Show More
@@ -1,492 +1,492 b'' | |||||
1 | # mdiff.py - diff and patch routines for mercurial |
|
1 | # mdiff.py - diff and patch routines for mercurial | |
2 | # |
|
2 | # | |
3 | # Copyright 2005, 2006 Matt Mackall <mpm@selenic.com> |
|
3 | # Copyright 2005, 2006 Matt Mackall <mpm@selenic.com> | |
4 | # |
|
4 | # | |
5 | # This software may be used and distributed according to the terms of the |
|
5 | # This software may be used and distributed according to the terms of the | |
6 | # GNU General Public License version 2 or any later version. |
|
6 | # GNU General Public License version 2 or any later version. | |
7 |
|
7 | |||
8 | from __future__ import absolute_import |
|
8 | from __future__ import absolute_import | |
9 |
|
9 | |||
10 | import re |
|
10 | import re | |
11 | import struct |
|
11 | import struct | |
12 | import zlib |
|
12 | import zlib | |
13 |
|
13 | |||
14 | from .i18n import _ |
|
14 | from .i18n import _ | |
15 | from . import ( |
|
15 | from . import ( | |
16 | error, |
|
16 | error, | |
17 | policy, |
|
17 | policy, | |
18 | pycompat, |
|
18 | pycompat, | |
19 | util, |
|
19 | util, | |
20 | ) |
|
20 | ) | |
21 |
|
21 | |||
22 | bdiff = policy.importmod(r'bdiff') |
|
22 | bdiff = policy.importmod(r'bdiff') | |
23 | mpatch = policy.importmod(r'mpatch') |
|
23 | mpatch = policy.importmod(r'mpatch') | |
24 |
|
24 | |||
25 | blocks = bdiff.blocks |
|
25 | blocks = bdiff.blocks | |
26 | fixws = bdiff.fixws |
|
26 | fixws = bdiff.fixws | |
27 | patches = mpatch.patches |
|
27 | patches = mpatch.patches | |
28 | patchedsize = mpatch.patchedsize |
|
28 | patchedsize = mpatch.patchedsize | |
29 | textdiff = bdiff.bdiff |
|
29 | textdiff = bdiff.bdiff | |
30 |
|
30 | |||
31 | def splitnewlines(text): |
|
31 | def splitnewlines(text): | |
32 | '''like str.splitlines, but only split on newlines.''' |
|
32 | '''like str.splitlines, but only split on newlines.''' | |
33 | lines = [l + '\n' for l in text.split('\n')] |
|
33 | lines = [l + '\n' for l in text.split('\n')] | |
34 | if lines: |
|
34 | if lines: | |
35 | if lines[-1] == '\n': |
|
35 | if lines[-1] == '\n': | |
36 | lines.pop() |
|
36 | lines.pop() | |
37 | else: |
|
37 | else: | |
38 | lines[-1] = lines[-1][:-1] |
|
38 | lines[-1] = lines[-1][:-1] | |
39 | return lines |
|
39 | return lines | |
40 |
|
40 | |||
41 | class diffopts(object): |
|
41 | class diffopts(object): | |
42 | '''context is the number of context lines |
|
42 | '''context is the number of context lines | |
43 | text treats all files as text |
|
43 | text treats all files as text | |
44 | showfunc enables diff -p output |
|
44 | showfunc enables diff -p output | |
45 | git enables the git extended patch format |
|
45 | git enables the git extended patch format | |
46 | nodates removes dates from diff headers |
|
46 | nodates removes dates from diff headers | |
47 | nobinary ignores binary files |
|
47 | nobinary ignores binary files | |
48 | noprefix disables the 'a/' and 'b/' prefixes (ignored in plain mode) |
|
48 | noprefix disables the 'a/' and 'b/' prefixes (ignored in plain mode) | |
49 | ignorews ignores all whitespace changes in the diff |
|
49 | ignorews ignores all whitespace changes in the diff | |
50 | ignorewsamount ignores changes in the amount of whitespace |
|
50 | ignorewsamount ignores changes in the amount of whitespace | |
51 | ignoreblanklines ignores changes whose lines are all blank |
|
51 | ignoreblanklines ignores changes whose lines are all blank | |
52 | upgrade generates git diffs to avoid data loss |
|
52 | upgrade generates git diffs to avoid data loss | |
53 | ''' |
|
53 | ''' | |
54 |
|
54 | |||
55 | defaults = { |
|
55 | defaults = { | |
56 | 'context': 3, |
|
56 | 'context': 3, | |
57 | 'text': False, |
|
57 | 'text': False, | |
58 | 'showfunc': False, |
|
58 | 'showfunc': False, | |
59 | 'git': False, |
|
59 | 'git': False, | |
60 | 'nodates': False, |
|
60 | 'nodates': False, | |
61 | 'nobinary': False, |
|
61 | 'nobinary': False, | |
62 | 'noprefix': False, |
|
62 | 'noprefix': False, | |
63 | 'index': 0, |
|
63 | 'index': 0, | |
64 | 'ignorews': False, |
|
64 | 'ignorews': False, | |
65 | 'ignorewsamount': False, |
|
65 | 'ignorewsamount': False, | |
66 | 'ignorewseol': False, |
|
66 | 'ignorewseol': False, | |
67 | 'ignoreblanklines': False, |
|
