annotate mercurial/bdiff.c @ 29540:4ce1fc91e30a

bdiff: rename functions and structs to be amenable for later exporting
author Maciej Fijalkowski <fijall@gmail.com>
date Wed, 13 Jul 2016 10:07:17 +0200
parents 666832b9e154
children 9631ff5ebbeb
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
1 /*
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
2 bdiff.c - efficient binary diff extension for Mercurial
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
3
2859
345bac2bc4ec update copyrights.
Vadim Gelfer <vadim.gelfer@gmail.com>
parents: 2600
diff changeset
4 Copyright 2005, 2006 Matt Mackall <mpm@selenic.com>
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
5
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
6 This software may be used and distributed according to the terms of
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
7 the GNU General Public License, incorporated herein by reference.
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
8
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
9 Based roughly on Python difflib
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
10 */
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
11
16749
eab8ca175262 bdiff: use Py_ssize_t instead of int
Adrian Buehlmann <adrian@cadifra.com>
parents: 16477
diff changeset
12 #define PY_SSIZE_T_CLEAN
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
13 #include <Python.h>
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
14 #include <stdlib.h>
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
15 #include <string.h>
5341
458acf92b49e bdiff: use INT_MAX to avoid some inner loop comparisons
Matt Mackall <mpm@selenic.com>
parents: 5340
diff changeset
16 #include <limits.h>
867
0cd2ee61b10a Allow Mercurial to build on HP-UX 11
tksoh@users.sourceforge.net
parents: 839
diff changeset
17
29539
666832b9e154 bdiff: use ssize_t in favor of Py_ssize_t in cpython-unaware locations
Maciej Fijalkowski <fijall@gmail.com>
parents: 29444
diff changeset
18 #include "compat.h"
11364
0044193a1c45 bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents: 10500
diff changeset
19 #include "util.h"
29444
284d742e5611 internals: move the bitmanipulation routines into its own file
Maciej Fijalkowski <fijall@gmail.com>
parents: 29323
diff changeset
20 #include "bitmanipulation.h"
11364
0044193a1c45 bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents: 10500
diff changeset
21
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
22 struct bdiff_line {
16749
eab8ca175262 bdiff: use Py_ssize_t instead of int
Adrian Buehlmann <adrian@cadifra.com>
parents: 16477
diff changeset
23 int hash, n, e;
29539
666832b9e154 bdiff: use ssize_t in favor of Py_ssize_t in cpython-unaware locations
Maciej Fijalkowski <fijall@gmail.com>
parents: 29444
diff changeset
24 ssize_t len;
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
25 const char *l;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
26 };
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
27
474
b2ae8283d1a6 Minor speed improvements for bdiff
mpm@selenic.com
parents: 472
diff changeset
28 struct pos {
b2ae8283d1a6 Minor speed improvements for bdiff
mpm@selenic.com
parents: 472
diff changeset
29 int pos, len;
b2ae8283d1a6 Minor speed improvements for bdiff
mpm@selenic.com
parents: 472
diff changeset
30 };
b2ae8283d1a6 Minor speed improvements for bdiff
mpm@selenic.com
parents: 472
diff changeset
31
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
32 struct bdiff_hunk;
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
33 struct bdiff_hunk {
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
34 int a1, a2, b1, b2;
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
35 struct bdiff_hunk *next;
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
36 };
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
37
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
38 static int bdiff_splitlines(const char *a, ssize_t len, struct bdiff_line **lr)
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
39 {
13732
afe9269dccec bdiff.c: rename all variables which hold a hash value to "hash"
Markus F.X.J. Oberhumer <markus@oberhumer.com>
parents: 13731
diff changeset
40 unsigned hash;
13731
5d0cdf4ec338 bdiff.c: use unsigned arithmetic for hash computation
Markus F.X.J. Oberhumer <markus@oberhumer.com>
parents: 13730
diff changeset
41 int i;
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
42 const char *p, *b = a;
5340
5737845fd974 bdiff: simple splitlines optimization
Christoph Spiel <cspiel@freenet.de>
parents: 5339
diff changeset
43 const char * const plast = a + len - 1;
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
44 struct bdiff_line *l;
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
45
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
46 /* count the lines */
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
47 i = 1; /* extra line for sentinel */
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
48 for (p = a; p < a + len; p++)
5340
5737845fd974 bdiff: simple splitlines optimization
Christoph Spiel <cspiel@freenet.de>
parents: 5339
diff changeset
49 if (*p == '\n' || p == plast)
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
50 i++;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
51
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
52 *lr = l = (struct bdiff_line *)malloc(sizeof(struct bdiff_line) * i);
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
53 if (!l)
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
54 return -1;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
55
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
56 /* build the line array and calculate hashes */
13732
afe9269dccec bdiff.c: rename all variables which hold a hash value to "hash"
Markus F.X.J. Oberhumer <markus@oberhumer.com>
parents: 13731
diff changeset
57 hash = 0;
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
58 for (p = a; p < a + len; p++) {
5342
d0c48891dd4a bdiff: switch to lyhash
Matt Mackall <mpm@selenic.com>
parents: 5341
diff changeset
59 /* Leonid Yuriev's hash */
13732
afe9269dccec bdiff.c: rename all variables which hold a hash value to "hash"
