~bzr-pqm/bzr/bzr.dev

0.5.93 by Aaron Bentley
Added patches.py
1
# Copyright (C) 2004, 2005 Aaron Bentley
2
# <aaron.bentley@utoronto.ca>
3
#
4
#    This program is free software; you can redistribute it and/or modify
5
#    it under the terms of the GNU General Public License as published by
6
#    the Free Software Foundation; either version 2 of the License, or
7
#    (at your option) any later version.
8
#
9
#    This program is distributed in the hope that it will be useful,
10
#    but WITHOUT ANY WARRANTY; without even the implied warranty of
11
#    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
12
#    GNU General Public License for more details.
13
#
14
#    You should have received a copy of the GNU General Public License
15
#    along with this program; if not, write to the Free Software
16
#    Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
17
18
class PatchSyntax(Exception):
19
    def __init__(self, msg):
20
        Exception.__init__(self, msg)
21
22
23
class MalformedPatchHeader(PatchSyntax):
24
    def __init__(self, desc, line):
25
        self.desc = desc
26
        self.line = line
27
        msg = "Malformed patch header.  %s\n%r" % (self.desc, self.line)
28
        PatchSyntax.__init__(self, msg)
29
30
class MalformedHunkHeader(PatchSyntax):
31
    def __init__(self, desc, line):
32
        self.desc = desc
33
        self.line = line
34
        msg = "Malformed hunk header.  %s\n%r" % (self.desc, self.line)
35
        PatchSyntax.__init__(self, msg)
36
37
class MalformedLine(PatchSyntax):
38
    def __init__(self, desc, line):
39
        self.desc = desc
40
        self.line = line
41
        msg = "Malformed line.  %s\n%s" % (self.desc, self.line)
42
        PatchSyntax.__init__(self, msg)
43
44
def get_patch_names(iter_lines):
45
    try:
46
        line = iter_lines.next()
47
        if not line.startswith("--- "):
48
            raise MalformedPatchHeader("No orig name", line)
49
        else:
50
            orig_name = line[4:].rstrip("\n")
51
    except StopIteration:
52
        raise MalformedPatchHeader("No orig line", "")
53
    try:
54
        line = iter_lines.next()
55
        if not line.startswith("+++ "):
56
            raise PatchSyntax("No mod name")
57
        else:
58
            mod_name = line[4:].rstrip("\n")
59
    except StopIteration:
60
        raise MalformedPatchHeader("No mod line", "")
61
    return (orig_name, mod_name)
62
63
def parse_range(textrange):
64
    """Parse a patch range, handling the "1" special-case
65
66
    :param textrange: The text to parse
67
    :type textrange: str
68
    :return: the position and range, as a tuple
69
    :rtype: (int, int)
70
    """
71
    tmp = textrange.split(',')
72
    if len(tmp) == 1:
73
        pos = tmp[0]
74
        range = "1"
75
    else:
76
        (pos, range) = tmp
77
    pos = int(pos)
78
    range = int(range)
79
    return (pos, range)
80
81
 