67 | 'ignoreblanklines': False, | |
68 | 'upgrade': False, |
|
68 | 'upgrade': False, | |
69 | 'showsimilarity': False, |
|
69 | 'showsimilarity': False, | |
70 | 'worddiff': False, |
|
70 | 'worddiff': False, | |
71 | } |
|
71 | } | |
72 |
|
72 | |||
73 | def __init__(self, **opts): |
|
73 | def __init__(self, **opts): | |
74 | opts = pycompat.byteskwargs(opts) |
|
74 | opts = pycompat.byteskwargs(opts) | |
75 | for k in self.defaults.keys(): |
|
75 | for k in self.defaults.keys(): | |
76 | v = opts.get(k) |
|
76 | v = opts.get(k) | |
77 | if v is None: |
|
77 | if v is None: | |
78 | v = self.defaults[k] |
|
78 | v = self.defaults[k] | |
79 | setattr(self, k, v) |
|
79 | setattr(self, k, v) | |
80 |
|
80 | |||
81 | try: |
|
81 | try: | |
82 | self.context = int(self.context) |
|
82 | self.context = int(self.context) | |
83 | except ValueError: |
|
83 | except ValueError: | |
84 | raise error.Abort(_('diff context lines count must be ' |
|
84 | raise error.Abort(_('diff context lines count must be ' | |
85 | 'an integer, not %r') % self.context) |
|
85 | 'an integer, not %r') % self.context) | |
86 |
|
86 | |||
87 | def copy(self, **kwargs): |
|
87 | def copy(self, **kwargs): | |
88 | opts = dict((k, getattr(self, k)) for k in self.defaults) |
|
88 | opts = dict((k, getattr(self, k)) for k in self.defaults) | |
89 | opts = pycompat.strkwargs(opts) |
|
89 | opts = pycompat.strkwargs(opts) | |
90 | opts.update(kwargs) |
|
90 | opts.update(kwargs) | |
91 | return diffopts(**opts) |
|
91 | return diffopts(**opts) | |
92 |
|
92 | |||
93 | defaultopts = diffopts() |
|
93 | defaultopts = diffopts() | |
94 |
|
94 | |||
95 | def wsclean(opts, text, blank=True): |
|
95 | def wsclean(opts, text, blank=True): | |
96 | if opts.ignorews: |
|
96 | if opts.ignorews: | |
97 | text = bdiff.fixws(text, 1) |
|
97 | text = bdiff.fixws(text, 1) | |
98 | elif opts.ignorewsamount: |
|
98 | elif opts.ignorewsamount: | |
99 | text = bdiff.fixws(text, 0) |
|
99 | text = bdiff.fixws(text, 0) | |
100 | if blank and opts.ignoreblanklines: |
|
100 | if blank and opts.ignoreblanklines: | |
101 | text = re.sub('\n+', '\n', text).strip('\n') |
|
101 | text = re.sub('\n+', '\n', text).strip('\n') | |
102 | if opts.ignorewseol: |
|
102 | if opts.ignorewseol: | |
103 | text = re.sub(r'[ \t\r\f]+\n', r'\n', text) |
|
103 | text = re.sub(r'[ \t\r\f]+\n', r'\n', text) | |
104 | return text |
|
104 | return text | |
105 |
|
105 | |||
106 | def splitblock(base1, lines1, base2, lines2, opts): |
|
106 | def splitblock(base1, lines1, base2, lines2, opts): | |
107 | # The input lines matches except for interwoven blank lines. We |
|
107 | # The input lines matches except for interwoven blank lines. We | |
108 | # transform it into a sequence of matching blocks and blank blocks. |
|
108 | # transform it into a sequence of matching blocks and blank blocks. | |
109 | lines1 = [(wsclean(opts, l) and 1 or 0) for l in lines1] |
|
109 | lines1 = [(wsclean(opts, l) and 1 or 0) for l in lines1] | |
110 | lines2 = [(wsclean(opts, l) and 1 or 0) for l in lines2] |
|
110 | lines2 = [(wsclean(opts, l) and 1 or 0) for l in lines2] | |
111 | s1, e1 = 0, len(lines1) |
|
111 | s1, e1 = 0, len(lines1) | |
112 | s2, e2 = 0, len(lines2) |
|
112 | s2, e2 = 0, len(lines2) | |
113 | while s1 < e1 or s2 < e2: |
|
113 | while s1 < e1 or s2 < e2: | |
114 | i1, i2, btype = s1, s2, '=' |
|
114 | i1, i2, btype = s1, s2, '=' | |
115 | if (i1 >= e1 or lines1[i1] == 0 |
|
115 | if (i1 >= e1 or lines1[i1] == 0 | |
116 | or i2 >= e2 or lines2[i2] == 0): |
|
116 | or i2 >= e2 or lines2[i2] == 0): | |
117 | # Consume the block of blank lines |
|
117 | # Consume the block of blank lines | |
118 | btype = '~' |
|
118 | btype = '~' | |
119 | while i1 < e1 and lines1[i1] == 0: |
|
119 | while i1 < e1 and lines1[i1] == 0: | |
120 | i1 += 1 |
|
120 | i1 += 1 | |
121 | while i2 < e2 and lines2[i2] == 0: |
|
121 | while i2 < e2 and lines2[i2] == 0: | |
122 | i2 += 1 |
|
122 | i2 += 1 | |
123 | else: |
|
123 | else: | |
124 | # Consume the matching lines |
|
124 | # Consume the matching lines | |
125 | while i1 < e1 and lines1[i1] == 1 and lines2[i2] == 1: |
|
125 | while i1 < e1 and lines1[i1] == 1 and lines2[i2] == 1: | |
126 | i1 += 1 |
|
126 | i1 += 1 | |
127 | i2 += 1 |
|
127 | i2 += 1 | |
128 | yield [base1 + s1, base1 + i1, base2 + s2, base2 + i2], btype |
