Show More
@@ -1,252 +1,257 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 |
|
5 | # This software may be used and distributed according to the terms | |
6 | # of the GNU General Public License, incorporated herein by reference. |
|
6 | # of the GNU General Public License, incorporated herein by reference. | |
7 |
|
7 | |||
8 | from demandload import demandload |
|
8 | from demandload import demandload | |
9 | import bdiff, mpatch |
|
9 | import bdiff, mpatch | |
10 | demandload(globals(), "re struct util") |
|
10 | demandload(globals(), "re struct util md5") | |
11 |
|
11 | |||
12 | def splitnewlines(text): |
|
12 | def splitnewlines(text): | |
13 | '''like str.splitlines, but only split on newlines.''' |
|
13 | '''like str.splitlines, but only split on newlines.''' | |
14 | lines = [l + '\n' for l in text.split('\n')] |
|
14 | lines = [l + '\n' for l in text.split('\n')] | |
15 | if lines: |
|
15 | if lines: | |
16 | if lines[-1] == '\n': |
|
16 | if lines[-1] == '\n': | |
17 | lines.pop() |
|
17 | lines.pop() | |
18 | else: |
|
18 | else: | |
19 | lines[-1] = lines[-1][:-1] |
|
19 | lines[-1] = lines[-1][:-1] | |
20 | return lines |
|
20 | return lines | |
21 |
|
21 | |||
22 | class diffopts(object): |
|
22 | class diffopts(object): | |
23 | '''context is the number of context lines |
|
23 | '''context is the number of context lines | |
24 | text treats all files as text |
|
24 | text treats all files as text | |
25 | showfunc enables diff -p output |
|
25 | showfunc enables diff -p output | |
26 | git enables the git extended patch format |
|
26 | git enables the git extended patch format | |
27 | nodates removes dates from diff headers |
|
27 | nodates removes dates from diff headers | |
28 | ignorews ignores all whitespace changes in the diff |
|
28 | ignorews ignores all whitespace changes in the diff | |
29 | ignorewsamount ignores changes in the amount of whitespace |
|
29 | ignorewsamount ignores changes in the amount of whitespace | |
30 | ignoreblanklines ignores changes whose lines are all blank''' |
|
30 | ignoreblanklines ignores changes whose lines are all blank''' | |
31 |
|
31 | |||
32 | defaults = { |
|
32 | defaults = { | |
33 | 'context': 3, |
|
33 | 'context': 3, | |
34 | 'text': False, |
|
34 | 'text': False, | |
35 | 'showfunc': True, |
|
35 | 'showfunc': True, | |
36 | 'git': False, |
|
36 | 'git': False, | |
37 | 'nodates': False, |
|
37 | 'nodates': False, | |
38 | 'ignorews': False, |
|
38 | 'ignorews': False, | |
39 | 'ignorewsamount': False, |
|
39 | 'ignorewsamount': False, | |
40 | 'ignoreblanklines': False, |
|
40 | 'ignoreblanklines': False, | |
41 | } |
|
41 | } | |
42 |
|
42 | |||
43 | __slots__ = defaults.keys() |
|
43 | __slots__ = defaults.keys() | |
44 |
|
44 | |||
45 | def __init__(self, **opts): |
|
45 | def __init__(self, **opts): | |
46 | for k in self.__slots__: |
|
46 | for k in self.__slots__: | |
47 | v = opts.get(k) |
|
47 | v = opts.get(k) | |
48 | if v is None: |
|
48 | if v is None: | |
49 | v = self.defaults[k] |
|
49 | v = self.defaults[k] | |
50 | setattr(self, k, v) |
|
50 | setattr(self, k, v) | |
51 |
|
51 | |||
52 | defaultopts = diffopts() |
|
52 | defaultopts = diffopts() | |
53 |
|
53 | |||
54 | def unidiff(a, ad, b, bd, fn, r=None, opts=defaultopts): |
|
54 | def unidiff(a, ad, b, bd, fn, r=None, opts=defaultopts): | |
55 | def datetag(date): |
|
55 | def datetag(date): | |