82
def hunk_from_header(line):
83
    if not line.startswith("@@") or not line.endswith("@@\n") \
84
        or not len(line) > 4:
85
        raise MalformedHunkHeader("Does not start and end with @@.", line)
86
    try:
87
        (orig, mod) = line[3:-4].split(" ")
88
    except Exception, e:
89
        raise MalformedHunkHeader(str(e), line)
90
    if not orig.startswith('-') or not mod.startswith('+'):
91
        raise MalformedHunkHeader("Positions don't start with + or -.", line)
92
    try:
93
        (orig_pos, orig_range) = parse_range(orig[1:])
94
        (mod_pos, mod_range) = parse_range(mod[1:])
95
    except Exception, e:
96
        raise MalformedHunkHeader(str(e), line)
97
    if mod_range < 0 or orig_range < 0:
98
        raise MalformedHunkHeader("Hunk range is negative", line)
99
    return Hunk(orig_pos, orig_range, mod_pos, mod_range)
100
101
102
class HunkLine:
103
    def __init__(self, contents):
104
        self.contents = contents
105
106
    def get_str(self, leadchar):
107
        if self.contents == "\n" and leadchar == " " and False:
108
            return "\n"
109
        if not self.contents.endswith('\n'):
110
            terminator = '\n' + NO_NL
111
        else:
112
            terminator = ''
113
        return leadchar + self.contents + terminator
114
115
    def no_nl(self):
116
        assert self.contents.endswith('\n')
117
        self.contents = self.contents[:-1]
118
119
class ContextLine(HunkLine):
120
    def __init__(self, contents):
121
        HunkLine.__init__(self, contents)
122
123
    def __str__(self):
124
        return self.get_str(" ")
125
126
127
class InsertLine(HunkLine):
128
    def __init__(self, contents):
129
        HunkLine.__init__(self, contents)
130
131
    def __str__(self):
132
        return self.get_str("+")
133
134
135
class RemoveLine(HunkLine):
136
    def __init__(self, contents):
137
        HunkLine.__init__(self, contents)
138
139
    def __str__(self):
140
        return self.get_str("-")
141
142
NO_NL = '\\ No newline at end of file\n'
143
__pychecker__="no-returnvalues"
144
145
def parse_line(line):
146
    if line.startswith("\n"):
147
        return ContextLine(line)
148
    elif line.startswith(" "):
149
        return ContextLine(line[1:])
150
    elif line.startswith("+"):
151
        return InsertLine(line[1:])
152
    elif line.startswith("-"):
153
        return RemoveLine(line[1:])
154
    elif line == NO_NL:
155
        return NO_NL
156
    else:
157
        raise MalformedLine("Unknown line type", line)
158
__pychecker__=""
159
160
161
class Hunk:
162
    def __init__(self, orig_pos, orig_range, mod_pos, mod_range):
163
        self.orig_pos = orig_pos
164
        self.orig_range = orig_range
165
        self.mod_pos = mod_pos
166
        self.mod_range = mod_range
167
        self.lines = []
168
169
    def get_header(self):
170
        return "@@ -%s +%s @@\n" % (self.range_str(self.orig_pos, 
171
                                                   self.orig_range),
172
                                    self.range_str(self.mod_pos, 
173
                                                   self.mod_range))
174
175
    def range_str(self, pos, range):
176
        """Return a file range, special-casing for 1-line files.
177
178
        :param pos: The position in the file
179
        :type pos: int
180
        :range: The range in the file
181
        :type range: int
182
        :return: a string in the format 1,4 except when range == pos == 1
183
        """
184
        if range == 1:
185
            return "%i" % pos
186
        else:
187
            return "%i,%i" % (pos, range)
188
189
    def __str__(self):
190
        lines = [self.get_header()]
191
        for line in self.lines:
192
            lines.append(str(line))
193
        return "".join(lines)
194
195
    def shift_to_mod(self, pos):
196
        if pos < self.orig_pos-1:
197
            return 0
198
        elif pos > self.orig_pos+self.orig_range:
199
            return self.mod_range - self.orig_range
200
        else:
201
            return self.shift_to_mod_lines(pos)
202
203
    def shift_to_mod_lines(self, pos):
204
        assert (pos >= self.orig_pos-1 and pos <= self.orig_pos+self.orig_range)
205
        position = self.orig_pos-1
206
        shift = 0
207
        for line in self.lines:
208
            if isinstance(line, InsertLine):
209
                shift += 1
210
            elif isinstance(line, RemoveLine):
211
                if position == pos:
212
                    return None
213
                shift -= 1
214
                position += 1
215
            elif isinstance(line, ContextLine):
216
                position += 1
217
            if position > pos:
218
                break
219
        return shift
220
221
def iter_hunks(iter_lines):
222
    hunk = None
223
    for line in iter_lines:
224
        if line == NO_NL:
225
            hunk.lines[-1].no_nl()
226
            yield hunk
227
            hunk = None
228
            continue
229
        elif line == "\n":
230
            if hunk is not None:
231
                yield hunk
232
                hunk = None
233
            continue
234
        if hunk is not None:
235
            yield hunk
236
        hunk = hunk_from_header(line)
237
        orig_size = 0
238
        mod_size = 0
239
        while orig_size < hunk.orig_range or mod_size < hunk.mod_range:
240
            hunk_line = parse_line(iter_lines.next())
241
            if hunk_line is NO_NL:
242
                hunk.lines[-1].no_nl()
243
            else:
244
                hunk.lines.append(hunk_line)
245
            if isinstance(hunk_line, (RemoveLine, ContextLine)):
246
                orig_size += 1
247
            if isinstance(hunk_line, (InsertLine, ContextLine)):
248
                mod_size += 1
249
    if hunk is not None:
250
        yield hunk
251
252
class Patch:
253
    def __init__(self, oldname, newname):
254
        self.oldname = oldname
255
        self.newname = newname
256
        self.hunks = []
257
258
    def __str__(self):
259
        ret =  "--- %s\n+++ %s\n" % (self.oldname, self.newname) 
260
        ret += "".join([str(h) for h in self.hunks])
261
        return ret
262
263
    def stats_str(self):
264
        """Return a string of patch statistics"""
265
        removes = 0
266
        inserts = 0
267
        for hunk in self.hunks:
268
            for line in hunk.lines:
269
                if isinstance(line, InsertLine):
270
                     inserts+=1;
271
                elif isinstance(line, RemoveLine):
272
                     removes+=1;
273
        return "%i inserts, %i removes in %i hunks" % \
274
            (inserts, removes, len(self.hunks))
275
276
    def pos_in_mod(self, position):
277
        newpos = position
278
        for hunk in self.hunks:
279
            shift = hunk.shift_to_mod(position)
280
            if shift is None:
281
                return None
282
            newpos += shift
283
        return newpos
284
            