|
128 | yield [base1 + s1, base1 + i1, base2 + s2, base2 + i2], btype | |
129 | s1 = i1 |
|
129 | s1 = i1 | |
130 | s2 = i2 |
|
130 | s2 = i2 | |
131 |
|
131 | |||
132 | def hunkinrange(hunk, linerange): |
|
132 | def hunkinrange(hunk, linerange): | |
133 | """Return True if `hunk` defined as (start, length) is in `linerange` |
|
133 | """Return True if `hunk` defined as (start, length) is in `linerange` | |
134 | defined as (lowerbound, upperbound). |
|
134 | defined as (lowerbound, upperbound). | |
135 |
|
135 | |||
136 | >>> hunkinrange((5, 10), (2, 7)) |
|
136 | >>> hunkinrange((5, 10), (2, 7)) | |
137 | True |
|
137 | True | |
138 | >>> hunkinrange((5, 10), (6, 12)) |
|
138 | >>> hunkinrange((5, 10), (6, 12)) | |
139 | True |
|
139 | True | |
140 | >>> hunkinrange((5, 10), (13, 17)) |
|
140 | >>> hunkinrange((5, 10), (13, 17)) | |
141 | True |
|
141 | True | |
142 | >>> hunkinrange((5, 10), (3, 17)) |
|
142 | >>> hunkinrange((5, 10), (3, 17)) | |
143 | True |
|
143 | True | |
144 | >>> hunkinrange((5, 10), (1, 3)) |
|
144 | >>> hunkinrange((5, 10), (1, 3)) | |
145 | False |
|
145 | False | |
146 | >>> hunkinrange((5, 10), (18, 20)) |
|
146 | >>> hunkinrange((5, 10), (18, 20)) | |
147 | False |
|
147 | False | |
148 | >>> hunkinrange((5, 10), (1, 5)) |
|
148 | >>> hunkinrange((5, 10), (1, 5)) | |
149 | False |
|
149 | False | |
150 | >>> hunkinrange((5, 10), (15, 27)) |
|
150 | >>> hunkinrange((5, 10), (15, 27)) | |
151 | False |
|
151 | False | |
152 | """ |
|
152 | """ | |
153 | start, length = hunk |
|
153 | start, length = hunk | |
154 | lowerbound, upperbound = linerange |
|
154 | lowerbound, upperbound = linerange | |
155 | return lowerbound < start + length and start < upperbound |
|
155 | return lowerbound < start + length and start < upperbound | |
156 |
|
156 | |||
157 | def blocksinrange(blocks, rangeb): |
|
157 | def blocksinrange(blocks, rangeb): | |
158 | """filter `blocks` like (a1, a2, b1, b2) from items outside line range |
|
158 | """filter `blocks` like (a1, a2, b1, b2) from items outside line range | |
159 | `rangeb` from ``(b1, b2)`` point of view. |
|
159 | `rangeb` from ``(b1, b2)`` point of view. | |
160 |
|
160 | |||
161 | Return `filteredblocks, rangea` where: |
|
161 | Return `filteredblocks, rangea` where: | |
162 |
|
162 | |||
163 | * `filteredblocks` is list of ``block = (a1, a2, b1, b2), stype`` items of |
|
163 | * `filteredblocks` is list of ``block = (a1, a2, b1, b2), stype`` items of | |
164 | `blocks` that are inside `rangeb` from ``(b1, b2)`` point of view; a |
|
164 | `blocks` that are inside `rangeb` from ``(b1, b2)`` point of view; a | |
165 | block ``(b1, b2)`` being inside `rangeb` if |
|
165 | block ``(b1, b2)`` being inside `rangeb` if | |
166 | ``rangeb[0] < b2 and b1 < rangeb[1]``; |
|
166 | ``rangeb[0] < b2 and b1 < rangeb[1]``; | |
167 | * `rangea` is the line range w.r.t. to ``(a1, a2)`` parts of `blocks`. |
|
167 | * `rangea` is the line range w.r.t. to ``(a1, a2)`` parts of `blocks`. | |
168 | """ |
|
168 | """ | |
169 | lbb, ubb = rangeb |
|
169 | lbb, ubb = rangeb | |
170 | lba, uba = None, None |
|
170 | lba, uba = None, None | |
171 | filteredblocks = [] |
|
171 | filteredblocks = [] | |
172 | for block in blocks: |
|
172 | for block in blocks: | |
173 | (a1, a2, b1, b2), stype = block |
|
173 | (a1, a2, b1, b2), stype = block | |
174 | if lbb >= b1 and ubb <= b2 and stype == '=': |
|
174 | if lbb >= b1 and ubb <= b2 and stype == '=': | |
175 | # rangeb is within a single "=" hunk, restrict back linerange1 |
|
175 | # rangeb is within a single "=" hunk, restrict back linerange1 | |
176 | # by offsetting rangeb |
|
176 | # by offsetting rangeb | |
177 | lba = lbb - b1 + a1 |
|
177 | lba = lbb - b1 + a1 | |
178 | uba = ubb - b1 + a1 |
|
178 | uba = ubb - b1 + a1 | |
179 | else: |
|
179 | else: | |
180 | if b1 <= lbb < b2: |
|
180 | if b1 <= lbb < b2: | |
181 | if stype == '=': |
|
181 | if stype == '=': | |
182 | lba = a2 - (b2 - lbb) |
|
182 | lba = a2 - (b2 - lbb) | |
183 | else: |
|
183 | else: | |
184 | lba = a1 |
|
184 | lba = a1 | |
185 | if b1 < ubb <= b2: |
|
185 | if b1 < ubb <= b2: | |
186 | if stype == '=': |
|
186 | if stype == '=': | |
187 | uba = a1 + (ubb - b1) |
|
187 | uba = a1 + (ubb - b1) | |
188 | else: |
|
188 | else: | |
189 | uba = a2 |