Markus F.X.J. Oberhumer <markus@oberhumer.com>
parents: 13731
diff changeset
60 hash = (hash * 1664525) + (unsigned char)*p + 1013904223;
5342
d0c48891dd4a bdiff: switch to lyhash
Matt Mackall <mpm@selenic.com>
parents: 5341
diff changeset
61
5340
5737845fd974 bdiff: simple splitlines optimization
Christoph Spiel <cspiel@freenet.de>
parents: 5339
diff changeset
62 if (*p == '\n' || p == plast) {
13732
afe9269dccec bdiff.c: rename all variables which hold a hash value to "hash"
Markus F.X.J. Oberhumer <markus@oberhumer.com>
parents: 13731
diff changeset
63 l->hash = hash;
afe9269dccec bdiff.c: rename all variables which hold a hash value to "hash"
Markus F.X.J. Oberhumer <markus@oberhumer.com>
parents: 13731
diff changeset
64 hash = 0;
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
65 l->len = p - b + 1;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
66 l->l = b;
5341
458acf92b49e bdiff: use INT_MAX to avoid some inner loop comparisons
Matt Mackall <mpm@selenic.com>
parents: 5340
diff changeset
67 l->n = INT_MAX;
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
68 l++;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
69 b = p + 1;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
70 }
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
71 }
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
72
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
73 /* set up a sentinel */
13732
afe9269dccec bdiff.c: rename all variables which hold a hash value to "hash"
Markus F.X.J. Oberhumer <markus@oberhumer.com>
parents: 13731
diff changeset
74 l->hash = 0;
13731
5d0cdf4ec338 bdiff.c: use unsigned arithmetic for hash computation
Markus F.X.J. Oberhumer <markus@oberhumer.com>
parents: 13730
diff changeset
75 l->len = 0;
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
76 l->l = a + len;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
77 return i - 1;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
78 }
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
79
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
80 static inline int cmp(struct bdiff_line *a, struct bdiff_line *b)
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
81 {
13732
afe9269dccec bdiff.c: rename all variables which hold a hash value to "hash"
Markus F.X.J. Oberhumer <markus@oberhumer.com>
parents: 13731
diff changeset
82 return a->hash != b->hash || a->len != b->len || memcmp(a->l, b->l, a->len);
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
83 }
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
84
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
85 static int equatelines(struct bdiff_line *a, int an, struct bdiff_line *b,
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
86 int bn)
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
87 {
5452
82b4ff3abbcd bdiff: tweaks for large files
Matt Mackall <mpm@selenic.com>
parents: 5342
diff changeset
88 int i, j, buckets = 1, t, scale;
82b4ff3abbcd bdiff: tweaks for large files
Matt Mackall <mpm@selenic.com>
parents: 5342
diff changeset
89 struct pos *h = NULL;
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
90
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
91 /* build a hash table of the next highest power of 2 */
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
92 while (buckets < bn + 1)
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
93 buckets *= 2;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
94
5339
058e93c3d07d I have spotted the biggest bottleneck in "bdiff.c". Actually it was
Christoph Spiel <cspiel@freenet.de>
parents: 4134
diff changeset
95 /* try to allocate a large hash table to avoid collisions */
5452
82b4ff3abbcd bdiff: tweaks for large files
Matt Mackall <mpm@selenic.com>
parents: 5342
diff changeset
96 for (scale = 4; scale; scale /= 2) {
5339
058e93c3d07d I have spotted the biggest bottleneck in "bdiff.c". Actually it was
Christoph Spiel <cspiel@freenet.de>
parents: 4134
diff changeset
97 h = (struct pos *)malloc(scale * buckets * sizeof(struct pos));
5452
82b4ff3abbcd bdiff: tweaks for large files
Matt Mackall <mpm@selenic.com>
parents: 5342
diff changeset
98 if (h)
82b4ff3abbcd bdiff: tweaks for large files
Matt Mackall <mpm@selenic.com>
parents: 5342
diff changeset
99 break;
82b4ff3abbcd bdiff: tweaks for large files
Matt Mackall <mpm@selenic.com>
parents: 5342
diff changeset
100 }
5339
058e93c3d07d I have spotted the biggest bottleneck in "bdiff.c". Actually it was
Christoph Spiel <cspiel@freenet.de>
parents: 4134
diff changeset
101
474
b2ae8283d1a6 Minor speed improvements for bdiff
mpm@selenic.com
parents: 472
diff changeset
102 if (!h)
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
103 return 0;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
104
5339
058e93c3d07d I have spotted the biggest bottleneck in "bdiff.c". Actually it was
Christoph Spiel <cspiel@freenet.de>
parents: 4134
diff changeset
105 buckets = buckets * scale - 1;
058e93c3d07d I have spotted the biggest bottleneck in "bdiff.c". Actually it was
Christoph Spiel <cspiel@freenet.de>
parents: 4134
diff changeset
106
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
107 /* clear the hash table */
474
b2ae8283d1a6 Minor speed improvements for bdiff
mpm@selenic.com
parents: 472
diff changeset
108 for (i = 0; i <= buckets; i++) {
29013
9a8363d23419 bdiff: deal better with duplicate lines
Matt Mackall <mpm@selenic.com>
parents: 29012
diff changeset
109 h[i].pos = -1;
474
b2ae8283d1a6 Minor speed improvements for bdiff
mpm@selenic.com
parents: 472
diff changeset
110 h[i].len = 0;
b2ae8283d1a6 Minor speed improvements for bdiff
mpm@selenic.com
parents: 472
diff changeset
111 }
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
112
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
113 /* add lines to the hash table chains */
29013
9a8363d23419 bdiff: deal better with duplicate lines
Matt Mackall <mpm@selenic.com>
parents: 29012
diff changeset
114 for (i = 0; i < bn; i++) {
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
115 /* find the equivalence class */
29013
9a8363d23419 bdiff: deal better with duplicate lines
Matt Mackall <mpm@selenic.com>
parents: 29012
diff changeset
