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