author | Solomon Matthews <smat@fb.com> |
Thu, 15 Jan 2015 20:03:28 -0800 | |
changeset 23905 | 8b5b963ba95a |
parent 19962 | 66b21ce60a19 |
child 29010 | e868d8ee7c8f |
permissions | -rw-r--r-- |
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 | 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 |
|
11364
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
18 |
#include "util.h" |
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
19 |
|
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
20 |
struct line { |
16749
eab8ca175262
bdiff: use Py_ssize_t instead of int
Adrian Buehlmann <adrian@cadifra.com>
parents:
16477
diff
changeset
|
21 |
int hash, n, e; |
eab8ca175262
bdiff: use Py_ssize_t instead of int
Adrian Buehlmann <adrian@cadifra.com>
parents:
16477
diff
changeset
|
22 |
Py_ssize_t len; |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
23 |
const char *l; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
24 |
}; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
25 |
|
474 | 26 |
struct pos { |
27 |
int pos, len; |
|
28 |
}; |
|
29 |
||
13089
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
30 |
struct hunk; |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
31 |
struct hunk { |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
32 |
int a1, a2, b1, b2; |
13089
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
33 |
struct hunk *next; |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
34 |
}; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
35 |
|
16749
eab8ca175262
bdiff: use Py_ssize_t instead of int
Adrian Buehlmann <adrian@cadifra.com>
parents:
16477
diff
changeset
|
36 |
static int splitlines(const char *a, Py_ssize_t len, struct line **lr) |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
37 |
{ |
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
|
38 |
unsigned hash; |
13731
5d0cdf4ec338
bdiff.c: use unsigned arithmetic for hash computation
Markus F.X.J. Oberhumer <markus@oberhumer.com>
parents:
13730
diff
changeset
|
39 |
int i; |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
40 |
const char *p, *b = a; |
5340
5737845fd974
bdiff: simple splitlines optimization
Christoph Spiel <cspiel@freenet.de>
parents:
5339
diff
changeset
|
41 |
const char * const plast = a + len - 1; |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
42 |
struct line *l; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
43 |
|
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
44 |
/* count the lines */ |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
45 |
i = 1; /* extra line for sentinel */ |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
46 |
for (p = a; p < a + len; p++) |
5340
5737845fd974
bdiff: simple splitlines optimization
Christoph Spiel <cspiel@freenet.de>
parents:
5339
diff
changeset
|
47 |
if (*p == '\n' || p == plast) |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
48 |
i++; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
49 |
|
1978
10606ee61107
do proper typecasting on malloc() and calloc() calls
TK Soh <teekaysoh@yahoo.com>
parents:
1759
diff
changeset
|
50 |
*lr = l = (struct line *)malloc(sizeof(struct line) * i); |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
51 |
if (!l) |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
52 |
return -1; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
53 |
|
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
54 |
/* 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
|
55 |
hash = 0; |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
56 |
for (p = a; p < a + len; p++) { |
5342 | 57 |
/* 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
|
58 |
hash = (hash * 1664525) + (unsigned char)*p + 1013904223; |
5342 | 59 |
|
5340
5737845fd974
bdiff: simple splitlines optimization
Christoph Spiel <cspiel@freenet.de>
parents:
5339
diff
changeset
|
60 |
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
|
61 |
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
|
62 |
hash = 0; |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
63 |
l->len = p - b + 1; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
64 |
l->l = b; |
5341
458acf92b49e
bdiff: use INT_MAX to avoid some inner loop comparisons
Matt Mackall <mpm@selenic.com>
parents:
5340
diff
changeset
|
65 |
l->n = INT_MAX; |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
66 |
l++; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
67 |
b = p + 1; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
68 |
} |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
69 |
} |
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 |
/* 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
|
72 |
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
|
73 |
l->len = 0; |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
74 |
l->l = a + len; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
75 |
return i - 1; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
76 |
} |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
77 |
|
13729
4a9c09239ba1
bdiff.c: make all local functions static
Markus F.X.J. Oberhumer <markus@oberhumer.com>
parents:
13302
diff
changeset
|
78 |
static inline int cmp(struct line *a, struct line *b) |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
79 |
{ |
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
|
80 |
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
|
81 |
} |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
82 |
|
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
83 |
static int equatelines(struct line *a, int an, struct line *b, int bn) |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
84 |
{ |
5452
82b4ff3abbcd
bdiff: tweaks for large files
Matt Mackall <mpm@selenic.com>