116 for (j = b[i].hash & buckets; h[j].pos != -1;
474
b2ae8283d1a6 Minor speed improvements for bdiff
mpm@selenic.com
parents: 472
diff changeset
117 j = (j + 1) & buckets)
b2ae8283d1a6 Minor speed improvements for bdiff
mpm@selenic.com
parents: 472
diff changeset
118 if (!cmp(b + i, b + h[j].pos))
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
119 break;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
120
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
121 /* add to the head of the equivalence class */
474
b2ae8283d1a6 Minor speed improvements for bdiff
mpm@selenic.com
parents: 472
diff changeset
122 b[i].n = h[j].pos;
433
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
123 b[i].e = j;
474
b2ae8283d1a6 Minor speed improvements for bdiff
mpm@selenic.com
parents: 472
diff changeset
124 h[j].pos = i;
b2ae8283d1a6 Minor speed improvements for bdiff
mpm@selenic.com
parents: 472
diff changeset
125 h[j].len++; /* keep track of popularity */
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
126 }
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
127
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
128 /* compute popularity threshold */
9534
8e202431d620 bdiff: gradually enable the popularity hack
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents: 8858
diff changeset
129 t = (bn >= 31000) ? bn / 1000 : 1000000 / (bn + 1);
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
130
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
131 /* match items in a to their equivalence class in b */
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
132 for (i = 0; i < an; i++) {
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
133 /* find the equivalence class */
29013
9a8363d23419 bdiff: deal better with duplicate lines
Matt Mackall <mpm@selenic.com>
parents: 29012
diff changeset
134 for (j = a[i].hash & buckets; h[j].pos != -1;
474
b2ae8283d1a6 Minor speed improvements for bdiff
mpm@selenic.com
parents: 472
diff changeset
135 j = (j + 1) & buckets)
b2ae8283d1a6 Minor speed improvements for bdiff
mpm@selenic.com
parents: 472
diff changeset
136 if (!cmp(a + i, b + h[j].pos))
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
137 break;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
138
433
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
139 a[i].e = j; /* use equivalence class for quick compare */
1542
8e80eefb3de6 made C src formatting more consistent
twaldmann@thinkmo.de
parents: 1397
diff changeset
140 if (h[j].len <= t)
474
b2ae8283d1a6 Minor speed improvements for bdiff
mpm@selenic.com
parents: 472
diff changeset
141 a[i].n = h[j].pos; /* point to head of match list */
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
142 else
29013
9a8363d23419 bdiff: deal better with duplicate lines
Matt Mackall <mpm@selenic.com>
parents: 29012
diff changeset
143 a[i].n = -1; /* too popular */
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
144 }
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
145
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
146 /* discard hash tables */
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
147 free(h);
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
148 return 1;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
149 }
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
150
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
151 static int longest_match(struct bdiff_line *a, struct bdiff_line *b,
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
152 struct pos *pos,
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
153 int a1, int a2, int b1, int b2, int *omi, int *omj)
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
154 {
29323
d29cb5e735e9 bdiff: remove effectively dead code
Matt Mackall <mpm@selenic.com>
parents: 29322
diff changeset
155 int mi = a1, mj = b1, mk = 0, i, j, k, half;
29015
87d4a6c5567e bdiff: further restrain potential quadratic performance
Matt Mackall <mpm@selenic.com>
parents: 29014
diff changeset
156
87d4a6c5567e bdiff: further restrain potential quadratic performance
Matt Mackall <mpm@selenic.com>
parents: 29014
diff changeset
157 /* window our search on large regions to better bound
87d4a6c5567e bdiff: further restrain potential quadratic performance
Matt Mackall <mpm@selenic.com>
parents: 29014
diff changeset
158 worst-case performance. by choosing a window at the end, we
87d4a6c5567e bdiff: further restrain potential quadratic performance
Matt Mackall <mpm@selenic.com>
parents: 29014
diff changeset
159 reduce skipping overhead on the b chains. */
87d4a6c5567e bdiff: further restrain potential quadratic performance
Matt Mackall <mpm@selenic.com>
parents: 29014
diff changeset
160 if (a2 - a1 > 30000)
87d4a6c5567e bdiff: further restrain potential quadratic performance
Matt Mackall <mpm@selenic.com>
parents: 29014
diff changeset
161 a1 = a2 - 30000;
87d4a6c5567e bdiff: further restrain potential quadratic performance
Matt Mackall <mpm@selenic.com>
parents: 29014
diff changeset
162
87d4a6c5567e bdiff: further restrain potential quadratic performance
Matt Mackall <mpm@selenic.com>
parents: 29014
diff changeset
163 half = (a1 + a2) / 2;
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
164
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
165 for (i = a1; i < a2; i++) {
29013
9a8363d23419 bdiff: deal better with duplicate lines
Matt Mackall <mpm@selenic.com>
parents: 29012
diff changeset
166 /* skip all lines in b after the current block */
9a8363d23419 bdiff: deal better with duplicate lines
Matt Mackall <mpm@selenic.com>
parents: 29012
diff changeset
167 for (j = a[i].n; j >= b2; j = b[j].n)
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
168 ;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
169
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
170 /* loop through all lines match a[i] in b */
29013
9a8363d23419 bdiff: deal better with duplicate lines
Matt Mackall <mpm@selenic.com>
parents: 29012
diff changeset
171 for (; j >= b1; j = b[j].n) {
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
172 /* does this extend an earlier match? */
29322
66dbdd3cc2b9 bdiff: extend matches across popular lines
Matt Mackall <mpm@selenic.com>
