2 * LibXDiff by Davide Libenzi ( File Differential Library )
3 * Copyright (C) 2003 Davide Libenzi
5 * This library is free software; you can redistribute it and/or
6 * modify it under the terms of the GNU Lesser General Public
7 * License as published by the Free Software Foundation; either
8 * version 2.1 of the License, or (at your option) any later version.
10 * This library is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 * Lesser General Public License for more details.
15 * You should have received a copy of the GNU Lesser General Public
16 * License along with this library; if not, see
17 * <http://www.gnu.org/licenses/>.
19 * Davide Libenzi <davidel@xmailserver.org>
30 long xdl_bogosqrt(long n
) {
34 * Classical integer square root approximation using shifts.
36 for (i
= 1; n
> 0; n
>>= 2)
43 int xdl_emit_diffrec(char const *rec
, long size
, char const *pre
, long psize
,
48 mb
[0].ptr
= (char *) pre
;
50 mb
[1].ptr
= (char *) rec
;
52 if (size
> 0 && rec
[size
- 1] != '\n') {
53 mb
[2].ptr
= (char *) "\n\\ No newline at end of file\n";
54 mb
[2].size
= strlen(mb
[2].ptr
);
57 if (ecb
->outf(ecb
->priv
, mb
, i
) < 0) {
65 void *xdl_mmfile_first(mmfile_t
*mmf
, long *size
)
72 long xdl_mmfile_size(mmfile_t
*mmf
)
78 int xdl_cha_init(chastore_t
*cha
, long isize
, long icount
) {
80 cha
->head
= cha
->tail
= NULL
;
82 cha
->nsize
= icount
* isize
;
83 cha
->ancur
= cha
->sncur
= NULL
;
90 void xdl_cha_free(chastore_t
*cha
) {
93 for (cur
= cha
->head
; (tmp
= cur
) != NULL
;) {
100 void *xdl_cha_alloc(chastore_t
*cha
) {
104 if (!(ancur
= cha
->ancur
) || ancur
->icurr
== cha
->nsize
) {
105 if (!(ancur
= (chanode_t
*) xdl_malloc(sizeof(chanode_t
) + cha
->nsize
))) {
112 cha
->tail
->next
= ancur
;
119 data
= (char *) ancur
+ sizeof(chanode_t
) + ancur
->icurr
;
120 ancur
->icurr
+= cha
->isize
;
125 long xdl_guess_lines(mmfile_t
*mf
, long sample
) {
126 long nl
= 0, size
, tsize
= 0;
127 char const *data
, *cur
, *top
;
129 if ((cur
= data
= xdl_mmfile_first(mf
, &size
)) != NULL
) {
130 for (top
= data
+ size
; nl
< sample
&& cur
< top
; ) {
132 if (!(cur
= memchr(cur
, '\n', top
- cur
)))
137 tsize
+= (long) (cur
- data
);
141 nl
= xdl_mmfile_size(mf
) / (tsize
/ nl
);
146 int xdl_blankline(const char *line
, long size
, long flags
)
150 if (!(flags
& XDF_WHITESPACE_FLAGS
))
153 for (i
= 0; i
< size
&& XDL_ISSPACE(line
[i
]); i
++)
160 * Have we eaten everything on the line, except for an optional
161 * CR at the very end?
163 static int ends_with_optional_cr(const char *l
, long s
, long i
)
165 int complete
= s
&& l
[s
-1] == '\n';
171 /* do not ignore CR at the end of an incomplete line */
172 if (complete
&& s
== i
+ 1 && l
[i
] == '\r')
177 int xdl_recmatch(const char *l1
, long s1
, const char *l2
, long s2
, long flags
)
181 if (s1
== s2
&& !memcmp(l1
, l2
, s1
))
183 if (!(flags
& XDF_WHITESPACE_FLAGS
))
190 * -w matches everything that matches with -b, and -b in turn
191 * matches everything that matches with --ignore-space-at-eol,
192 * which in turn matches everything that matches with --ignore-cr-at-eol.
194 * Each flavor of ignoring needs different logic to skip whitespaces
195 * while we have both sides to compare.
197 if (flags
& XDF_IGNORE_WHITESPACE
) {
199 while (i1
< s1
&& i2
< s2
) {
200 if (l1
[i1
++] != l2
[i2
++])
203 while (i1
< s1
&& XDL_ISSPACE(l1
[i1
]))
205 while (i2
< s2
&& XDL_ISSPACE(l2
[i2
]))
208 } else if (flags
& XDF_IGNORE_WHITESPACE_CHANGE
) {
209 while (i1
< s1
&& i2
< s2
) {
210 if (XDL_ISSPACE(l1
[i1
]) && XDL_ISSPACE(l2
[i2
])) {
211 /* Skip matching spaces and try again */
212 while (i1
< s1
&& XDL_ISSPACE(l1
[i1
]))
214 while (i2
< s2
&& XDL_ISSPACE(l2
[i2
]))
218 if (l1
[i1
++] != l2
[i2
++])
221 } else if (flags
& XDF_IGNORE_WHITESPACE_AT_EOL
) {
222 while (i1
< s1
&& i2
< s2
&& l1
[i1
] == l2
[i2
]) {
226 } else if (flags
& XDF_IGNORE_CR_AT_EOL
) {
227 /* Find the first difference and see how the line ends */
228 while (i1
< s1
&& i2
< s2
&& l1
[i1
] == l2
[i2
]) {
232 return (ends_with_optional_cr(l1
, s1
, i1
) &&
233 ends_with_optional_cr(l2
, s2
, i2
));
237 * After running out of one side, the remaining side must have
238 * nothing but whitespace for the lines to match. Note that
239 * ignore-whitespace-at-eol case may break out of the loop
240 * while there still are characters remaining on both lines.
