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>
26 long xdl_bogosqrt(long n) {
30 * Classical integer square root approximation using shifts.
32 for (i = 1; n > 0; n >>= 2)
39 int xdl_emit_diffrec(char const *rec, long size, char const *pre, long psize,
44 mb[0].ptr = (char *) pre;
46 mb[1].ptr = (char *) rec;
48 if (size > 0 && rec[size - 1] != '\n') {
49 mb[2].ptr = (char *) "\n\\ No newline at end of file\n";
50 mb[2].size = strlen(mb[2].ptr);
53 if (ecb->out_line(ecb->priv, mb, i) < 0) {
61 void *xdl_mmfile_first(mmfile_t *mmf, long *size)
68 long xdl_mmfile_size(mmfile_t *mmf)
74 int xdl_cha_init(chastore_t *cha, long isize, long icount) {
76 cha->head = cha->tail = NULL;
78 cha->nsize = icount * isize;
79 cha->ancur = cha->sncur = NULL;
86 void xdl_cha_free(chastore_t *cha) {
89 for (cur = cha->head; (tmp = cur) != NULL;) {
96 void *xdl_cha_alloc(chastore_t *cha) {
100 if (!(ancur = cha->ancur) || ancur->icurr == cha->nsize) {
101 if (!(ancur = (chanode_t *) xdl_malloc(sizeof(chanode_t) + cha->nsize))) {
108 cha->tail->next = ancur;
115 data = (char *) ancur + sizeof(chanode_t) + ancur->icurr;
116 ancur->icurr += cha->isize;
121 long xdl_guess_lines(mmfile_t *mf, long sample) {
122 long nl = 0, size, tsize = 0;
123 char const *data, *cur, *top;
125 if ((cur = data = xdl_mmfile_first(mf, &size)) != NULL) {
126 for (top = data + size; nl < sample && cur < top; ) {
128 if (!(cur = memchr(cur, '\n', top - cur)))
133 tsize += (long) (cur - data);
137 nl = xdl_mmfile_size(mf) / (tsize / nl);
142 int xdl_blankline(const char *line, long size, long flags)
146 if (!(flags & XDF_WHITESPACE_FLAGS))
149 for (i = 0; i < size && XDL_ISSPACE(line[i]); i++)
156 * Have we eaten everything on the line, except for an optional
157 * CR at the very end?
159 static int ends_with_optional_cr(const char *l, long s, long i)
161 int complete = s && l[s-1] == '\n';
167 /* do not ignore CR at the end of an incomplete line */
168 if (complete && s == i + 1 && l[i] == '\r')
173 int xdl_recmatch(const char *l1, long s1, const char *l2, long s2, long flags)
177 if (s1 == s2 && !memcmp(l1, l2, s1))
179 if (!(flags & XDF_WHITESPACE_FLAGS))
186 * -w matches everything that matches with -b, and -b in turn
187 * matches everything that matches with --ignore-space-at-eol,
188 * which in turn matches everything that matches with --ignore-cr-at-eol.
190 * Each flavor of ignoring needs different logic to skip whitespaces
191 * while we have both sides to compare.
193 if (flags & XDF_IGNORE_WHITESPACE) {
195 while (i1 < s1 && i2 < s2) {
196 if (l1[i1++] != l2[i2++])
199 while (i1 < s1 && XDL_ISSPACE(l1[i1]))
201 while (i2 < s2 && XDL_ISSPACE(l2[i2]))
204 } else if (flags & XDF_IGNORE_WHITESPACE_CHANGE) {
205 while (i1 < s1 && i2 < s2) {
206 if (XDL_ISSPACE(l1[i1]) && XDL_ISSPACE(l2[i2])) {
207 /* Skip matching spaces and try again */
208 while (i1 < s1 && XDL_ISSPACE(l1[i1]))
210 while (i2 < s2 && XDL_ISSPACE(l2[i2]))
214 if (l1[i1++] != l2[i2++])
217 } else if (flags & XDF_IGNORE_WHITESPACE_AT_EOL) {
218 while (i1 < s1 && i2 < s2 && l1[i1] == l2[i2]) {
222 } else if (flags & XDF_IGNORE_CR_AT_EOL) {
223 /* Find the first difference and see how the line ends */
224 while (i1 < s1 && i2 < s2 && l1[i1] == l2[i2]) {
228 return (ends_with_optional_cr(l1, s1, i1) &&
229 ends_with_optional_cr(l2, s2, i2));
233 * After running out of one side, the remaining side must have
234 * nothing but whitespace for the lines to match. Note that
235 * ignore-whitespace-at-eol case may break out of the loop
236 * while there still are characters remaining on both lines.