parents: 29015
diff changeset
173 for (k = 1; j - k >= b1 && i - k >= a1; k++) {
66dbdd3cc2b9 bdiff: extend matches across popular lines
Matt Mackall <mpm@selenic.com>
parents: 29015
diff changeset
174 /* reached an earlier match? */
66dbdd3cc2b9 bdiff: extend matches across popular lines
Matt Mackall <mpm@selenic.com>
parents: 29015
diff changeset
175 if (pos[j - k].pos == i - k) {
66dbdd3cc2b9 bdiff: extend matches across popular lines
Matt Mackall <mpm@selenic.com>
parents: 29015
diff changeset
176 k += pos[j - k].len;
66dbdd3cc2b9 bdiff: extend matches across popular lines
Matt Mackall <mpm@selenic.com>
parents: 29015
diff changeset
177 break;
66dbdd3cc2b9 bdiff: extend matches across popular lines
Matt Mackall <mpm@selenic.com>
parents: 29015
diff changeset
178 }
66dbdd3cc2b9 bdiff: extend matches across popular lines
Matt Mackall <mpm@selenic.com>
parents: 29015
diff changeset
179 /* previous line mismatch? */
66dbdd3cc2b9 bdiff: extend matches across popular lines
Matt Mackall <mpm@selenic.com>
parents: 29015
diff changeset
180 if (a[i - k].e != b[j - k].e)
66dbdd3cc2b9 bdiff: extend matches across popular lines
Matt Mackall <mpm@selenic.com>
parents: 29015
diff changeset
181 break;
66dbdd3cc2b9 bdiff: extend matches across popular lines
Matt Mackall <mpm@selenic.com>
parents: 29015
diff changeset
182 }
66dbdd3cc2b9 bdiff: extend matches across popular lines
Matt Mackall <mpm@selenic.com>
parents: 29015
diff changeset
183
474
b2ae8283d1a6 Minor speed improvements for bdiff
mpm@selenic.com
parents: 472
diff changeset
184 pos[j].pos = i;
b2ae8283d1a6 Minor speed improvements for bdiff
mpm@selenic.com
parents: 472
diff changeset
185 pos[j].len = k;
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
186
29014
f1ca249696ed bdiff: balance recursion to avoid quadratic behavior (issue4704)
Matt Mackall <mpm@selenic.com>
parents: 29013
diff changeset
187 /* best match so far? we prefer matches closer
f1ca249696ed bdiff: balance recursion to avoid quadratic behavior (issue4704)
Matt Mackall <mpm@selenic.com>
parents: 29013
diff changeset
188 to the middle to balance recursion */
f1ca249696ed bdiff: balance recursion to avoid quadratic behavior (issue4704)
Matt Mackall <mpm@selenic.com>
parents: 29013
diff changeset
189 if (k > mk || (k == mk && (i <= mi || i < half))) {
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
190 mi = i;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
191 mj = j;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
192 mk = k;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
193 }
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
194 }
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
195 }
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
196
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
197 if (mk) {
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
198 mi = mi - mk + 1;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
199 mj = mj - mk + 1;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
200 }
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
201
29323
d29cb5e735e9 bdiff: remove effectively dead code
Matt Mackall <mpm@selenic.com>
parents: 29322
diff changeset
202 /* expand match to include subsequent popular lines */
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
203 while (mi + mk < a2 && mj + mk < b2 &&
433
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
204 a[mi + mk].e == b[mj + mk].e)
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
205 mk++;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
206
29323
d29cb5e735e9 bdiff: remove effectively dead code
Matt Mackall <mpm@selenic.com>
parents: 29322
diff changeset
207 *omi = mi;
d29cb5e735e9 bdiff: remove effectively dead code
Matt Mackall <mpm@selenic.com>
parents: 29322
diff changeset
208 *omj = mj;
5341
458acf92b49e bdiff: use INT_MAX to avoid some inner loop comparisons
Matt Mackall <mpm@selenic.com>
parents: 5340
diff changeset
209
29323
d29cb5e735e9 bdiff: remove effectively dead code
Matt Mackall <mpm@selenic.com>
parents: 29322
diff changeset
210 return mk;
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
211 }
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
212
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
213 static struct bdiff_hunk *recurse(struct bdiff_line *a, struct bdiff_line *b,
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
214 struct pos *pos,
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
215 int a1, int a2, int b1, int b2, struct bdiff_hunk *l)
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
216 {
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
217 int i, j, k;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
218
10500
e96597c8d0ea bdiff: do not use recursion / avoid stackoverflow (issue1940)
Alistair Bell <alistair@bellsonline.com>
parents: 10282
diff changeset
219 while (1) {
e96597c8d0ea bdiff: do not use recursion / avoid stackoverflow (issue1940)
Alistair Bell <alistair@bellsonline.com>
parents: 10282
diff changeset
220 /* find the longest match in this chunk */
e96597c8d0ea bdiff: do not use recursion / avoid stackoverflow (issue1940)
Alistair Bell <alistair@bellsonline.com>
parents: 10282
diff changeset
221 k = longest_match(a, b, pos, a1, a2, b1, b2, &i, &j);
e96597c8d0ea bdiff: do not use recursion / avoid stackoverflow (issue1940)
Alistair Bell <alistair@bellsonline.com>
parents: 10282
diff changeset
222 if (!k)
13089
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
223 return l;
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
224
10500
e96597c8d0ea bdiff: do not use recursion / avoid stackoverflow (issue1940)
Alistair Bell <alistair@bellsonline.com>
parents: 10282
diff changeset
225 /* and recurse on the remaining chunks on either side */
13089
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
226 l = recurse(a, b, pos, a1, i, b1, j, l);
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
227 if (!l)
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
228 return NULL;
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
229
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
230 l->next = (struct bdiff_hunk *)malloc(sizeof(struct bdiff_hunk));
