
start = end |
|
97 |
||
|
0.9.1
by Aaron Bentley
Get trivial case passing |
98 |
|
99 |
class NewText(object): |
|
100 |
||
101 |
def __init__(self, lines): |
|
102 |
self.lines = lines |
|
103 |
||
104 |
def __eq__(self, other): |
|
105 |
if self.__class__ is not other.__class__: |
|
106 |
return False |
|
107 |
return (other.lines == self.lines) |
|
|
0.9.2
by Aaron Bentley
Get single-parent comparison working |
108 |
|
109 |
def __repr__(self): |
|
110 |
return 'NewText(%r)' % self.lines |
|
111 |
||
|
0.9.4
by Aaron Bentley
Start supporting serialization |
112 |
def to_patch(self): |
113 |
yield 'i %d\n' % len(self.lines) |
|
114 |
for line in self.lines: |
|
115 |
yield line |
|
116 |
yield '\n' |
|
117 |
||
|
0.9.2
by Aaron Bentley
Get single-parent comparison working |
118 |
|
119 |
class ParentText(object): |
|
120 |
||
121 |
def __init__(self, parent, parent_pos, child_pos, num_lines): |
|
122 |
self.parent = parent |
|
123 |
self.parent_pos = parent_pos |
|
124 |
self.child_pos = child_pos |
|
125 |
self.num_lines = num_lines |
|
126 |
||
127 |
def __repr__(self): |
|
128 |
return 'ParentText(%(parent)r, %(parent_pos)r, %(child_pos)r,'\ |
|
129 |
' %(num_lines)r)' % self.__dict__ |
|
130 |
||
131 |
def __eq__(self, other): |
|
132 |
if self.__class__ != other.__class__: |
|
133 |
return False |
|
134 |
return (self.__dict__ == other.__dict__) |
|
|
0.9.4
by Aaron Bentley
Start supporting serialization |
135 |
|
136 |
def to_patch(self): |
|
137 |
yield 'c %(parent)d %(parent_pos)d %(child_pos)d %(num_lines)d\n'\ |
|
138 |
% self.__dict__ |
|
|
0.9.8
by Aaron Bentley
get add_version working |
139 |
|
140 |
||
141 |
class MultiVersionedFile(object): |
|
142 |
||
143 |
def __init__(self): |
|
144 |
self._diffs = {} |
|
145 |
self._lines = {} |
|
146 |
self._parents = {} |
|
147 |
||
148 |
def add_version(self, lines, version_id, parent_ids): |
|
149 |
parent_lines = [self._lines[p] for p in parent_ids] |
|
150 |
diff = MultiParent.from_lines(lines, parent_lines) |
|
151 |
self.add_diff(diff, version_id, parent_ids) |
|
152 |
self._lines[version_id] = lines |
|
153 |
||
154 |
def add_diff(self, diff, version_id, parent_ids): |
|
155 |
self._diffs[version_id] = diff |
|
156 |
self._parents[version_id] = parent_ids |
|
157 |
||
158 |
def clear_cache(self): |
|
159 |
self._lines.clear() |
|
|
0.9.9
by Aaron Bentley
Much progress on non-naive text reconstruction |
160 |
|
161 |
def get_line_list(self, version_ids): |
|
162 |
return [self.cache_version(v) for v in version_ids] |
|
163 |
||
164 |
def cache_version(self, version_id): |
|
165 |
try: |
|
166 |
return self._lines[version_id] |
|
167 |
except KeyError: |
|
168 |
pass
|
|
169 |
diff = self._diffs[version_id] |
|
170 |
lines = [] |
|
171 |
reconstructor = _Reconstructor(self._diffs, self._lines, self._parents) |
|
172 |
for hunk in diff.hunks: |
|
173 |
if isinstance(hunk, NewText): |
|
174 |
lines.extend(hunk.lines) |
|
175 |
else: |
|
176 |
reconstructor.reconstruct(lines, hunk, version_id) |
|
177 |
self._lines[version_id] = lines |
|
178 |
return lines |
|
179 |
||
180 |
||
181 |
class _Reconstructor(object): |
|
182 |
||
183 |
def __init__(self, diffs, lines, parents): |
|
184 |
||
185 |
self.diffs = diffs |
|
186 |
self.lines = lines |
|
187 |
self.parents = parents |
|
188 |
self.cursor = {} |
|
189 |
||
190 |
def reconstruct(self, lines, parent_text, version_id): |
|
191 |
parent_id = self.parents[version_id][parent_text.parent] |
|
192 |
end = parent_text.parent_pos + parent_text.num_lines |
|
193 |
return self._reconstruct(lines, parent_id, parent_text.parent_pos, end) |
|
194 |
||
195 |
def _reconstruct(self, lines, req_version_id, req_start, req_end): |
|
196 |
pending_reqs = [(req_version_id, req_start, req_end)] |
|
197 |
while len(pending_reqs) > 0: |
|
198 |
req_version_id, req_start, req_end = pending_reqs.pop() |
|
199 |
try: |
|
200 |
start, end, kind, data, iterator = self.cursor[req_version_id] |
|
201 |
except KeyError: |
|
202 |
iterator = self.diffs[req_version_id].range_iterator() |
|
203 |
start, end, kind, data = iterator.next() |
|
204 |
while end < req_start: |
|
205 |
start, end, kind, data = iterator.next() |
|
206 |
self.cursor[req_version_id] = start, end, kind, data, iterator |
|
207 |
if kind == 'new': |
|
208 |
lines.extend(data[req_start - start: (req_end - start)]) |
|
209 |
else: |
|
210 |
parent, parent_start, parent_end = data |
|
211 |
version_id = self.parents[req_version_id][parent] |
|
212 |
sub_start = parent_start + req_start - start |
|
213 |
sub_end = parent_end + req_end - end |
|
214 |
pending_reqs.append((version_id, sub_start, sub_end)) |