56 | return (opts.git or opts.nodates) and '\n' or '\t%s\n' % date |
|
56 | return (opts.git or opts.nodates) and '\n' or '\t%s\n' % date | |
57 |
|
57 | |||
58 | if not a and not b: return "" |
|
58 | if not a and not b: return "" | |
59 | epoch = util.datestr((0, 0)) |
|
59 | epoch = util.datestr((0, 0)) | |
60 |
|
60 | |||
61 | if not opts.text and (util.binary(a) or util.binary(b)): |
|
61 | if not opts.text and (util.binary(a) or util.binary(b)): | |
|
62 | def h(v): | |||
|
63 | # md5 is used instead of sha1 because md5 is supposedly faster | |||
|
64 | return md5.new(v).digest() | |||
|
65 | if a and b and len(a) == len(b) and h(a) == h(b): | |||
|
66 | return "" | |||
62 | l = ['Binary file %s has changed\n' % fn] |
|
67 | l = ['Binary file %s has changed\n' % fn] | |
63 | elif not a: |
|
68 | elif not a: | |
64 | b = splitnewlines(b) |
|
69 | b = splitnewlines(b) | |
65 | if a is None: |
|
70 | if a is None: | |
66 | l1 = '--- /dev/null%s' % datetag(epoch) |
|
71 | l1 = '--- /dev/null%s' % datetag(epoch) | |
67 | else: |
|
72 | else: | |
68 | l1 = "--- %s%s" % ("a/" + fn, datetag(ad)) |
|
73 | l1 = "--- %s%s" % ("a/" + fn, datetag(ad)) | |
69 | l2 = "+++ %s%s" % ("b/" + fn, datetag(bd)) |
|
74 | l2 = "+++ %s%s" % ("b/" + fn, datetag(bd)) | |
70 | l3 = "@@ -0,0 +1,%d @@\n" % len(b) |
|
75 | l3 = "@@ -0,0 +1,%d @@\n" % len(b) | |
71 | l = [l1, l2, l3] + ["+" + e for e in b] |
|
76 | l = [l1, l2, l3] + ["+" + e for e in b] | |
72 | elif not b: |
|
77 | elif not b: | |
73 | a = splitnewlines(a) |
|
78 | a = splitnewlines(a) | |
74 | l1 = "--- %s%s" % ("a/" + fn, datetag(ad)) |
|
79 | l1 = "--- %s%s" % ("a/" + fn, datetag(ad)) | |
75 | if b is None: |
|
80 | if b is None: | |
76 | l2 = '+++ /dev/null%s' % datetag(epoch) |
|
81 | l2 = '+++ /dev/null%s' % datetag(epoch) | |
77 | else: |
|
82 | else: | |
78 | l2 = "+++ %s%s" % ("b/" + fn, datetag(bd)) |
|
83 | l2 = "+++ %s%s" % ("b/" + fn, datetag(bd)) | |
79 | l3 = "@@ -1,%d +0,0 @@\n" % len(a) |
|
84 | l3 = "@@ -1,%d +0,0 @@\n" % len(a) | |
80 | l = [l1, l2, l3] + ["-" + e for e in a] |
|
85 | l = [l1, l2, l3] + ["-" + e for e in a] | |
81 | else: |
|
86 | else: | |
82 | al = splitnewlines(a) |
|
87 | al = splitnewlines(a) | |
83 | bl = splitnewlines(b) |
|
88 | bl = splitnewlines(b) | |
84 | l = list(bunidiff(a, b, al, bl, "a/" + fn, "b/" + fn, opts=opts)) |
|
89 | l = list(bunidiff(a, b, al, bl, "a/" + fn, "b/" + fn, opts=opts)) | |
85 | if not l: return "" |
|
90 | if not l: return "" | |
86 | # difflib uses a space, rather than a tab |
|
91 | # difflib uses a space, rather than a tab | |
87 | l[0] = "%s%s" % (l[0][:-2], datetag(ad)) |
|
92 | l[0] = "%s%s" % (l[0][:-2], datetag(ad)) | |
88 | l[1] = "%s%s" % (l[1][:-2], datetag(bd)) |
|
93 | l[1] = "%s%s" % (l[1][:-2], datetag(bd)) | |
89 |
|
94 | |||
90 | for ln in xrange(len(l)): |
|
95 | for ln in xrange(len(l)): | |
91 | if l[ln][-1] != '\n': |
|
96 | if l[ln][-1] != '\n': | |
92 | l[ln] += "\n\ No newline at end of file\n" |
|
97 | l[ln] += "\n\ No newline at end of file\n" | |
93 |
|
98 | |||
94 | if r: |
|
99 | if r: | |
95 | l.insert(0, "diff %s %s\n" % |
|
100 | l.insert(0, "diff %s %s\n" % | |
96 | (' '.join(["-r %s" % rev for rev in r]), fn)) |
|
101 | (' '.join(["-r %s" % rev for rev in r]), fn)) | |
97 |
|
102 | |||
98 | return "".join(l) |
|
103 | return "".join(l) | |
99 |
|
104 | |||
100 | # somewhat self contained replacement for difflib.unified_diff |