285
    def iter_inserted(self):
286
        """Iteraties through inserted lines
287
        
288
        :return: Pair of line number, line
289
        :rtype: iterator of (int, InsertLine)
290
        """
291
        for hunk in self.hunks:
292
            pos = hunk.mod_pos - 1;
293
            for line in hunk.lines:
294
                if isinstance(line, InsertLine):
295
                    yield (pos, line)
296
                    pos += 1
297
                if isinstance(line, ContextLine):
298
                    pos += 1
299
300
def parse_patch(iter_lines):
301
    (orig_name, mod_name) = get_patch_names(iter_lines)
302
    patch = Patch(orig_name, mod_name)
303
    for hunk in iter_hunks(iter_lines):
304
        patch.hunks.append(hunk)
305
    return patch
306
307
308
def iter_file_patch(iter_lines):
309
    saved_lines = []
310
    for line in iter_lines:
311
        if line.startswith('*** '):
312
            continue
313
        elif line.startswith('--- '):
314
            if len(saved_lines) > 0:
315
                yield saved_lines
316
            saved_lines = []
317
        saved_lines.append(line)
318
    if len(saved_lines) > 0:
319
        yield saved_lines
320
321
322
def parse_patches(iter_lines):
323
    return [parse_patch(f.__iter__()) for f in iter_file_patch(iter_lines)]
324
325
326
def difference_index(atext, btext):
327
    """Find the indext of the first character that differs betweeen two texts
328
329
    :param atext: The first text
330
    :type atext: str
331
    :param btext: The second text
332
    :type str: str
333
    :return: The index, or None if there are no differences within the range
334
    :rtype: int or NoneType
335
    """
336
    length = len(atext)
337
    if len(btext) < length:
338
        length = len(btext)
339
    for i in range(length):
340
        if atext[i] != btext[i]:
341
            return i;
342
    return None
343
344
class PatchConflict(Exception):
345
    def __init__(self, line_no, orig_line, patch_line):
346
        orig = orig_line.rstrip('\n')
347
        patch = str(patch_line).rstrip('\n')
348
        msg = 'Text contents mismatch at line %d.  Original has "%s",'\
349
            ' but patch says it should be "%s"' % (line_no, orig, patch)
350
        Exception.__init__(self, msg)
351
352
353
def iter_patched(orig_lines, patch_lines):
354
    """Iterate through a series of lines with a patch applied.
355
    This handles a single file, and does exact, not fuzzy patching.
356
    """
357
    if orig_lines is not None:
358
        orig_lines = orig_lines.__iter__()
359
    seen_patch = []
360
    patch_lines = patch_lines.__iter__()
361
    get_patch_names(patch_lines)
362
    line_no = 1
363
    for hunk in iter_hunks(patch_lines):
364
        while line_no < hunk.orig_pos:
365
            orig_line = orig_lines.next()
366
            yield orig_line
367
            line_no += 1
368
        for hunk_line in hunk.lines:
369
            seen_patch.append(str(hunk_line))
370
            if isinstance(hunk_line, InsertLine):
371
                yield hunk_line.contents
372
            elif isinstance(hunk_line, (ContextLine, RemoveLine)):
373
                orig_line = orig_lines.next()
374
                if orig_line != hunk_line.contents:
375
                    raise PatchConflict(line_no, orig_line, "".join(seen_patch))
376
                if isinstance(hunk_line, ContextLine):
377
                    yield orig_line
378
                else:
379
                    assert isinstance(hunk_line, RemoveLine)
380
                line_no += 1
381
                    