13089
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
231 if (!l->next)
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
232 return NULL;
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
233
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
234 l = l->next;
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
235 l->a1 = i;
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
236 l->a2 = i + k;
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
237 l->b1 = j;
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
238 l->b2 = j + k;
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
239 l->next = NULL;
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
240
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
241 /* tail-recursion didn't happen, so do equivalent iteration */
10500
e96597c8d0ea bdiff: do not use recursion / avoid stackoverflow (issue1940)
Alistair Bell <alistair@bellsonline.com>
parents: 10282
diff changeset
242 a1 = i + k;
e96597c8d0ea bdiff: do not use recursion / avoid stackoverflow (issue1940)
Alistair Bell <alistair@bellsonline.com>
parents: 10282
diff changeset
243 b1 = j + k;
e96597c8d0ea bdiff: do not use recursion / avoid stackoverflow (issue1940)
Alistair Bell <alistair@bellsonline.com>
parents: 10282
diff changeset
244 }
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
245 }
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
246
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
247 static int bdiff_diff(struct bdiff_line *a, int an, struct bdiff_line *b,
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
248 int bn, struct bdiff_hunk *base)
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
249 {
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
250 struct bdiff_hunk *curr;
474
b2ae8283d1a6 Minor speed improvements for bdiff
mpm@selenic.com
parents: 472
diff changeset
251 struct pos *pos;
13089
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
252 int t, count = 0;
433
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
253
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
254 /* allocate and fill arrays */
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
255 t = equatelines(a, an, b, bn);
5571
f84bb2e1cc3a fix calloc(0, ...) issue
Jim Hague <jim.hague@acm.org>
parents: 5452
diff changeset
256 pos = (struct pos *)calloc(bn ? bn : 1, sizeof(struct pos));
13089
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
257
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
258 if (pos && t) {
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
259 /* generate the matching block list */
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
260
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
261 curr = recurse(a, b, pos, 0, an, 0, bn, base);
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
262 if (!curr)
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
263 return -1;
433
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
264
13089
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
265 /* sentinel end hunk */
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
266 curr->next = (struct bdiff_hunk *)malloc(sizeof(struct bdiff_hunk));
13089
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
267 if (!curr->next)
13090
c73745762f33 bdiff: Fix bogus NULL return
Matt Mackall <mpm@selenic.com>
parents: 13089
diff changeset
268 return -1;
13089
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
269 curr = curr->next;
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
270 curr->a1 = curr->a2 = an;
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
271 curr->b1 = curr->b2 = bn;
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
272 curr->next = NULL;
433
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
273 }
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
274
474
b2ae8283d1a6 Minor speed improvements for bdiff
mpm@selenic.com
parents: 472
diff changeset
275 free(pos);
7104
9514cbb6e4f6 bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents: 7036
diff changeset
276
7625
930a2be7e875 bdiff: add comment about normalization
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents: 7189
diff changeset
277 /* normalize the hunk list, try to push each hunk towards the end */
13089
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
278 for (curr = base->next; curr; curr = curr->next) {
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
279 struct bdiff_hunk *next = curr->next;
7104
9514cbb6e4f6 bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents: 7036
diff changeset
280
13089
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
281 if (!next)
7104
9514cbb6e4f6 bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents: 7036
diff changeset
282 break;
9514cbb6e4f6 bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents: 7036
diff changeset
283
29010
e868d8ee7c8f bdiff: unify duplicate normalize loops
Matt Mackall <mpm@selenic.com>
parents: 19962
diff changeset
284 if (curr->a2 == next->a1 || curr->b2 == next->b1)
29011
8bcda4c76820 bdiff: fold in shift calculation in normalize
Matt Mackall <mpm@selenic.com>
parents: 29010
diff changeset
285 while (curr->a2 < an && curr->b2 < bn
29012
4bd67ae7d75a bdiff: fix latent normalization bug
Matt Mackall <mpm@selenic.com>
parents: 29011
diff changeset
286 && next->a1 < next->a2
4bd67ae7d75a bdiff: fix latent normalization bug
Matt Mackall <mpm@selenic.com>
parents: 29011
diff changeset
287 && next->b1 < next->b2
29011
8bcda4c76820 bdiff: fold in shift calculation in normalize
Matt Mackall <mpm@selenic.com>
parents: 29010
diff changeset
288 && !cmp(a + curr->a2, b + curr->b2)) {
8bcda4c76820 bdiff: fold in shift calculation in normalize
Matt Mackall <mpm@selenic.com>
parents: 29010
diff changeset
289 curr->a2++;
8bcda4c76820 bdiff: fold in shift calculation in normalize
Matt Mackall <mpm@selenic.com>
parents: 29010
diff changeset
290 next->a1++;
8bcda4c76820 bdiff: fold in shift calculation in normalize
Matt Mackall <mpm@selenic.com>
parents: 29010
diff changeset
291 curr->b2++;
8bcda4c76820 bdiff: fold in shift calculation in normalize
Matt Mackall <mpm@selenic.com>
parents: 29010
diff changeset
292 next->b1++;