|
105 | # somewhat self contained replacement for difflib.unified_diff | |
101 | # t1 and t2 are the text to be diffed |
|
106 | # t1 and t2 are the text to be diffed | |
102 | # l1 and l2 are the text broken up into lines |
|
107 | # l1 and l2 are the text broken up into lines | |
103 | # header1 and header2 are the filenames for the diff output |
|
108 | # header1 and header2 are the filenames for the diff output | |
104 | def bunidiff(t1, t2, l1, l2, header1, header2, opts=defaultopts): |
|
109 | def bunidiff(t1, t2, l1, l2, header1, header2, opts=defaultopts): | |
105 | def contextend(l, len): |
|
110 | def contextend(l, len): | |
106 | ret = l + opts.context |
|
111 | ret = l + opts.context | |
107 | if ret > len: |
|
112 | if ret > len: | |
108 | ret = len |
|
113 | ret = len | |
109 | return ret |
|
114 | return ret | |
110 |
|
115 | |||
111 | def contextstart(l): |
|
116 | def contextstart(l): | |
112 | ret = l - opts.context |
|
117 | ret = l - opts.context | |
113 | if ret < 0: |
|
118 | if ret < 0: | |
114 | return 0 |
|
119 | return 0 | |
115 | return ret |
|
120 | return ret | |
116 |
|
121 | |||
117 | def yieldhunk(hunk, header): |
|
122 | def yieldhunk(hunk, header): | |
118 | if header: |
|
123 | if header: | |
119 | for x in header: |
|
124 | for x in header: | |
120 | yield x |
|
125 | yield x | |
121 | (astart, a2, bstart, b2, delta) = hunk |
|
126 | (astart, a2, bstart, b2, delta) = hunk | |
122 | aend = contextend(a2, len(l1)) |
|
127 | aend = contextend(a2, len(l1)) | |
123 | alen = aend - astart |
|
128 | alen = aend - astart | |
124 | blen = b2 - bstart + aend - a2 |
|
129 | blen = b2 - bstart + aend - a2 | |
125 |
|
130 | |||
126 | func = "" |
|
131 | func = "" | |
127 | if opts.showfunc: |
|
132 | if opts.showfunc: | |
128 | # walk backwards from the start of the context |
|
133 | # walk backwards from the start of the context | |
129 | # to find a line starting with an alphanumeric char. |
|
134 | # to find a line starting with an alphanumeric char. | |
130 | for x in xrange(astart, -1, -1): |
|
135 | for x in xrange(astart, -1, -1): | |
131 | t = l1[x].rstrip() |
|
136 | t = l1[x].rstrip() | |
132 | if funcre.match(t): |
|
137 | if funcre.match(t): | |
133 | func = ' ' + t[:40] |
|
138 | func = ' ' + t[:40] | |
134 | break |
|
139 | break | |
135 |
|
140 | |||
136 | yield "@@ -%d,%d +%d,%d @@%s\n" % (astart + 1, alen, |
|
141 | yield "@@ -%d,%d +%d,%d @@%s\n" % (astart + 1, alen, | |
137 | bstart + 1, blen, func) |
|
142 | bstart + 1, blen, func) | |
138 | for x in delta: |
|
143 | for x in delta: | |
139 | yield x |
|
144 | yield x | |
140 | for x in xrange(a2, aend): |
|
145 | for x in xrange(a2, aend): | |
141 | yield ' ' + l1[x] |
|
146 | yield ' ' + l1[x] | |
142 |
|
147 | |||
143 | header = [ "--- %s\t\n" % header1, "+++ %s\t\n" % header2 ] |
|
148 | header = [ "--- %s\t\n" % header1, "+++ %s\t\n" % header2 ] | |
144 |
|
149 | |||
145 | if opts.showfunc: |
|
150 | if opts.showfunc: | |
146 | funcre = re.compile('\w') |
|
151 | funcre = re.compile('\w') | |
147 | if opts.ignorewsamount: |
|
152 | if opts.ignorewsamount: | |
148 | wsamountre = re.compile('[ \t]+') |
|
153 | wsamountre = re.compile('[ \t]+') | |
149 | wsappendedre = re.compile(' \n') |
|
154 | wsappendedre = re.compile(' \n') | |
150 | if opts.ignoreblanklines: |
|
155 | if opts.ignoreblanklines: | |