|
189 | uba = a2 | |
190 | if hunkinrange((b1, (b2 - b1)), rangeb): |
|
190 | if hunkinrange((b1, (b2 - b1)), rangeb): | |
191 | filteredblocks.append(block) |
|
191 | filteredblocks.append(block) | |
192 | if lba is None or uba is None or uba < lba: |
|
192 | if lba is None or uba is None or uba < lba: | |
193 | raise error.Abort(_('line range exceeds file size')) |
|
193 | raise error.Abort(_('line range exceeds file size')) | |
194 | return filteredblocks, (lba, uba) |
|
194 | return filteredblocks, (lba, uba) | |
195 |
|
195 | |||
196 | def allblocks(text1, text2, opts=None, lines1=None, lines2=None): |
|
196 | def allblocks(text1, text2, opts=None, lines1=None, lines2=None): | |
197 | """Return (block, type) tuples, where block is an mdiff.blocks |
|
197 | """Return (block, type) tuples, where block is an mdiff.blocks | |
198 | line entry. type is '=' for blocks matching exactly one another |
|
198 | line entry. type is '=' for blocks matching exactly one another | |
199 | (bdiff blocks), '!' for non-matching blocks and '~' for blocks |
|
199 | (bdiff blocks), '!' for non-matching blocks and '~' for blocks | |
200 | matching only after having filtered blank lines. |
|
200 | matching only after having filtered blank lines. | |
201 | line1 and line2 are text1 and text2 split with splitnewlines() if |
|
201 | line1 and line2 are text1 and text2 split with splitnewlines() if | |
202 | they are already available. |
|
202 | they are already available. | |
203 | """ |
|
203 | """ | |
204 | if opts is None: |
|
204 | if opts is None: | |
205 | opts = defaultopts |
|
205 | opts = defaultopts | |
206 | if opts.ignorews or opts.ignorewsamount or opts.ignorewseol: |
|
206 | if opts.ignorews or opts.ignorewsamount or opts.ignorewseol: | |
207 | text1 = wsclean(opts, text1, False) |
|
207 | text1 = wsclean(opts, text1, False) | |
208 | text2 = wsclean(opts, text2, False) |
|
208 | text2 = wsclean(opts, text2, False) | |
209 | diff = bdiff.blocks(text1, text2) |
|
209 | diff = bdiff.blocks(text1, text2) | |
210 | for i, s1 in enumerate(diff): |
|
210 | for i, s1 in enumerate(diff): | |
211 | # The first match is special. |
|
211 | # The first match is special. | |
212 | # we've either found a match starting at line 0 or a match later |
|
212 | # we've either found a match starting at line 0 or a match later | |
213 | # in the file. If it starts later, old and new below will both be |
|
213 | # in the file. If it starts later, old and new below will both be | |
214 | # empty and we'll continue to the next match. |
|
214 | # empty and we'll continue to the next match. | |
215 | if i > 0: |
|
215 | if i > 0: | |
216 | s = diff[i - 1] |
|
216 | s = diff[i - 1] | |
217 | else: |
|
217 | else: | |
218 | s = [0, 0, 0, 0] |
|
218 | s = [0, 0, 0, 0] | |
219 | s = [s[1], s1[0], s[3], s1[2]] |
|
219 | s = [s[1], s1[0], s[3], s1[2]] | |
220 |
|
220 | |||
221 | # bdiff sometimes gives huge matches past eof, this check eats them, |
|
221 | # bdiff sometimes gives huge matches past eof, this check eats them, | |
222 | # and deals with the special first match case described above |
|
222 | # and deals with the special first match case described above | |
223 | if s[0] != s[1] or s[2] != s[3]: |
|
223 | if s[0] != s[1] or s[2] != s[3]: | |
224 | type = '!' |
|
224 | type = '!' | |
225 | if opts.ignoreblanklines: |
|
225 | if opts.ignoreblanklines: | |
226 | if lines1 is None: |
|
226 | if lines1 is None: | |
227 | lines1 = splitnewlines(text1) |
|
227 | lines1 = splitnewlines(text1) | |
228 | if lines2 is None: |
|
228 | if lines2 is None: | |
229 | lines2 = splitnewlines(text2) |
|
229 | lines2 = splitnewlines(text2) | |
230 | old = wsclean(opts, "".join(lines1[s[0]:s[1]])) |
|
230 | old = wsclean(opts, "".join(lines1[s[0]:s[1]])) | |
231 | new = wsclean(opts, "".join(lines2[s[2]:s[3]])) |
|
231 | new = wsclean(opts, "".join(lines2[s[2]:s[3]])) | |
232 | if old == new: |
|
232 | if old == new: | |
233 | type = '~' |
|
233 | type = '~' | |
234 | yield s, type |
|
234 | yield s, type | |
235 | yield s1, '=' |
|
235 | yield s1, '=' | |
236 |
|
236 | |||
237 | def unidiff(a, ad, b, bd, fn1, fn2, opts=defaultopts): |
|
237 | def unidiff(a, ad, b, bd, fn1, fn2, opts=defaultopts): | |
238 | """Return a unified diff as a (headers, hunks) tuple. |
|