8bcda4c76820 bdiff: fold in shift calculation in normalize
Matt Mackall <mpm@selenic.com>
parents: 29010
diff changeset
293 }
7104
9514cbb6e4f6 bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents: 7036
diff changeset
294 }
9514cbb6e4f6 bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents: 7036
diff changeset
295
13089
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
296 for (curr = base->next; curr; curr = curr->next)
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
297 count++;
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
298 return count;
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
299 }
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
300
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
301 static void bdiff_freehunks(struct bdiff_hunk *l)
13089
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
302 {
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
303 struct bdiff_hunk *n;
13089
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
304 for (; l; l = n) {
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
305 n = l->next;
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
306 free(l);
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
307 }
433
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
308 }
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
309
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
310 static PyObject *blocks(PyObject *self, PyObject *args)
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
311 {
435
e731d25ddab2 Fix a compile warning for bdiff
mpm@selenic.com
parents: 433
diff changeset
312 PyObject *sa, *sb, *rl = NULL, *m;
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
313 struct bdiff_line *a, *b;
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
314 struct bdiff_hunk l, *h;
13089
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
315 int an, bn, count, pos = 0;
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
316
19962
66b21ce60a19 bdiff: avoid a memory error on malloc failure
Matt Mackall <mpm@selenic.com>
parents: 18551
diff changeset
317 l.next = NULL;
66b21ce60a19 bdiff: avoid a memory error on malloc failure
Matt Mackall <mpm@selenic.com>
parents: 18551
diff changeset
318
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
319 if (!PyArg_ParseTuple(args, "SS:bdiff", &sa, &sb))
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
320 return NULL;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
321
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
322 an = bdiff_splitlines(PyBytes_AsString(sa), PyBytes_Size(sa), &a);
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
323 bn = bdiff_splitlines(PyBytes_AsString(sb), PyBytes_Size(sb), &b);
11364
0044193a1c45 bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents: 10500
diff changeset
324
433
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
325 if (!a || !b)
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
326 goto nomem;
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
327
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
328 count = bdiff_diff(a, an, b, bn, &l);
13089
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
329 if (count < 0)
433
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
330 goto nomem;
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
331
13089
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
332 rl = PyList_New(count);
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
333 if (!rl)
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
334 goto nomem;
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
335
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
336 for (h = l.next; h; h = h->next) {
433
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
337 m = Py_BuildValue("iiii", h->a1, h->a2, h->b1, h->b2);
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
338 PyList_SetItem(rl, pos, m);
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
339 pos++;
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
340 }
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
341
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
342 nomem:
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
343 free(a);
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
344 free(b);
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
345 bdiff_freehunks(l.next);
433
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
346 return rl ? rl : PyErr_NoMemory();
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
347 }
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
348
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
349 static PyObject *bdiff(PyObject *self, PyObject *args)
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
350 {
15222
73015301db86 bdiff: fix pointer aliasing
Matt Mackall <mpm@selenic.com>
parents: 13732
diff changeset
351 char *sa, *sb, *rb;
3335
9061613c1593 Teach bdiff to support buffer objects
Brendan Cully <brendan@kublai.com>
parents: 2859
diff changeset
352 PyObject *result = NULL;
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
353 struct bdiff_line *al, *bl;
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
354 struct bdiff_hunk l, *h;
16749
eab8ca175262 bdiff: use Py_ssize_t instead of int
Adrian Buehlmann <adrian@cadifra.com>
parents: 16477
diff changeset
355 int an, bn, count;
eab8ca175262 bdiff: use Py_ssize_t instead of int
Adrian Buehlmann <adrian@cadifra.com>
parents: 16477
diff changeset
356 Py_ssize_t len = 0, la, lb;
16477
70b5e25f1598 bdiff.bdiff: release the GIL before doing expensive diff operations
Augie Fackler <raf@durin42.com>
parents: 16437
diff changeset
357 PyThreadState *_save;
433
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
358
19962
66b21ce60a19 bdiff: avoid a memory error on malloc failure
Matt Mackall <mpm@selenic.com>
parents: 18551
diff changeset
359 l.next = NULL;
66b21ce60a19 bdiff: avoid a memory error on malloc failure
Matt Mackall <mpm@selenic.com>
parents: 18551
diff changeset
360
3369
4bad632913d8 python2.5 PyArg_ParseTuple fix
Alexis S. L. Carvalho <alexis@cecm.usp.br>
parents: 3335
diff changeset
361 if (!PyArg_ParseTuple(args, "s#s#:bdiff", &sa, &la, &sb, &lb))