parents:
5342
diff
changeset
|
85 |
int i, j, buckets = 1, t, scale; |
82b4ff3abbcd
bdiff: tweaks for large files
Matt Mackall <mpm@selenic.com>
parents:
5342
diff
changeset
|
86 |
struct pos *h = NULL; |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
87 |
|
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
88 |
/* 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
|
89 |
while (buckets < bn + 1) |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
90 |
buckets *= 2; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
91 |
|
5339
058e93c3d07d
I have spotted the biggest bottleneck in "bdiff.c". Actually it was
Christoph Spiel <cspiel@freenet.de>
parents:
4134
diff
changeset
|
92 |
/* 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
|
93 |
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
|
94 |
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
|
95 |
if (h) |
82b4ff3abbcd
bdiff: tweaks for large files
Matt Mackall <mpm@selenic.com>
parents:
5342
diff
changeset
|
96 |
break; |
82b4ff3abbcd
bdiff: tweaks for large files
Matt Mackall <mpm@selenic.com>
parents:
5342
diff
changeset
|
97 |
} |
5339
058e93c3d07d
I have spotted the biggest bottleneck in "bdiff.c". Actually it was
Christoph Spiel <cspiel@freenet.de>
parents:
4134
diff
changeset
|
98 |
|
474 | 99 |
if (!h) |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
100 |
return 0; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
101 |
|
5339
058e93c3d07d
I have spotted the biggest bottleneck in "bdiff.c". Actually it was
Christoph Spiel <cspiel@freenet.de>
parents:
4134
diff
changeset
|
102 |
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
|
103 |
|
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
104 |
/* clear the hash table */ |
474 | 105 |
for (i = 0; i <= buckets; i++) { |
5341
458acf92b49e
bdiff: use INT_MAX to avoid some inner loop comparisons
Matt Mackall <mpm@selenic.com>
parents:
5340
diff
changeset
|
106 |
h[i].pos = INT_MAX; |
474 | 107 |
h[i].len = 0; |
108 |
} |
|
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
109 |
|
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
110 |
/* add lines to the hash table chains */ |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
111 |
for (i = bn - 1; i >= 0; i--) { |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
112 |
/* find the equivalence class */ |
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
|
113 |
for (j = b[i].hash & buckets; h[j].pos != INT_MAX; |
474 | 114 |
j = (j + 1) & buckets) |
115 |
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
|
116 |
break; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
117 |
|
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
118 |
/* add to the head of the equivalence class */ |
474 | 119 |
b[i].n = h[j].pos; |
433
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
120 |
b[i].e = j; |
474 | 121 |
h[j].pos = i; |
122 |
h[j].len++; /* keep track of popularity */ |
|
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
123 |
} |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
124 |
|
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
125 |
/* compute popularity threshold */ |
9534
8e202431d620
bdiff: gradually enable the popularity hack
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
8858
diff
changeset
|
126 |
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
|
127 |
|
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
128 |
/* 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
|
129 |
for (i = 0; i < an; i++) { |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
130 |
/* find the equivalence class */ |
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
|
131 |
for (j = a[i].hash & buckets; h[j].pos != INT_MAX; |
474 | 132 |
j = (j + 1) & buckets) |
133 |
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
|
134 |
break; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
135 |
|
433
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
136 |
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
|
137 |
if (h[j].len <= t) |
474 | 138 |
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
|
139 |
else |
5341
458acf92b49e
bdiff: use INT_MAX to avoid some inner loop comparisons
Matt Mackall <mpm@selenic.com>
parents:
5340
diff
changeset
|
140 |
a[i].n = INT_MAX; /* too popular */ |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
141 |
} |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
142 |
|
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
143 |
/* discard hash tables */ |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
144 |
free(h); |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
145 |
return 1; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
146 |
} |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
147 |
|
474 | 148 |
static int longest_match(struct line *a, struct line *b, struct pos *pos, |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
149 |
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
|
150 |
{ |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
151 |
int mi = a1, mj = b1, mk = 0, mb = 0, i, j, k; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
152 |
|
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
153 |
for (i = a1; i < a2; i++) { |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
154 |
/* skip things before the current block */ |
5341
458acf92b49e
bdiff: use INT_MAX to avoid some inner loop comparisons
Matt Mackall <mpm@selenic.com>
parents:
5340
diff
changeset
|
155 |
for (j = a[i].n; j < b1; j = b[j].n) |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
156 |
; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
157 |
|
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
158 |