239 while (i1 < s1 && XDL_ISSPACE(l1[i1]))
245 while (i2 < s2 && XDL_ISSPACE(l2[i2]))
252 static unsigned long xdl_hash_record_with_whitespace(char const **data,
253 char const *top, long flags) {
254 unsigned long ha = 5381;
255 char const *ptr = *data;
256 int cr_at_eol_only = (flags & XDF_WHITESPACE_FLAGS) == XDF_IGNORE_CR_AT_EOL;
258 for (; ptr < top && *ptr != '\n'; ptr++) {
259 if (cr_at_eol_only) {
260 /* do not ignore CR at the end of an incomplete line */
262 (ptr + 1 < top && ptr[1] == '\n'))
265 else if (XDL_ISSPACE(*ptr)) {
266 const char *ptr2 = ptr;
268 while (ptr + 1 < top && XDL_ISSPACE(ptr[1])
271 at_eol = (top <= ptr + 1 || ptr[1] == '\n');
272 if (flags & XDF_IGNORE_WHITESPACE)
273 ; /* already handled */
274 else if (flags & XDF_IGNORE_WHITESPACE_CHANGE
277 ha ^= (unsigned long) ' ';
279 else if (flags & XDF_IGNORE_WHITESPACE_AT_EOL
281 while (ptr2 != ptr + 1) {
283 ha ^= (unsigned long) *ptr2;
290 ha ^= (unsigned long) *ptr;
292 *data = ptr < top ? ptr + 1: ptr;
297 unsigned long xdl_hash_record(char const **data, char const *top, long flags) {
298 unsigned long ha = 5381;
299 char const *ptr = *data;
301 if (flags & XDF_WHITESPACE_FLAGS)
302 return xdl_hash_record_with_whitespace(data, top, flags);
304 for (; ptr < top && *ptr != '\n'; ptr++) {
306 ha ^= (unsigned long) *ptr;
308 *data = ptr < top ? ptr + 1: ptr;
313 unsigned int xdl_hashbits(unsigned int size) {
314 unsigned int val = 1, bits = 0;
316 for (; val < size && bits < CHAR_BIT * sizeof(unsigned int); val <<= 1, bits++);
317 return bits ? bits: 1;
321 int xdl_num_out(char *out, long val) {
322 char *ptr, *str = out;
325 ptr = buf + sizeof(buf) - 1;
331 for (; val && ptr > buf; val /= 10)
332 *--ptr = "0123456789"[val % 10];
334 for (; *ptr; ptr++, str++)
343 static int xdl_format_hunk_hdr(long s1, long c1, long s2, long c2,
344 const char *func, long funclen,
350 memcpy(buf, "@@ -", 4);
353 nb += xdl_num_out(buf + nb, c1 ? s1: s1 - 1);
356 memcpy(buf + nb, ",", 1);
359 nb += xdl_num_out(buf + nb, c1);
362 memcpy(buf + nb, " +", 2);
365 nb += xdl_num_out(buf + nb, c2 ? s2: s2 - 1);
368 memcpy(buf + nb, ",", 1);
371 nb += xdl_num_out(buf + nb, c2);
374 memcpy(buf + nb, " @@", 3);
376 if (func && funclen) {
378 if (funclen > sizeof(buf) - nb - 1)
379 funclen = sizeof(buf) - nb - 1;
380 memcpy(buf + nb, func, funclen);
387 if (ecb->out_line(ecb->priv, &mb, 1) < 0)
392 int xdl_emit_hunk_hdr(long s1, long c1, long s2, long c2,
393 const char *func, long funclen,
396 return xdl_format_hunk_hdr(s1, c1, s2, c2, func, funclen, ecb);
397 if (ecb->out_hunk(ecb->priv,
398 c1 ? s1 : s1 - 1, c1,
399 c2 ? s2 : s2 - 1, c2,
405 int xdl_fall_back_diff(xdfenv_t *diff_env, xpparam_t const *xpp,
406 int line1, int count1, int line2, int count2)
409 * This probably does not work outside Git, since
410 * we have a very simple mmfile structure.
412 * Note: ideally, we would reuse the prepared environment, but
413 * the libxdiff interface does not (yet) allow for diffing only
414 * ranges of lines instead of the whole files.
416 mmfile_t subfile1, subfile2;
419 subfile1.ptr = (char *)diff_env->xdf1.recs[line1 - 1]->ptr;
420 subfile1.size = diff_env->xdf1.recs[line1 + count1 - 2]->ptr +
421 diff_env->xdf1.recs[line1 + count1 - 2]->size - subfile1.ptr;
422 subfile2.ptr = (char *)diff_env->xdf2.recs[line2 - 1]->ptr;
423 subfile2.size = diff_env->xdf2.recs[line2 + count2 - 2]->ptr +
424 diff_env->xdf2.recs[line2 + count2 - 2]->size - subfile2.ptr;
425 if (xdl_do_diff(&subfile1, &subfile2, xpp, &env) < 0)
428 memcpy(diff_env->xdf1.rchg + line1 - 1, env.xdf1.rchg, count1);
429 memcpy(diff_env->xdf2.rchg + line2 - 1, env.xdf2.rchg, count2);