433
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
362 return NULL;
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
363
18551
d6fb7bbec16a bdiff: simplify overflow checks
Matt Mackall <mpm@selenic.com>
parents: 16750
diff changeset
364 if (la > UINT_MAX || lb > UINT_MAX) {
d6fb7bbec16a bdiff: simplify overflow checks
Matt Mackall <mpm@selenic.com>
parents: 16750
diff changeset
365 PyErr_SetString(PyExc_ValueError, "bdiff inputs too large");
d6fb7bbec16a bdiff: simplify overflow checks
Matt Mackall <mpm@selenic.com>
parents: 16750
diff changeset
366 return NULL;
d6fb7bbec16a bdiff: simplify overflow checks
Matt Mackall <mpm@selenic.com>
parents: 16750
diff changeset
367 }
d6fb7bbec16a bdiff: simplify overflow checks
Matt Mackall <mpm@selenic.com>
parents: 16750
diff changeset
368
16477
70b5e25f1598 bdiff.bdiff: release the GIL before doing expensive diff operations
Augie Fackler <raf@durin42.com>
parents: 16437
diff changeset
369 _save = PyEval_SaveThread();
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
370 an = bdiff_splitlines(sa, la, &al);
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
371 bn = bdiff_splitlines(sb, lb, &bl);
433
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
372 if (!al || !bl)
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
373 goto nomem;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
374
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
375 count = bdiff_diff(al, an, bl, bn, &l);
13089
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
376 if (count < 0)
433
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
377 goto nomem;
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
378
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
379 /* calculate length of output */
3335
9061613c1593 Teach bdiff to support buffer objects
Brendan Cully <brendan@kublai.com>
parents: 2859
diff changeset
380 la = lb = 0;
13089
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
381 for (h = l.next; h; h = h->next) {
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
382 if (h->a1 != la || h->b1 != lb)
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
383 len += 12 + bl[h->b1].l - bl[lb].l;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
384 la = h->a2;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
385 lb = h->b2;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
386 }
16477
70b5e25f1598 bdiff.bdiff: release the GIL before doing expensive diff operations
Augie Fackler <raf@durin42.com>
parents: 16437
diff changeset
387 PyEval_RestoreThread(_save);
70b5e25f1598 bdiff.bdiff: release the GIL before doing expensive diff operations
Augie Fackler <raf@durin42.com>
parents: 16437
diff changeset
388 _save = NULL;
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
389
11364
0044193a1c45 bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents: 10500
diff changeset
390 result = PyBytes_FromStringAndSize(NULL, len);
0044193a1c45 bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents: 10500
diff changeset
391
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
392 if (!result)
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
393 goto nomem;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
394
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
395 /* build binary patch */
11364
0044193a1c45 bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents: 10500
diff changeset
396 rb = PyBytes_AsString(result);
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
397 la = lb = 0;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
398
13089
faee0ffbc24b bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents: 11364
diff changeset
399 for (h = l.next; h; h = h->next) {
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
400 if (h->a1 != la || h->b1 != lb) {
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
401 len = bl[h->b1].l - bl[lb].l;
18551
d6fb7bbec16a bdiff: simplify overflow checks
Matt Mackall <mpm@selenic.com>
parents: 16750
diff changeset
402 putbe32((uint32_t)(al[la].l - al->l), rb);
d6fb7bbec16a bdiff: simplify overflow checks
Matt Mackall <mpm@selenic.com>
parents: 16750
diff changeset
403 putbe32((uint32_t)(al[h->a1].l - al->l), rb + 4);
d6fb7bbec16a bdiff: simplify overflow checks
Matt Mackall <mpm@selenic.com>
parents: 16750
diff changeset
404 putbe32((uint32_t)len, rb + 8);
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
405 memcpy(rb + 12, bl[lb].l, len);
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
406 rb += 12 + len;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
407 }
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
408 la = h->a2;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
409 lb = h->b2;
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
410 }
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
411
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
412 nomem:
16477
70b5e25f1598 bdiff.bdiff: release the GIL before doing expensive diff operations
Augie Fackler <raf@durin42.com>
parents: 16437
diff changeset
413 if (_save)
70b5e25f1598 bdiff.bdiff: release the GIL before doing expensive diff operations
Augie Fackler <raf@durin42.com>
parents: 16437
diff changeset
414 PyEval_RestoreThread(_save);
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
415 free(al);
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
416 free(bl);
29540
4ce1fc91e30a bdiff: rename functions and structs to be amenable for later exporting
Maciej Fijalkowski <fijall@gmail.com>
parents: 29539
diff changeset
417 bdiff_freehunks(l.next);
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
418 return result ? result : PyErr_NoMemory();
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
419 }
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
420
15530
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
421 /*
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
422 * If allws != 0, remove all whitespace (' ', \t and \r). Otherwise,
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
423 * reduce whitespace sequences to a single space and trim remaining whitespace
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
424 * from end of lines.