/* loop through all lines match a[i] in b */ |
5341
458acf92b49e
bdiff: use INT_MAX to avoid some inner loop comparisons
Matt Mackall <mpm@selenic.com>
parents:
5340
diff
changeset
|
159 |
for (; j < b2; j = b[j].n) { |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
160 |
/* does this extend an earlier match? */ |
474 | 161 |
if (i > a1 && j > b1 && pos[j - 1].pos == i - 1) |
162 |
k = pos[j - 1].len + 1; |
|
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
163 |
else |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
164 |
k = 1; |
474 | 165 |
pos[j].pos = i; |
166 |
pos[j].len = k; |
|
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
167 |
|
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
168 |
/* best match so far? */ |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
169 |
if (k > mk) { |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
170 |
mi = i; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
171 |
mj = j; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
172 |
mk = k; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
173 |
} |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
174 |
} |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
175 |
} |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
176 |
|
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
177 |
if (mk) { |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
178 |
mi = mi - mk + 1; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
179 |
mj = mj - mk + 1; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
180 |
} |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
181 |
|
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
182 |
/* expand match to include neighboring popular lines */ |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
183 |
while (mi - mb > a1 && mj - mb > b1 && |
433
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
184 |
a[mi - mb - 1].e == b[mj - mb - 1].e) |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
185 |
mb++; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
186 |
while (mi + mk < a2 && mj + mk < b2 && |
433
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
187 |
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
|
188 |
mk++; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
189 |
|
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
190 |
*omi = mi - mb; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
191 |
*omj = mj - mb; |
5341
458acf92b49e
bdiff: use INT_MAX to avoid some inner loop comparisons
Matt Mackall <mpm@selenic.com>
parents:
5340
diff
changeset
|
192 |
|
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
193 |
return mk + mb; |
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 |
|
13089
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
196 |
static struct hunk *recurse(struct line *a, struct line *b, struct pos *pos, |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
197 |
int a1, int a2, int b1, int b2, struct hunk *l) |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
198 |
{ |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
199 |
int i, j, k; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
200 |
|
10500
e96597c8d0ea
bdiff: do not use recursion / avoid stackoverflow (issue1940)
Alistair Bell <alistair@bellsonline.com>
parents:
10282
diff
changeset
|
201 |
while (1) { |
e96597c8d0ea
bdiff: do not use recursion / avoid stackoverflow (issue1940)
Alistair Bell <alistair@bellsonline.com>
parents:
10282
diff
changeset
|
202 |
/* 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
|
203 |
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
|
204 |
if (!k) |
13089
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
205 |
return l; |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
206 |
|
10500
e96597c8d0ea
bdiff: do not use recursion / avoid stackoverflow (issue1940)
Alistair Bell <alistair@bellsonline.com>
parents:
10282
diff
changeset
|
207 |
/* and recurse on the remaining chunks on either side */ |
13089
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
208 |
l = recurse(a, b, pos, a1, i, b1, j, l); |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
209 |
if (!l) |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
210 |
return NULL; |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
211 |
|
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
212 |
l->next = (struct hunk *)malloc(sizeof(struct hunk)); |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
213 |
if (!l->next) |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
214 |
return NULL; |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
215 |
|
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
216 |
l = l->next; |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
217 |
l->a1 = i; |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
218 |
l->a2 = i + k; |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
219 |
l->b1 = j; |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
220 |
l->b2 = j + k; |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
221 |
l->next = NULL; |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
222 |
|
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
223 |
/* 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
|
224 |
a1 = i + k; |
e96597c8d0ea
bdiff: do not use recursion / avoid stackoverflow (issue1940)
Alistair Bell <alistair@bellsonline.com>
parents:
10282
diff
changeset
|
225 |
b1 = j + k; |
e96597c8d0ea
bdiff: do not use recursion / avoid stackoverflow (issue1940)
Alistair Bell <alistair@bellsonline.com>
parents:
10282
diff
changeset
|
226 |
} |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
227 |
} |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