238 | """Return a unified diff as a (headers, hunks) tuple. | |
239 |
|
239 | |||
240 | If the diff is not null, `headers` is a list with unified diff header |
|
240 | If the diff is not null, `headers` is a list with unified diff header | |
241 | lines "--- <original>" and "+++ <new>" and `hunks` is a generator yielding |
|
241 | lines "--- <original>" and "+++ <new>" and `hunks` is a generator yielding | |
242 | (hunkrange, hunklines) coming from _unidiff(). |
|
242 | (hunkrange, hunklines) coming from _unidiff(). | |
243 | Otherwise, `headers` and `hunks` are empty. |
|
243 | Otherwise, `headers` and `hunks` are empty. | |
244 | """ |
|
244 | """ | |
245 | def datetag(date, fn=None): |
|
245 | def datetag(date, fn=None): | |
246 | if not opts.git and not opts.nodates: |
|
246 | if not opts.git and not opts.nodates: | |
247 | return '\t%s' % date |
|
247 | return '\t%s' % date | |
248 | if fn and ' ' in fn: |
|
248 | if fn and ' ' in fn: | |
249 | return '\t' |
|
249 | return '\t' | |
250 | return '' |
|
250 | return '' | |
251 |
|
251 | |||
252 | sentinel = [], () |
|
252 | sentinel = [], () | |
253 | if not a and not b: |
|
253 | if not a and not b: | |
254 | return sentinel |
|
254 | return sentinel | |
255 |
|
255 | |||
256 | if opts.noprefix: |
|
256 | if opts.noprefix: | |
257 | aprefix = bprefix = '' |
|
257 | aprefix = bprefix = '' | |
258 | else: |
|
258 | else: | |
259 | aprefix = 'a/' |
|
259 | aprefix = 'a/' | |
260 | bprefix = 'b/' |
|
260 | bprefix = 'b/' | |
261 |
|
261 | |||
262 | epoch = util.datestr((0, 0)) |
|
262 | epoch = util.datestr((0, 0)) | |
263 |
|
263 | |||
264 | fn1 = util.pconvert(fn1) |
|
264 | fn1 = util.pconvert(fn1) | |
265 | fn2 = util.pconvert(fn2) |
|
265 | fn2 = util.pconvert(fn2) | |
266 |
|
266 | |||
267 | def checknonewline(lines): |
|
267 | def checknonewline(lines): | |
268 | for text in lines: |
|
268 | for text in lines: | |
269 | if text[-1:] != '\n': |
|
269 | if text[-1:] != '\n': | |
270 | text += "\n\ No newline at end of file\n" |
|
270 | text += "\n\ No newline at end of file\n" | |
271 | yield text |
|
271 | yield text | |
272 |
|
272 | |||
273 | if not opts.text and (util.binary(a) or util.binary(b)): |
|
273 | if not opts.text and (util.binary(a) or util.binary(b)): | |
274 | if a and b and len(a) == len(b) and a == b: |
|
274 | if a and b and len(a) == len(b) and a == b: | |
275 | return sentinel |
|
275 | return sentinel | |
276 | headerlines = [] |
|
276 | headerlines = [] | |
277 | hunks = (None, ['Binary file %s has changed\n' % fn1]), |
|
277 | hunks = (None, ['Binary file %s has changed\n' % fn1]), | |
278 | elif not a: |
|
278 | elif not a: | |
279 | b = splitnewlines(b) |
|
279 | b = splitnewlines(b) | |
280 | if a is None: |
|
280 | if a is None: | |
281 | l1 = '--- /dev/null%s' % datetag(epoch) |
|
281 | l1 = '--- /dev/null%s' % datetag(epoch) | |
282 | else: |
|
282 | else: | |
283 | l1 = "--- %s%s%s" % (aprefix, fn1, datetag(ad, fn1)) |
|
283 | l1 = "--- %s%s%s" % (aprefix, fn1, datetag(ad, fn1)) | |
284 | l2 = "+++ %s%s" % (bprefix + fn2, datetag(bd, fn2)) |
|
284 | l2 = "+++ %s%s" % (bprefix + fn2, datetag(bd, fn2)) | |
285 | headerlines = [l1, l2] |
|
285 | headerlines = [l1, l2] | |
286 | size = len(b) |
|
286 | size = len(b) | |
287 | hunkrange = (0, 0, 1, size) |
|
287 | hunkrange = (0, 0, 1, size) | |
288 | hunklines = ["@@ -0,0 +1,%d @@\n" % size] + ["+" + e for e in b] |
|
288 | hunklines = ["@@ -0,0 +1,%d @@\n" % size] + ["+" + e for e in b] | |
289 | hunks = (hunkrange, checknonewline(hunklines)), |
|
289 | hunks = (hunkrange, checknonewline(hunklines)), | |
290 | elif not b: |
|
290 | elif not b: | |
291 | a = splitnewlines(a) |
|
291 | a = splitnewlines(a) | |
292 | l1 = "--- %s%s%s" % (aprefix, fn1, datetag(ad, fn1)) |
|
292 | l1 = "--- %s%s%s" % (aprefix, fn1, datetag(ad, fn1)) | |
293 | if b is None: |
|
293 | if b is None: | |
294 | l2 = '+++ /dev/null%s' % datetag(epoch) |
|
294 | l2 = '+++ /dev/null%s' % datetag(epoch) | |
295 | else: |
|
295 | else: | |
296 | l2 = "+++ %s%s%s" % (bprefix, fn2, datetag(bd, fn2)) |
|
296 | l2 = "+++ %s%s%s" % (bprefix, fn2, datetag(bd, fn2)) | |
297 | headerlines = [l1, l2] |
|
297 | headerlines = [l1, l2] | |
298 | size = len(a) |
|
298 | size = len(a) | |
299 | hunkrange = (1, size, 0, 0) |
|
299 | hunkrange = (1, size, 0, 0) | |