382
import unittest
383
import os.path
384
class PatchesTester(unittest.TestCase):
385
    def datafile(self, filename):
386
        data_path = os.path.join(os.path.dirname(__file__), "testdata", 
387
                                 filename)
388
        return file(data_path, "rb")
389
390
    def testValidPatchHeader(self):
391
        """Parse a valid patch header"""
392
        lines = "--- orig/commands.py\n+++ mod/dommands.py\n".split('\n')
393
        (orig, mod) = get_patch_names(lines.__iter__())
394
        assert(orig == "orig/commands.py")
395
        assert(mod == "mod/dommands.py")
396
397
    def testInvalidPatchHeader(self):
398
        """Parse an invalid patch header"""
399
        lines = "-- orig/commands.py\n+++ mod/dommands.py".split('\n')
400
        self.assertRaises(MalformedPatchHeader, get_patch_names,
401
                          lines.__iter__())
402
403
    def testValidHunkHeader(self):
404
        """Parse a valid hunk header"""
405
        header = "@@ -34,11 +50,6 @@\n"
406
        hunk = hunk_from_header(header);
407
        assert (hunk.orig_pos == 34)
408
        assert (hunk.orig_range == 11)
409
        assert (hunk.mod_pos == 50)
410
        assert (hunk.mod_range == 6)
411
        assert (str(hunk) == header)
412
413
    def testValidHunkHeader2(self):
414
        """Parse a tricky, valid hunk header"""
415
        header = "@@ -1 +0,0 @@\n"
416
        hunk = hunk_from_header(header);
417
        assert (hunk.orig_pos == 1)
418
        assert (hunk.orig_range == 1)
419
        assert (hunk.mod_pos == 0)
420
        assert (hunk.mod_range == 0)
421
        assert (str(hunk) == header)
422
423
    def makeMalformed(self, header):
424
        self.assertRaises(MalformedHunkHeader, hunk_from_header, header)
425
426
    def testInvalidHeader(self):
427
        """Parse an invalid hunk header"""
428
        self.makeMalformed(" -34,11 +50,6 \n")
429
        self.makeMalformed("@@ +50,6 -34,11 @@\n")
430
        self.makeMalformed("@@ -34,11 +50,6 @@")
431
        self.makeMalformed("@@ -34.5,11 +50,6 @@\n")
432
        self.makeMalformed("@@-34,11 +50,6@@\n")
433
        self.makeMalformed("@@ 34,11 50,6 @@\n")
434
        self.makeMalformed("@@ -34,11 @@\n")
435
        self.makeMalformed("@@ -34,11 +50,6.5 @@\n")
436
        self.makeMalformed("@@ -34,11 +50,-6 @@\n")
437
438
    def lineThing(self,text, type):
439
        line = parse_line(text)
440
        assert(isinstance(line, type))
441
        assert(str(line)==text)
442
443
    def makeMalformedLine(self, text):
444
        self.assertRaises(MalformedLine, parse_line, text)
445
446
    def testValidLine(self):
447
        """Parse a valid hunk line"""
448
        self.lineThing(" hello\n", ContextLine)
449
        self.lineThing("+hello\n", InsertLine)
450
        self.lineThing("-hello\n", RemoveLine)
451
    