228 |
|
13089
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
229 |
static int diff(struct line *a, int an, struct line *b, int bn, |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
230 |
struct hunk *base) |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
231 |
{ |
7104
9514cbb6e4f6
bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
7036
diff
changeset
|
232 |
struct hunk *curr; |
474 | 233 |
struct pos *pos; |
13089
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
234 |
int t, count = 0; |
433
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
235 |
|
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
236 |
/* allocate and fill arrays */ |
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
237 |
t = equatelines(a, an, b, bn); |
5571 | 238 |
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
|
239 |
|
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
240 |
if (pos && t) { |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
241 |
/* generate the matching block list */ |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
242 |
|
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
243 |
curr = recurse(a, b, pos, 0, an, 0, bn, base); |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
244 |
if (!curr) |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
245 |
return -1; |
433
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
246 |
|
13089
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
247 |
/* sentinel end hunk */ |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
248 |
curr->next = (struct hunk *)malloc(sizeof(struct hunk)); |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
249 |
if (!curr->next) |
13090
c73745762f33
bdiff: Fix bogus NULL return
Matt Mackall <mpm@selenic.com>
parents:
13089
diff
changeset
|
250 |
return -1; |
13089
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
251 |
curr = curr->next; |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
252 |
curr->a1 = curr->a2 = an; |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
253 |
curr->b1 = curr->b2 = bn; |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
254 |
curr->next = NULL; |
433
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
255 |
} |
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
256 |
|
474 | 257 |
free(pos); |
7104
9514cbb6e4f6
bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
7036
diff
changeset
|
258 |
|
7625
930a2be7e875
bdiff: add comment about normalization
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
7189
diff
changeset
|
259 |
/* 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
|
260 |
for (curr = base->next; curr; curr = curr->next) { |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
261 |
struct hunk *next = curr->next; |
7104
9514cbb6e4f6
bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
7036
diff
changeset
|
262 |
int shift = 0; |
9514cbb6e4f6
bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
7036
diff
changeset
|
263 |
|
13089
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
264 |
if (!next) |
7104
9514cbb6e4f6
bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
7036
diff
changeset
|
265 |
break; |
9514cbb6e4f6
bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
7036
diff
changeset
|
266 |
|
9514cbb6e4f6
bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
7036
diff
changeset
|
267 |
if (curr->a2 == next->a1) |
10282
08a0f04b56bd
many, many trivial check-code fixups
Matt Mackall <mpm@selenic.com>
parents:
9534
diff
changeset
|
268 |
while (curr->a2 + shift < an && curr->b2 + shift < bn |
08a0f04b56bd
many, many trivial check-code fixups
Matt Mackall <mpm@selenic.com>
parents:
9534
diff
changeset
|
269 |
&& !cmp(a + curr->a2 + shift, |
08a0f04b56bd
many, many trivial check-code fixups
Matt Mackall <mpm@selenic.com>
parents:
9534
diff
changeset
|
270 |
b + curr->b2 + shift)) |
7104
9514cbb6e4f6
bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
7036
diff
changeset
|
271 |
shift++; |
9514cbb6e4f6
bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
7036
diff
changeset
|
272 |
else if (curr->b2 == next->b1) |
10282
08a0f04b56bd
many, many trivial check-code fixups
Matt Mackall <mpm@selenic.com>
parents:
9534
diff
changeset
|
273 |
while (curr->b2 + shift < bn && curr->a2 + shift < an |
08a0f04b56bd
many, many trivial check-code fixups
Matt Mackall <mpm@selenic.com>
parents:
9534
diff
changeset
|
274 |
&& !cmp(b + curr->b2 + shift, |
08a0f04b56bd
many, many trivial check-code fixups
Matt Mackall <mpm@selenic.com>
parents:
9534
diff
changeset
|
275 |
a + curr->a2 + shift)) |
7104
9514cbb6e4f6
bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
7036
diff
changeset
|
276 |
shift++; |
9514cbb6e4f6
bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
7036
diff
changeset
|
277 |
if (!shift) |
9514cbb6e4f6
bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
7036
diff
changeset
|
278 |
continue; |
9514cbb6e4f6
bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
7036
diff
changeset
|
279 |
curr->b2 += shift; |
9514cbb6e4f6
bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
7036
diff
changeset
|
280 |
next->b1 += shift; |
9514cbb6e4f6
bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
7036
diff
changeset
|
281 |
curr->a2 += shift; |
9514cbb6e4f6
bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
7036
diff
changeset
|
282 |
next->a1 += shift; |
9514cbb6e4f6
bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
7036
diff
changeset
|
283 |
} |
9514cbb6e4f6
bdiff: normalize the diff (issue1295)
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
7036
diff
changeset
|
284 |
|
13089
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
285 |
for (curr = base->next; curr; curr = curr->next) |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