300 | hunklines = ["@@ -1,%d +0,0 @@\n" % size] + ["-" + e for e in a] |
|
300 | hunklines = ["@@ -1,%d +0,0 @@\n" % size] + ["-" + e for e in a] | |
301 | hunks = (hunkrange, checknonewline(hunklines)), |
|
301 | hunks = (hunkrange, checknonewline(hunklines)), | |
302 | else: |
|
302 | else: | |
303 | diffhunks = _unidiff(a, b, opts=opts) |
|
303 | diffhunks = _unidiff(a, b, opts=opts) | |
304 | try: |
|
304 | try: | |
305 | hunkrange, hunklines = next(diffhunks) |
|
305 | hunkrange, hunklines = next(diffhunks) | |
306 | except StopIteration: |
|
306 | except StopIteration: | |
307 | return sentinel |
|
307 | return sentinel | |
308 |
|
308 | |||
309 | headerlines = [ |
|
309 | headerlines = [ | |
310 | "--- %s%s%s" % (aprefix, fn1, datetag(ad, fn1)), |
|
310 | "--- %s%s%s" % (aprefix, fn1, datetag(ad, fn1)), | |
311 | "+++ %s%s%s" % (bprefix, fn2, datetag(bd, fn2)), |
|
311 | "+++ %s%s%s" % (bprefix, fn2, datetag(bd, fn2)), | |
312 | ] |
|
312 | ] | |
313 | def rewindhunks(): |
|
313 | def rewindhunks(): | |
314 | yield hunkrange, checknonewline(hunklines) |
|
314 | yield hunkrange, checknonewline(hunklines) | |
315 | for hr, hl in diffhunks: |
|
315 | for hr, hl in diffhunks: | |
316 | yield hr, checknonewline(hl) |
|
316 | yield hr, checknonewline(hl) | |
317 |
|
317 | |||
318 | hunks = rewindhunks() |
|
318 | hunks = rewindhunks() | |
319 |
|
319 | |||
320 | return headerlines, hunks |
|
320 | return headerlines, hunks | |
321 |
|
321 | |||
322 | def _unidiff(t1, t2, opts=defaultopts): |
|
322 | def _unidiff(t1, t2, opts=defaultopts): | |
323 | """Yield hunks of a headerless unified diff from t1 and t2 texts. |
|
323 | """Yield hunks of a headerless unified diff from t1 and t2 texts. | |
324 |
|
324 | |||
325 | Each hunk consists of a (hunkrange, hunklines) tuple where `hunkrange` is a |
|
325 | Each hunk consists of a (hunkrange, hunklines) tuple where `hunkrange` is a | |
326 | tuple (s1, l1, s2, l2) representing the range information of the hunk to |
|
326 | tuple (s1, l1, s2, l2) representing the range information of the hunk to | |
327 | form the '@@ -s1,l1 +s2,l2 @@' header and `hunklines` is a list of lines |
|
327 | form the '@@ -s1,l1 +s2,l2 @@' header and `hunklines` is a list of lines | |
328 | of the hunk combining said header followed by line additions and |
|
328 | of the hunk combining said header followed by line additions and | |
329 | deletions. |
|
329 | deletions. | |
330 | """ |
|
330 | """ | |
331 | l1 = splitnewlines(t1) |
|
331 | l1 = splitnewlines(t1) | |
332 | l2 = splitnewlines(t2) |
|
332 | l2 = splitnewlines(t2) | |
333 | def contextend(l, len): |
|
333 | def contextend(l, len): | |
334 | ret = l + opts.context |
|
334 | ret = l + opts.context | |
335 | if ret > len: |
|
335 | if ret > len: | |
336 | ret = len |
|
336 | ret = len | |
337 | return ret |
|
337 | return ret | |
338 |
|
338 | |||
339 | def contextstart(l): |
|
339 | def contextstart(l): | |
340 | ret = l - opts.context |
|
340 | ret = l - opts.context | |
341 | if ret < 0: |
|
341 | if ret < 0: | |
342 | return 0 |
|
342 | return 0 | |
343 | return ret |
|
343 | return ret | |
344 |
|
344 | |||
345 | lastfunc = [0, ''] |
|
345 | lastfunc = [0, ''] | |
346 | def yieldhunk(hunk): |
|
346 | def yieldhunk(hunk): | |
347 | (astart, a2, bstart, b2, delta) = hunk |
|
347 | (astart, a2, bstart, b2, delta) = hunk | |
348 | aend = contextend(a2, len(l1)) |
|
348 | aend = contextend(a2, len(l1)) | |
349 | alen = aend - astart |
|
349 | alen = aend - astart | |
350 | blen = b2 - bstart + aend - a2 |
|
350 | blen = b2 - bstart + aend - a2 | |
351 |
|
351 | |||
352 | func = "" |
|
352 | func = "" | |
353 | if opts.showfunc: |
|
353 | if opts.showfunc: | |
354 | lastpos, func = lastfunc |
|
354 | lastpos, func = lastfunc | |
355 | # walk backwards from the start of the context up to the start of |
|
355 | # walk backwards from the start of the context up to the start of | |
356 | # the previous hunk context until we find a line starting with an |
|
356 | # the previous hunk context until we find a line starting with an | |
357 | # alphanumeric char. |
|
357 | # alphanumeric char. | |
358 | for i in xrange(astart - 1, lastpos - 1, -1): |
|
358 | for i in xrange(astart - 1, lastpos - 1, -1): | |
359 | if l1[i][0].isalnum(): |
|
359 | if l1[i][0:1].isalnum(): | |