151 | wsblanklinesre = re.compile('\n') |
|
156 | wsblanklinesre = re.compile('\n') | |
152 | if opts.ignorews: |
|
157 | if opts.ignorews: | |
153 | wsre = re.compile('[ \t]') |
|
158 | wsre = re.compile('[ \t]') | |
154 |
|
159 | |||
155 | # bdiff.blocks gives us the matching sequences in the files. The loop |
|
160 | # bdiff.blocks gives us the matching sequences in the files. The loop | |
156 | # below finds the spaces between those matching sequences and translates |
|
161 | # below finds the spaces between those matching sequences and translates | |
157 | # them into diff output. |
|
162 | # them into diff output. | |
158 | # |
|
163 | # | |
159 | diff = bdiff.blocks(t1, t2) |
|
164 | diff = bdiff.blocks(t1, t2) | |
160 | hunk = None |
|
165 | hunk = None | |
161 | for i in xrange(len(diff)): |
|
166 | for i in xrange(len(diff)): | |
162 | # The first match is special. |
|
167 | # The first match is special. | |
163 | # we've either found a match starting at line 0 or a match later |
|
168 | # we've either found a match starting at line 0 or a match later | |
164 | # in the file. If it starts later, old and new below will both be |
|
169 | # in the file. If it starts later, old and new below will both be | |
165 | # empty and we'll continue to the next match. |
|
170 | # empty and we'll continue to the next match. | |
166 | if i > 0: |
|
171 | if i > 0: | |
167 | s = diff[i-1] |
|
172 | s = diff[i-1] | |
168 | else: |
|
173 | else: | |
169 | s = [0, 0, 0, 0] |
|
174 | s = [0, 0, 0, 0] | |
170 | delta = [] |
|
175 | delta = [] | |
171 | s1 = diff[i] |
|
176 | s1 = diff[i] | |
172 | a1 = s[1] |
|
177 | a1 = s[1] | |
173 | a2 = s1[0] |
|
178 | a2 = s1[0] | |
174 | b1 = s[3] |
|
179 | b1 = s[3] | |
175 | b2 = s1[2] |
|
180 | b2 = s1[2] | |
176 |
|
181 | |||
177 | old = l1[a1:a2] |
|
182 | old = l1[a1:a2] | |
178 | new = l2[b1:b2] |
|
183 | new = l2[b1:b2] | |
179 |
|
184 | |||
180 | # bdiff sometimes gives huge matches past eof, this check eats them, |
|
185 | # bdiff sometimes gives huge matches past eof, this check eats them, | |
181 | # and deals with the special first match case described above |
|
186 | # and deals with the special first match case described above | |
182 | if not old and not new: |
|
187 | if not old and not new: | |
183 | continue |
|
188 | continue | |
184 |
|
189 | |||
185 | if opts.ignoreblanklines: |
|
190 | if opts.ignoreblanklines: | |
186 | wsold = wsblanklinesre.sub('', "".join(old)) |
|
191 | wsold = wsblanklinesre.sub('', "".join(old)) | |
187 | wsnew = wsblanklinesre.sub('', "".join(new)) |
|
192 | wsnew = wsblanklinesre.sub('', "".join(new)) | |
188 | if wsold == wsnew: |
|
193 | if wsold == wsnew: | |
189 | continue |
|
194 | continue | |
190 |
|
195 | |||
191 | if opts.ignorewsamount: |
|
196 | if opts.ignorewsamount: | |
192 | wsold = wsamountre.sub(' ', "".join(old)) |
|
197 | wsold = wsamountre.sub(' ', "".join(old)) | |
193 | wsold = wsappendedre.sub('\n', wsold) |
|
198 | wsold = wsappendedre.sub('\n', wsold) | |
194 | wsnew = wsamountre.sub(' ', "".join(new)) |
|
199 | wsnew = wsamountre.sub(' ', "".join(new)) | |
195 | wsnew = wsappendedre.sub('\n', wsnew) |
|
200 | wsnew = wsappendedre.sub('\n', wsnew) | |
196 | if wsold == wsnew: |
|
201 | if wsold == wsnew: | |
197 | continue |
|
202 | continue | |
198 |
|
203 | |||
199 | if opts.ignorews: |
|
204 | if opts.ignorews: | |
200 | wsold = wsre.sub('', "".join(old)) |