286 |
count++; |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
287 |
return count; |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
288 |
} |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
289 |
|
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
290 |
static void freehunks(struct hunk *l) |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
291 |
{ |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
292 |
struct hunk *n; |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
293 |
for (; l; l = n) { |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
294 |
n = l->next; |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
295 |
free(l); |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
296 |
} |
433
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
297 |
} |
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
298 |
|
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
299 |
static PyObject *blocks(PyObject *self, PyObject *args) |
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
300 |
{ |
435 | 301 |
PyObject *sa, *sb, *rl = NULL, *m; |
433
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
302 |
struct line *a, *b; |
13089
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
303 |
struct hunk l, *h; |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
304 |
int an, bn, count, pos = 0; |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
305 |
|
19962
66b21ce60a19
bdiff: avoid a memory error on malloc failure
Matt Mackall <mpm@selenic.com>
parents:
18551
diff
changeset
|
306 |
l.next = NULL; |
66b21ce60a19
bdiff: avoid a memory error on malloc failure
Matt Mackall <mpm@selenic.com>
parents:
18551
diff
changeset
|
307 |
|
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
308 |
if (!PyArg_ParseTuple(args, "SS:bdiff", &sa, &sb)) |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
309 |
return NULL; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
310 |
|
11364
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
311 |
an = splitlines(PyBytes_AsString(sa), PyBytes_Size(sa), &a); |
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
312 |
bn = splitlines(PyBytes_AsString(sb), PyBytes_Size(sb), &b); |
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
313 |
|
433
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
314 |
if (!a || !b) |
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
315 |
goto nomem; |
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
316 |
|
13089
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
317 |
count = diff(a, an, b, bn, &l); |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
318 |
if (count < 0) |
433
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
319 |
goto nomem; |
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
320 |
|
13089
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
321 |
rl = PyList_New(count); |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
322 |
if (!rl) |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
323 |
goto nomem; |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
324 |
|
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
325 |
for (h = l.next; h; h = h->next) { |
433
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
326 |
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
|
327 |
PyList_SetItem(rl, pos, m); |
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
328 |
pos++; |
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
329 |
} |
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
330 |
|
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
331 |
nomem: |
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
332 |
free(a); |
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
333 |
free(b); |
13089
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
334 |
freehunks(l.next); |
433
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
335 |
return rl ? rl : PyErr_NoMemory(); |
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
336 |
} |
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
337 |
|
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
338 |
static PyObject *bdiff(PyObject *self, PyObject *args) |
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
339 |
{ |
15222
73015301db86
bdiff: fix pointer aliasing
Matt Mackall <mpm@selenic.com>
parents:
13732
diff
changeset
|
340 |
char *sa, *sb, *rb; |
3335
9061613c1593
Teach bdiff to support buffer objects
Brendan Cully <brendan@kublai.com>
parents:
2859
diff
changeset
|
341 |
PyObject *result = NULL; |
433
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
342 |
struct line *al, *bl; |
13089
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
343 |
struct hunk l, *h; |
16749
eab8ca175262
bdiff: use Py_ssize_t instead of int
Adrian Buehlmann <adrian@cadifra.com>
parents:
16477
diff
changeset
|
344 |
int an, bn, count; |
eab8ca175262
bdiff: use Py_ssize_t instead of int
Adrian Buehlmann <adrian@cadifra.com>
parents:
16477
diff
changeset
|
345 |
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
|
346 |
PyThreadState *_save; |
433
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
347 |
|
19962
66b21ce60a19
bdiff: avoid a memory error on malloc failure
Matt Mackall <mpm@selenic.com>
parents:
18551
diff
changeset
|
348 |
l.next = NULL; |
66b21ce60a19
bdiff: avoid a memory error on malloc failure
Matt Mackall <mpm@selenic.com>
parents:
18551
diff
changeset
|
349 |
|
3369
4bad632913d8
python2.5 PyArg_ParseTuple fix
Alexis S. L. Carvalho <alexis@cecm.usp.br>
parents:
3335
diff
changeset
|
350 |
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
|
351 |
return NULL; |
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
352 |