360 | func = ' ' + l1[i].rstrip()[:40] |
|
360 | func = ' ' + l1[i].rstrip()[:40] | |
361 | lastfunc[1] = func |
|
361 | lastfunc[1] = func | |
362 | break |
|
362 | break | |
363 | # by recording this hunk's starting point as the next place to |
|
363 | # by recording this hunk's starting point as the next place to | |
364 | # start looking for function lines, we avoid reading any line in |
|
364 | # start looking for function lines, we avoid reading any line in | |
365 | # the file more than once. |
|
365 | # the file more than once. | |
366 | lastfunc[0] = astart |
|
366 | lastfunc[0] = astart | |
367 |
|
367 | |||
368 | # zero-length hunk ranges report their start line as one less |
|
368 | # zero-length hunk ranges report their start line as one less | |
369 | if alen: |
|
369 | if alen: | |
370 | astart += 1 |
|
370 | astart += 1 | |
371 | if blen: |
|
371 | if blen: | |
372 | bstart += 1 |
|
372 | bstart += 1 | |
373 |
|
373 | |||
374 | hunkrange = astart, alen, bstart, blen |
|
374 | hunkrange = astart, alen, bstart, blen | |
375 | hunklines = ( |
|
375 | hunklines = ( | |
376 | ["@@ -%d,%d +%d,%d @@%s\n" % (hunkrange + (func,))] |
|
376 | ["@@ -%d,%d +%d,%d @@%s\n" % (hunkrange + (func,))] | |
377 | + delta |
|
377 | + delta | |
378 | + [' ' + l1[x] for x in xrange(a2, aend)] |
|
378 | + [' ' + l1[x] for x in xrange(a2, aend)] | |
379 | ) |
|
379 | ) | |
380 | yield hunkrange, hunklines |
|
380 | yield hunkrange, hunklines | |
381 |
|
381 | |||
382 | # bdiff.blocks gives us the matching sequences in the files. The loop |
|
382 | # bdiff.blocks gives us the matching sequences in the files. The loop | |
383 | # below finds the spaces between those matching sequences and translates |
|
383 | # below finds the spaces between those matching sequences and translates | |
384 | # them into diff output. |
|
384 | # them into diff output. | |
385 | # |
|
385 | # | |
386 | hunk = None |
|
386 | hunk = None | |
387 | ignoredlines = 0 |
|
387 | ignoredlines = 0 | |
388 | for s, stype in allblocks(t1, t2, opts, l1, l2): |
|
388 | for s, stype in allblocks(t1, t2, opts, l1, l2): | |
389 | a1, a2, b1, b2 = s |
|
389 | a1, a2, b1, b2 = s | |
390 | if stype != '!': |
|
390 | if stype != '!': | |
391 | if stype == '~': |
|
391 | if stype == '~': | |
392 | # The diff context lines are based on t1 content. When |
|
392 | # The diff context lines are based on t1 content. When | |
393 | # blank lines are ignored, the new lines offsets must |
|
393 | # blank lines are ignored, the new lines offsets must | |
394 | # be adjusted as if equivalent blocks ('~') had the |
|
394 | # be adjusted as if equivalent blocks ('~') had the | |
395 | # same sizes on both sides. |
|
395 | # same sizes on both sides. | |
396 | ignoredlines += (b2 - b1) - (a2 - a1) |
|
396 | ignoredlines += (b2 - b1) - (a2 - a1) | |
397 | continue |
|
397 | continue | |
398 | delta = [] |
|
398 | delta = [] | |
399 | old = l1[a1:a2] |
|
399 | old = l1[a1:a2] | |
400 | new = l2[b1:b2] |
|
400 | new = l2[b1:b2] | |
401 |
|
401 | |||
402 | b1 -= ignoredlines |
|
402 | b1 -= ignoredlines | |
403 | b2 -= ignoredlines |
|
403 | b2 -= ignoredlines | |
404 | astart = contextstart(a1) |
|
404 | astart = contextstart(a1) | |
405 | bstart = contextstart(b1) |
|
405 | bstart = contextstart(b1) | |
406 | prev = None |
|
406 | prev = None | |
407 | if hunk: |
|
407 | if hunk: | |
408 | # join with the previous hunk if it falls inside the context |
|
408 | # join with the previous hunk if it falls inside the context | |
409 | if astart < hunk[1] + opts.context + 1: |
|
409 | if astart < hunk[1] + opts.context + 1: | |
410 | prev = hunk |
|
410 | prev = hunk | |
411 | astart = hunk[1] |
|
411 | astart = hunk[1] | |
412 | bstart = hunk[3] |
|
412 | bstart = hunk[3] | |
413 | else: |
|
413 | else: | |
414 | for x in yieldhunk(hunk): |
|
414 | for x in yieldhunk(hunk): | |
415 | yield x |
|
415 | yield x | |
416 | if prev: |
|
416 | if prev: | |
417 | # we've joined the previous hunk, record the new ending points. |
|
417 | # we've joined the previous hunk, record the new ending points. | |
418 | hunk[1] = a2 |
|
418 | hunk[1] = a2 | |
419 | hunk[3] = b2 |
|
419 | hunk[3] = b2 | |
420 | delta = hunk[4] |
|
420 | delta = hunk[4] | |
421 | else: |
|
421 | else: | |
422 | # create a new hunk |
|