452
    def testMalformedLine(self):
453
        """Parse invalid valid hunk lines"""
454
        self.makeMalformedLine("hello\n")
455
    
456
    def compare_parsed(self, patchtext):
457
        lines = patchtext.splitlines(True)
458
        patch = parse_patch(lines.__iter__())
459
        pstr = str(patch)
460
        i = difference_index(patchtext, pstr)
461
        if i is not None:
462
            print "%i: \"%s\" != \"%s\"" % (i, patchtext[i], pstr[i])
463
        self.assertEqual (patchtext, str(patch))
464
465
    def testAll(self):
466
        """Test parsing a whole patch"""
467
        patchtext = """--- orig/commands.py
468
+++ mod/commands.py
469
@@ -1337,7 +1337,8 @@
470
 
471
     def set_title(self, command=None):
472
         try:
473
-            version = self.tree.tree_version.nonarch
474
+            version = pylon.alias_or_version(self.tree.tree_version, self.tree,
475
+                                             full=False)
476
         except:
477
             version = "[no version]"
478
         if command is None:
479
@@ -1983,7 +1984,11 @@
480
                                          version)
481
         if len(new_merges) > 0:
482
             if cmdutil.prompt("Log for merge"):
483
-                mergestuff = cmdutil.log_for_merge(tree, comp_version)
484
+                if cmdutil.prompt("changelog for merge"):
485
+                    mergestuff = "Patches applied:\\n"
486
+                    mergestuff += pylon.changelog_for_merge(new_merges)
487
+                else:
488
+                    mergestuff = cmdutil.log_for_merge(tree, comp_version)
489
                 log.description += mergestuff
490
         log.save()
491
     try:
492
"""
493
        self.compare_parsed(patchtext)
494
495
    def testInit(self):
496
        """Handle patches missing half the position, range tuple"""
497
        patchtext = \
498
"""--- orig/__init__.py
499
+++ mod/__init__.py
500
@@ -1 +1,2 @@
501
 __docformat__ = "restructuredtext en"
502
+__doc__ = An alternate Arch commandline interface
503
"""
504
        self.compare_parsed(patchtext)
505
        
506
507
508
    def testLineLookup(self):
509
        import sys
510
        """Make sure we can accurately look up mod line from orig"""
511
        patch = parse_patch(self.datafile("diff"))
512
        orig = list(self.datafile("orig"))
513
        mod = list(self.datafile("mod"))
514
        removals = []
515
        for i in range(len(orig)):
516
            mod_pos = patch.pos_in_mod(i)
517
            if mod_pos is None:
518
                removals.append(orig[i])
519
                continue
520
            assert(mod[mod_pos]==orig[i])
521
        rem_iter = removals.__iter__()
522
        for hunk in patch.hunks:
523
            for line in hunk.lines:
524
                if isinstance(line, RemoveLine):
525
                    next = rem_iter.next()
526
                    if line.contents != next:
527
                        sys.stdout.write(" orig:%spatch:%s" % (next,
528
                                         line.contents))
529
                    assert(line.contents == next)
530
        self.assertRaises(StopIteration, rem_iter.next)
531
532
    def testFirstLineRenumber(self):
533
        """Make sure we handle lines at the beginning of the hunk"""
534
        patch = parse_patch(self.datafile("insert_top.patch"))
535
        assert (patch.pos_in_mod(0)==1)
536
537
def test():
538
    patchesTestSuite = unittest.makeSuite(PatchesTester,'test')
539
    runner = unittest.TextTestRunner(verbosity=0)
540
    return runner.run(patchesTestSuite)
541
    
542
543
if __name__ == "__main__":
544
    test()
545
# arch-tag: d1541a25-eac5-4de9-a476-08a7cecd5683