|
18551
d6fb7bbec16a
bdiff: simplify overflow checks
Matt Mackall <mpm@selenic.com>
parents:
16750
diff
changeset
|
353 |
if (la > UINT_MAX || lb > UINT_MAX) { |
d6fb7bbec16a
bdiff: simplify overflow checks
Matt Mackall <mpm@selenic.com>
parents:
16750
diff
changeset
|
354 |
PyErr_SetString(PyExc_ValueError, "bdiff inputs too large"); |
d6fb7bbec16a
bdiff: simplify overflow checks
Matt Mackall <mpm@selenic.com>
parents:
16750
diff
changeset
|
355 |
return NULL; |
d6fb7bbec16a
bdiff: simplify overflow checks
Matt Mackall <mpm@selenic.com>
parents:
16750
diff
changeset
|
356 |
} |
d6fb7bbec16a
bdiff: simplify overflow checks
Matt Mackall <mpm@selenic.com>
parents:
16750
diff
changeset
|
357 |
|
16477
70b5e25f1598
bdiff.bdiff: release the GIL before doing expensive diff operations
Augie Fackler <raf@durin42.com>
parents:
16437
diff
changeset
|
358 |
_save = PyEval_SaveThread(); |
3335
9061613c1593
Teach bdiff to support buffer objects
Brendan Cully <brendan@kublai.com>
parents:
2859
diff
changeset
|
359 |
an = splitlines(sa, la, &al); |
9061613c1593
Teach bdiff to support buffer objects
Brendan Cully <brendan@kublai.com>
parents:
2859
diff
changeset
|
360 |
bn = splitlines(sb, lb, &bl); |
433
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
361 |
if (!al || !bl) |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
362 |
goto nomem; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
363 |
|
13089
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
364 |
count = diff(al, an, bl, bn, &l); |
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
365 |
if (count < 0) |
433
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
366 |
goto nomem; |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
367 |
|
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
368 |
/* calculate length of output */ |
3335
9061613c1593
Teach bdiff to support buffer objects
Brendan Cully <brendan@kublai.com>
parents:
2859
diff
changeset
|
369 |
la = lb = 0; |
13089
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
370 |
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
|
371 |
if (h->a1 != la || h->b1 != lb) |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
372 |
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
|
373 |
la = h->a2; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
374 |
lb = h->b2; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
375 |
} |
16477
70b5e25f1598
bdiff.bdiff: release the GIL before doing expensive diff operations
Augie Fackler <raf@durin42.com>
parents:
16437
diff
changeset
|
376 |
PyEval_RestoreThread(_save); |
70b5e25f1598
bdiff.bdiff: release the GIL before doing expensive diff operations
Augie Fackler <raf@durin42.com>
parents:
16437
diff
changeset
|
377 |
_save = NULL; |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
378 |
|
11364
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
379 |
result = PyBytes_FromStringAndSize(NULL, len); |
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
380 |
|
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
381 |
if (!result) |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
382 |
goto nomem; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
383 |
|
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
384 |
/* build binary patch */ |
11364
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
385 |
rb = PyBytes_AsString(result); |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
386 |
la = lb = 0; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
387 |
|
13089
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
388 |
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
|
389 |
if (h->a1 != la || h->b1 != lb) { |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
390 |
len = bl[h->b1].l - bl[lb].l; |
18551
d6fb7bbec16a
bdiff: simplify overflow checks
Matt Mackall <mpm@selenic.com>
parents:
16750
diff
changeset
|
391 |
putbe32((uint32_t)(al[la].l - al->l), rb); |
d6fb7bbec16a
bdiff: simplify overflow checks
Matt Mackall <mpm@selenic.com>
parents:
16750
diff
changeset
|
392 |
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
|
393 |
putbe32((uint32_t)len, rb + 8); |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
394 |
memcpy(rb + 12, bl[lb].l, len); |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
395 |
rb += 12 + len; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
396 |
} |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
397 |
la = h->a2; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
398 |
lb = h->b2; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
399 |
} |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
400 |
|
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
401 |
nomem: |
16477
70b5e25f1598
bdiff.bdiff: release the GIL before doing expensive diff operations
Augie Fackler <raf@durin42.com>
parents:
16437
diff
changeset
|
402 |
if (_save) |
70b5e25f1598
bdiff.bdiff: release the GIL before doing expensive diff operations
Augie Fackler <raf@durin42.com>
parents:
16437
diff
changeset
|
403 |
PyEval_RestoreThread(_save); |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
404 |
free(al); |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
405 |
free(bl); |
13089
faee0ffbc24b
bdiff: dynamically allocate hunks
Matt Mackall <mpm@selenic.com>
parents:
11364
diff
changeset
|
406 |
freehunks(l.next); |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
407 |
return result ? result : PyErr_NoMemory(); |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
408 |
} |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
409 |
|
15530
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
410 |
/* |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
411 |
* 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