422 | # create a new hunk | |
423 | hunk = [astart, a2, bstart, b2, delta] |
|
423 | hunk = [astart, a2, bstart, b2, delta] | |
424 |
|
424 | |||
425 | delta[len(delta):] = [' ' + x for x in l1[astart:a1]] |
|
425 | delta[len(delta):] = [' ' + x for x in l1[astart:a1]] | |
426 | delta[len(delta):] = ['-' + x for x in old] |
|
426 | delta[len(delta):] = ['-' + x for x in old] | |
427 | delta[len(delta):] = ['+' + x for x in new] |
|
427 | delta[len(delta):] = ['+' + x for x in new] | |
428 |
|
428 | |||
429 | if hunk: |
|
429 | if hunk: | |
430 | for x in yieldhunk(hunk): |
|
430 | for x in yieldhunk(hunk): | |
431 | yield x |
|
431 | yield x | |
432 |
|
432 | |||
433 | def b85diff(to, tn): |
|
433 | def b85diff(to, tn): | |
434 | '''print base85-encoded binary diff''' |
|
434 | '''print base85-encoded binary diff''' | |
435 | def fmtline(line): |
|
435 | def fmtline(line): | |
436 | l = len(line) |
|
436 | l = len(line) | |
437 | if l <= 26: |
|
437 | if l <= 26: | |
438 | l = chr(ord('A') + l - 1) |
|
438 | l = chr(ord('A') + l - 1) | |
439 | else: |
|
439 | else: | |
440 | l = chr(l - 26 + ord('a') - 1) |
|
440 | l = chr(l - 26 + ord('a') - 1) | |
441 | return '%c%s\n' % (l, util.b85encode(line, True)) |
|
441 | return '%c%s\n' % (l, util.b85encode(line, True)) | |
442 |
|
442 | |||
443 | def chunk(text, csize=52): |
|
443 | def chunk(text, csize=52): | |
444 | l = len(text) |
|
444 | l = len(text) | |
445 | i = 0 |
|
445 | i = 0 | |
446 | while i < l: |
|
446 | while i < l: | |
447 | yield text[i:i + csize] |
|
447 | yield text[i:i + csize] | |
448 | i += csize |
|
448 | i += csize | |
449 |
|
449 | |||
450 | if to is None: |
|
450 | if to is None: | |
451 | to = '' |
|
451 | to = '' | |
452 | if tn is None: |
|
452 | if tn is None: | |
453 | tn = '' |
|
453 | tn = '' | |
454 |
|
454 | |||
455 | if to == tn: |
|
455 | if to == tn: | |
456 | return '' |
|
456 | return '' | |
457 |
|
457 | |||
458 | # TODO: deltas |
|
458 | # TODO: deltas | |
459 | ret = [] |
|
459 | ret = [] | |
460 | ret.append('GIT binary patch\n') |
|
460 | ret.append('GIT binary patch\n') | |
461 | ret.append('literal %d\n' % len(tn)) |
|
461 | ret.append('literal %d\n' % len(tn)) | |
462 | for l in chunk(zlib.compress(tn)): |
|
462 | for l in chunk(zlib.compress(tn)): | |
463 | ret.append(fmtline(l)) |
|
463 | ret.append(fmtline(l)) | |
464 | ret.append('\n') |
|
464 | ret.append('\n') | |
465 |
|
465 | |||
466 | return ''.join(ret) |
|
466 | return ''.join(ret) | |
467 |
|
467 | |||
468 | def patchtext(bin): |
|
468 | def patchtext(bin): | |
469 | pos = 0 |
|
469 | pos = 0 | |
470 | t = [] |
|
470 | t = [] | |
471 | while pos < len(bin): |
|
471 | while pos < len(bin): | |
472 | p1, p2, l = struct.unpack(">lll", bin[pos:pos + 12]) |
|
472 | p1, p2, l = struct.unpack(">lll", bin[pos:pos + 12]) | |
473 | pos += 12 |
|
473 | pos += 12 | |
474 | t.append(bin[pos:pos + l]) |
|
474 | t.append(bin[pos:pos + l]) | |
475 | pos += l |
|
475 | pos += l | |
476 | return "".join(t) |
|
476 | return "".join(t) | |
477 |
|
477 | |||
478 | def patch(a, bin): |
|
478 | def patch(a, bin): | |
479 | if len(a) == 0: |
|
479 | if len(a) == 0: | |
480 | # skip over trivial delta header |
|
480 | # skip over trivial delta header | |
481 | return util.buffer(bin, 12) |
|
481 | return util.buffer(bin, 12) | |
482 | return mpatch.patches(a, [bin]) |
|
482 | return mpatch.patches(a, [bin]) | |
483 |
|
483 | |||
484 | # similar to difflib.SequenceMatcher.get_matching_blocks |
|
484 | # similar to difflib.SequenceMatcher.get_matching_blocks | |
485 | def get_matching_blocks(a, b): |
|
485 | def get_matching_blocks(a, b): | |
486 | return [(d[0], d[2], d[1] - d[0]) for d in bdiff.blocks(a, b)] |
|
486 | return [(d[0], d[2], d[1] - d[0]) for d in bdiff.blocks(a, b)] | |
487 |
|
487 | |||
488 | def trivialdiffheader(length): |
|
488 | def trivialdiffheader(length): | |
489 | return struct.pack(">lll", 0, 0, length) if length else '' |
|
489 | return struct.pack(">lll", 0, 0, length) if length else '' | |
490 |
|
490 | |||
491 | def replacediffheader(oldlen, newlen): |
|
491 | def replacediffheader(oldlen, newlen): | |
492 | return struct.pack(">lll", 0, oldlen, newlen) |
|
492 | return struct.pack(">lll", 0, oldlen, newlen) |
General Comments 0
You need to be logged in to leave comments.
Login now