|
205 | wsold = wsre.sub('', "".join(old)) | |
201 | wsnew = wsre.sub('', "".join(new)) |
|
206 | wsnew = wsre.sub('', "".join(new)) | |
202 | if wsold == wsnew: |
|
207 | if wsold == wsnew: | |
203 | continue |
|
208 | continue | |
204 |
|
209 | |||
205 | astart = contextstart(a1) |
|
210 | astart = contextstart(a1) | |
206 | bstart = contextstart(b1) |
|
211 | bstart = contextstart(b1) | |
207 | prev = None |
|
212 | prev = None | |
208 | if hunk: |
|
213 | if hunk: | |
209 | # join with the previous hunk if it falls inside the context |
|
214 | # join with the previous hunk if it falls inside the context | |
210 | if astart < hunk[1] + opts.context + 1: |
|
215 | if astart < hunk[1] + opts.context + 1: | |
211 | prev = hunk |
|
216 | prev = hunk | |
212 | astart = hunk[1] |
|
217 | astart = hunk[1] | |
213 | bstart = hunk[3] |
|
218 | bstart = hunk[3] | |
214 | else: |
|
219 | else: | |
215 | for x in yieldhunk(hunk, header): |
|
220 | for x in yieldhunk(hunk, header): | |
216 | yield x |
|
221 | yield x | |
217 | # we only want to yield the header if the files differ, and |
|
222 | # we only want to yield the header if the files differ, and | |
218 | # we only want to yield it once. |
|
223 | # we only want to yield it once. | |
219 | header = None |
|
224 | header = None | |
220 | if prev: |
|
225 | if prev: | |
221 | # we've joined the previous hunk, record the new ending points. |
|
226 | # we've joined the previous hunk, record the new ending points. | |
222 | hunk[1] = a2 |
|
227 | hunk[1] = a2 | |
223 | hunk[3] = b2 |
|
228 | hunk[3] = b2 | |
224 | delta = hunk[4] |
|
229 | delta = hunk[4] | |
225 | else: |
|
230 | else: | |
226 | # create a new hunk |
|
231 | # create a new hunk | |
227 | hunk = [ astart, a2, bstart, b2, delta ] |
|
232 | hunk = [ astart, a2, bstart, b2, delta ] | |
228 |
|
233 | |||
229 | delta[len(delta):] = [ ' ' + x for x in l1[astart:a1] ] |
|
234 | delta[len(delta):] = [ ' ' + x for x in l1[astart:a1] ] | |
230 | delta[len(delta):] = [ '-' + x for x in old ] |
|
235 | delta[len(delta):] = [ '-' + x for x in old ] | |
231 | delta[len(delta):] = [ '+' + x for x in new ] |
|
236 | delta[len(delta):] = [ '+' + x for x in new ] | |
232 |
|
237 | |||
233 | if hunk: |
|
238 | if hunk: | |
234 | for x in yieldhunk(hunk, header): |
|
239 | for x in yieldhunk(hunk, header): | |
235 | yield x |
|
240 | yield x | |
236 |
|
241 | |||
237 | def patchtext(bin): |
|
242 | def patchtext(bin): | |
238 | pos = 0 |
|
243 | pos = 0 | |
239 | t = [] |
|
244 | t = [] | |
240 | while pos < len(bin): |
|
245 | while pos < len(bin): | |
241 | p1, p2, l = struct.unpack(">lll", bin[pos:pos + 12]) |
|
246 | p1, p2, l = struct.unpack(">lll", bin[pos:pos + 12]) | |
242 | pos += 12 |
|
247 | pos += 12 | |
243 | t.append(bin[pos:pos + l]) |
|
248 | t.append(bin[pos:pos + l]) | |
244 | pos += l |
|
249 | pos += l | |
245 | return "".join(t) |
|
250 | return "".join(t) | |
246 |
|
251 | |||
247 | def patch(a, bin): |
|
252 | def patch(a, bin): | |
248 | return mpatch.patches(a, [bin]) |
|
253 | return mpatch.patches(a, [bin]) | |
249 |
|
254 | |||
250 | patches = mpatch.patches |
|
255 | patches = mpatch.patches | |
251 | patchedsize = mpatch.patchedsize |
|
256 | patchedsize = mpatch.patchedsize | |
252 | textdiff = bdiff.bdiff |
|
257 | textdiff = bdiff.bdiff |
General Comments 0
You need to be logged in to leave comments.
Login now