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
425 */
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
426 static PyObject *fixws(PyObject *self, PyObject *args)
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
427 {
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
428 PyObject *s, *result = NULL;
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
429 char allws, c;
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
430 const char *r;
16749
eab8ca175262 bdiff: use Py_ssize_t instead of int
Adrian Buehlmann <adrian@cadifra.com>
parents: 16477
diff changeset
431 Py_ssize_t i, rlen, wlen = 0;
15530
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
432 char *w;
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
433
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
434 if (!PyArg_ParseTuple(args, "Sb:fixws", &s, &allws))
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
435 return NULL;
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
436 r = PyBytes_AsString(s);
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
437 rlen = PyBytes_Size(s);
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
438
16071
8134ec8627e7 bdiff: fix malloc(0) issue in fixws()
Jim Hague <jim.hague@acm.org>
parents: 15530
diff changeset
439 w = (char *)malloc(rlen ? rlen : 1);
15530
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
440 if (!w)
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
441 goto nomem;
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
442
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
443 for (i = 0; i != rlen; i++) {
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
444 c = r[i];
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
445 if (c == ' ' || c == '\t' || c == '\r') {
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
446 if (!allws && (wlen == 0 || w[wlen - 1] != ' '))
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
447 w[wlen++] = ' ';
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
448 } else if (c == '\n' && !allws
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
449 && wlen > 0 && w[wlen - 1] == ' ') {
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
450 w[wlen - 1] = '\n';
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
451 } else {
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
452 w[wlen++] = c;
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
453 }
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
454 }
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
455
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
456 result = PyBytes_FromStringAndSize(w, wlen);
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
457
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
458 nomem:
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
459 free(w);
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
460 return result ? result : PyErr_NoMemory();
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
461 }
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
462
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
463
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
464 static char mdiff_doc[] = "Efficient binary diff.";
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
465
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
466 static PyMethodDef methods[] = {
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
467 {"bdiff", bdiff, METH_VARARGS, "calculate a binary diff\n"},
433
79c694462294 Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents: 411
diff changeset
468 {"blocks", blocks, METH_VARARGS, "find a list of matching lines\n"},
15530
eeac5e179243 mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents: 15222
diff changeset
469 {"fixws", fixws, METH_VARARGS, "normalize diff whitespaces\n"},
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
470 {NULL, NULL}
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
471 };
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
472
11364
0044193a1c45 bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents: 10500
diff changeset
473 #ifdef IS_PY3K
0044193a1c45 bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents: 10500
diff changeset
474 static struct PyModuleDef bdiff_module = {
0044193a1c45 bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents: 10500
diff changeset
475 PyModuleDef_HEAD_INIT,
0044193a1c45 bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents: 10500
diff changeset
476 "bdiff",
0044193a1c45 bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents: 10500
diff changeset
477 mdiff_doc,
0044193a1c45 bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents: 10500
diff changeset
478 -1,
0044193a1c45 bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents: 10500
diff changeset
479 methods
0044193a1c45 bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents: 10500
diff changeset
480 };
0044193a1c45 bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents: 10500
diff changeset
481
0044193a1c45 bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents: 10500
diff changeset
482 PyMODINIT_FUNC PyInit_bdiff(void)
0044193a1c45 bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents: 10500
diff changeset
483 {
0044193a1c45 bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents: 10500
diff changeset
484 return PyModule_Create(&bdiff_module);
0044193a1c45 bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents: 10500
diff changeset
485 }
0044193a1c45 bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents: 10500
diff changeset
486 #else
400
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
487 PyMODINIT_FUNC initbdiff(void)
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
488 {
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
489 Py_InitModule3("bdiff", methods, mdiff_doc);
8b067bde6679 Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff changeset
490 }
11364
0044193a1c45 bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents: 10500
diff changeset
491 #endif
1542
8e80eefb3de6 made C src formatting more consistent
twaldmann@thinkmo.de
parents: 1397
diff changeset
492