|
412 |
* 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
|
413 |
* from end of lines. |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
414 |
*/ |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
415 |
static PyObject *fixws(PyObject *self, PyObject *args) |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
416 |
{ |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
417 |
PyObject *s, *result = NULL; |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
418 |
char allws, c; |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
419 |
const char *r; |
16749
eab8ca175262
bdiff: use Py_ssize_t instead of int
Adrian Buehlmann <adrian@cadifra.com>
parents:
16477
diff
changeset
|
420 |
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
|
421 |
char *w; |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
422 |
|
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
423 |
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
|
424 |
return NULL; |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
425 |
r = PyBytes_AsString(s); |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
426 |
rlen = PyBytes_Size(s); |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
427 |
|
16071
8134ec8627e7
bdiff: fix malloc(0) issue in fixws()
Jim Hague <jim.hague@acm.org>
parents:
15530
diff
changeset
|
428 |
w = (char *)malloc(rlen ? rlen : 1); |
15530
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
429 |
if (!w) |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
430 |
goto nomem; |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
431 |
|
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
432 |
for (i = 0; i != rlen; i++) { |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
433 |
c = r[i]; |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
434 |
if (c == ' ' || c == '\t' || c == '\r') { |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
435 |
if (!allws && (wlen == 0 || w[wlen - 1] != ' ')) |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
436 |
w[wlen++] = ' '; |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
437 |
} else if (c == '\n' && !allws |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
438 |
&& wlen > 0 && w[wlen - 1] == ' ') { |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
439 |
w[wlen - 1] = '\n'; |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
440 |
} else { |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
441 |
w[wlen++] = c; |
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 |
} |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
444 |
|
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
445 |
result = PyBytes_FromStringAndSize(w, wlen); |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
446 |
|
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
447 |
nomem: |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
448 |
free(w); |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
449 |
return result ? result : PyErr_NoMemory(); |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
450 |
} |
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
451 |
|
eeac5e179243
mdiff: replace wscleanup() regexps with C loops
Patrick Mezard <pmezard@gmail.com>
parents:
15222
diff
changeset
|
452 |
|
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
453 |
static char mdiff_doc[] = "Efficient binary diff."; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
454 |
|
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
455 |
static PyMethodDef methods[] = { |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
456 |
{"bdiff", bdiff, METH_VARARGS, "calculate a binary diff\n"}, |
433
79c694462294
Add bdiff.blocks / minor performance tweaks
mpm@selenic.com
parents:
411
diff
changeset
|
457 |
{"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
|
458 |
{"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
|
459 |
{NULL, NULL} |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
460 |
}; |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
461 |
|
11364
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
462 |
#ifdef IS_PY3K |
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
463 |
static struct PyModuleDef bdiff_module = { |
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
464 |
PyModuleDef_HEAD_INIT, |
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
465 |
"bdiff", |
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
466 |
mdiff_doc, |
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
467 |
-1, |
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
468 |
methods |
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
469 |
}; |
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
470 |
|
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
471 |
PyMODINIT_FUNC PyInit_bdiff(void) |
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
472 |
{ |
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
473 |
return PyModule_Create(&bdiff_module); |
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
474 |
} |
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
475 |
#else |
400
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
476 |
PyMODINIT_FUNC initbdiff(void) |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
477 |
{ |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
478 |
Py_InitModule3("bdiff", methods, mdiff_doc); |
8b067bde6679
Add a fast binary diff extension (not yet used)
mpm@selenic.com
parents:
diff
changeset
|
479 |
} |
11364
0044193a1c45
bdiff.c: Added support for py3k.
Renato Cunha <renatoc@gmail.com>
parents:
10500
diff
changeset
|
480 |
#endif |
1542
8e80eefb3de6
made C src formatting more consistent
twaldmann@thinkmo.de
parents:
1397
diff
changeset
|
481 |