243 while (i1
< s1
&& XDL_ISSPACE(l1
[i1
]))
249 while (i2
< s2
&& XDL_ISSPACE(l2
[i2
]))
256 static unsigned long xdl_hash_record_with_whitespace(char const **data
,
257 char const *top
, long flags
) {
258 unsigned long ha
= 5381;
259 char const *ptr
= *data
;
260 int cr_at_eol_only
= (flags
& XDF_WHITESPACE_FLAGS
) == XDF_IGNORE_CR_AT_EOL
;
262 for (; ptr
< top
&& *ptr
!= '\n'; ptr
++) {
263 if (cr_at_eol_only
) {
264 /* do not ignore CR at the end of an incomplete line */
266 (ptr
+ 1 < top
&& ptr
[1] == '\n'))
269 else if (XDL_ISSPACE(*ptr
)) {
270 const char *ptr2
= ptr
;
272 while (ptr
+ 1 < top
&& XDL_ISSPACE(ptr
[1])
275 at_eol
= (top
<= ptr
+ 1 || ptr
[1] == '\n');
276 if (flags
& XDF_IGNORE_WHITESPACE
)
277 ; /* already handled */
278 else if (flags
& XDF_IGNORE_WHITESPACE_CHANGE
281 ha
^= (unsigned long) ' ';
283 else if (flags
& XDF_IGNORE_WHITESPACE_AT_EOL
285 while (ptr2
!= ptr
+ 1) {
287 ha
^= (unsigned long) *ptr2
;
294 ha
^= (unsigned long) *ptr
;
296 *data
= ptr
< top
? ptr
+ 1: ptr
;
301 unsigned long xdl_hash_record(char const **data
, char const *top
, long flags
) {
302 unsigned long ha
= 5381;
303 char const *ptr
= *data
;
305 if (flags
& XDF_WHITESPACE_FLAGS
)
306 return xdl_hash_record_with_whitespace(data
, top
, flags
);
308 for (; ptr
< top
&& *ptr
!= '\n'; ptr
++) {
310 ha
^= (unsigned long) *ptr
;
312 *data
= ptr
< top
? ptr
+ 1: ptr
;
317 unsigned int xdl_hashbits(unsigned int size
) {
318 unsigned int val
= 1, bits
= 0;
320 for (; val
< size
&& bits
< CHAR_BIT
* sizeof(unsigned int); val
<<= 1, bits
++);
321 return bits
? bits
: 1;
325 int xdl_num_out(char *out
, long val
) {
326 char *ptr
, *str
= out
;
329 ptr
= buf
+ sizeof(buf
) - 1;
335 for (; val
&& ptr
> buf
; val
/= 10)
336 *--ptr
= "0123456789"[val
% 10];
338 for (; *ptr
; ptr
++, str
++)
347 int xdl_emit_hunk_hdr(long s1
, long c1
, long s2
, long c2
,
348 const char *func
, long funclen
, xdemitcb_t
*ecb
) {
353 memcpy(buf
, "@@ -", 4);
356 nb
+= xdl_num_out(buf
+ nb
, c1
? s1
: s1
- 1);
359 memcpy(buf
+ nb
, ",", 1);
362 nb
+= xdl_num_out(buf
+ nb
, c1
);
365 memcpy(buf
+ nb
, " +", 2);
368 nb
+= xdl_num_out(buf
+ nb
, c2
? s2
: s2
- 1);
371 memcpy(buf
+ nb
, ",", 1);
374 nb
+= xdl_num_out(buf
+ nb
, c2
);
377 memcpy(buf
+ nb
, " @@", 3);
379 if (func
&& funclen
) {
381 if (funclen
> sizeof(buf
) - nb
- 1)
382 funclen
= sizeof(buf
) - nb
- 1;
383 memcpy(buf
+ nb
, func
, funclen
);
390 if (ecb
->outf(ecb
->priv
, &mb
, 1) < 0)
396 int xdl_fall_back_diff(xdfenv_t
*diff_env
, xpparam_t
const *xpp
,
397 int line1
, int count1
, int line2
, int count2
)
400 * This probably does not work outside Git, since
401 * we have a very simple mmfile structure.
403 * Note: ideally, we would reuse the prepared environment, but
404 * the libxdiff interface does not (yet) allow for diffing only
405 * ranges of lines instead of the whole files.
407 mmfile_t subfile1
, subfile2
;
410 subfile1
.ptr
= (char *)diff_env
->xdf1
.recs
[line1
- 1]->ptr
;
411 subfile1
.size
= diff_env
->xdf1
.recs
[line1
+ count1
- 2]->ptr
+
412 diff_env
->xdf1
.recs
[line1
+ count1
- 2]->size
- subfile1
.ptr
;
413 subfile2
.ptr
= (char *)diff_env
->xdf2
.recs
[line2
- 1]->ptr
;
414 subfile2
.size
= diff_env
->xdf2
.recs
[line2
+ count2
- 2]->ptr
+
415 diff_env
->xdf2
.recs
[line2
+ count2
- 2]->size
- subfile2
.ptr
;
416 if (xdl_do_diff(&subfile1
, &subfile2
, xpp
, &env
) < 0)
419 memcpy(diff_env
->xdf1
.rchg
+ line1
- 1, env
.xdf1
.rchg
, count1
);
420 memcpy(diff_env
->xdf2
.rchg
+ line2
- 1, env
.xdf2
.rchg
, count2
);