3 #include "run-command.h"
6 * convert.c - convert a file when checking it out and checking it in.
8 * This should use the pathname to decide on whether it wants to do some
9 * more interesting conversions (automatic gzip/unzip, general format
10 * conversions etc etc), but by default it just does automatic CRLF<->LF
11 * translation when the "auto_crlf" option is set.
14 #define CRLF_GUESS (-1)
20 /* CR, LF and CRLF counts */
21 unsigned cr, lf, crlf;
23 /* These are just approximations! */
24 unsigned printable, nonprintable;
27 static void gather_stats(const char *buf, unsigned long size, struct text_stat *stats)
31 memset(stats, 0, sizeof(*stats));
33 for (i = 0; i < size; i++) {
34 unsigned char c = buf[i];
37 if (i+1 < size && buf[i+1] == '\n')
47 stats->nonprintable++;
50 /* BS, HT, ESC and FF */
51 case '\b': case '\t': case '\033': case '\014':
55 stats->nonprintable++;
64 * The same heuristics as diff.c::mmfile_is_binary()
66 static int is_binary(unsigned long size, struct text_stat *stats)
69 if ((stats->printable >> 7) < stats->nonprintable)
72 * Other heuristics? Average line length might be relevant,
73 * as might LF vs CR vs CRLF counts..
75 * NOTE! It might be normal to have a low ratio of CRLF to LF
76 * (somebody starts with a LF-only file and edits it with an editor
77 * that adds CRLF only to lines that are added..). But do we
78 * want to support CR-only? Probably not.
83 static int crlf_to_git(const char *path, const char *src, size_t len,
84 struct strbuf *buf, int action)
86 struct text_stat stats;
89 if ((action == CRLF_BINARY) || !auto_crlf || !len)
92 gather_stats(src, len, &stats);
93 /* No CR? Nothing to convert, regardless. */
97 if (action == CRLF_GUESS) {
99 * We're currently not going to even try to convert stuff
100 * that has bare CR characters. Does anybody do that crazy
103 if (stats.cr != stats.crlf)
107 * And add some heuristics for binary vs text, of course...
109 if (is_binary(len, &stats))
113 /* only grow if not in place */
114 if (strbuf_avail(buf) + buf->len < len)
115 strbuf_grow(buf, len - buf->len);
117 if (action == CRLF_GUESS) {
119 * If we guessed, we already know we rejected a file with
120 * lone CR, and we can strip a CR without looking at what
124 unsigned char c = *src++;
130 unsigned char c = *src++;
131 if (! (c == '\r' && (1 < len && *src == '\n')))
135 strbuf_setlen(buf, dst - buf->buf);
139 static int crlf_to_worktree(const char *path, const char *src, size_t len,
140 struct strbuf *buf, int action)
142 char *to_free = NULL;
143 struct text_stat stats;
145 if ((action == CRLF_BINARY) || (action == CRLF_INPUT) ||
152 gather_stats(src, len, &stats);
154 /* No LF? Nothing to convert, regardless. */
158 /* Was it already in CRLF format? */
159 if (stats.lf == stats.crlf)
162 if (action == CRLF_GUESS) {
163 /* If we have any bare CR characters, we're not going to touch it */
164 if (stats.cr != stats.crlf)
167 if (is_binary(len, &stats))
171 /* are we "faking" in place editing ? */
173 to_free = strbuf_detach(buf, NULL);
175 strbuf_grow(buf, len + stats.lf - stats.crlf);
177 const char *nl = memchr(src, '\n', len);
180 if (nl > src && nl[-1] == '\r') {
181 strbuf_add(buf, src, nl + 1 - src);
183 strbuf_add(buf, src, nl - src);
184 strbuf_addstr(buf, "\r\n");
189 strbuf_add(buf, src, len);
195 static int filter_buffer(int fd, const char *src,
196 unsigned long size, const char *cmd)
199 * Spawn cmd and feed the buffer contents through its stdin.
201 struct child_process child_process;
202 int write_err, status;
203 const char *argv[] = { "sh", "-c", cmd, NULL };
205 memset(&child_process, 0, sizeof(child_process));
206 child_process.argv = argv;
207 child_process.in = -1;
208 child_process.out = fd;
210 if (start_command(&child_process))
211 return error("cannot fork to run external filter %s", cmd);
213 write_err = (write_in_full(child_process.in, src, size) < 0);
214 if (close(child_process.in))
217 error("cannot feed the input to external filter %s", cmd);
219 status = finish_command(&child_process);
221 error("external filter %s failed %d", cmd, -status);
222 return (write_err || status);
225 static int apply_filter(const char *path, const char *src, size_t len,
226 struct strbuf *dst, const char *cmd)
229 * Create a pipeline to have the command filter the buffer's
232 * (child --> cmd) --> us
236 struct child_process child_process;
242 memset(&child_process, 0, sizeof(child_process));
244 if (pipe(pipe_feed) < 0) {
245 error("cannot create pipe to run external filter %s", cmd);
250 child_process.pid = fork();
251 if (child_process.pid < 0) {
252 error("cannot fork to run external filter %s", cmd);
257 if (!child_process.pid) {
259 exit(filter_buffer(pipe_feed[1], src, len, cmd));
263 strbuf_init(&nbuf, 0);
264 if (strbuf_read(&nbuf, pipe_feed[0], len) < 0) {
265 error("read from external filter %s failed", cmd);
268 if (close(pipe_feed[0])) {
269 error("read from external filter %s failed", cmd);
272 status = finish_command(&child_process);
274 error("external filter %s failed %d", cmd, -status);
279 strbuf_swap(dst, &nbuf);
281 strbuf_release(&nbuf);
285 static struct convert_driver {
287 struct convert_driver *next;
290 } *user_convert, **user_convert_tail;
292 static int read_convert_config(const char *var, const char *value)
294 const char *ep, *name;
296 struct convert_driver *drv;
299 * External conversion drivers are configured using
300 * "filter.<name>.variable".
302 if (prefixcmp(var, "filter.") || (ep = strrchr(var, '.')) == var + 6)
306 for (drv = user_convert; drv; drv = drv->next)
307 if (!strncmp(drv->name, name, namelen) && !drv->name[namelen])
310 drv = xcalloc(1, sizeof(struct convert_driver));
311 drv->name = xmemdupz(name, namelen);
312 *user_convert_tail = drv;
313 user_convert_tail = &(drv->next);
319 * filter.<name>.smudge and filter.<name>.clean specifies
324 * The command-line will not be interpolated in any way.
327 if (!strcmp("smudge", ep)) {
329 return error("%s: lacks value", var);
330 drv->smudge = strdup(value);
334 if (!strcmp("clean", ep)) {
336 return error("%s: lacks value", var);
337 drv->clean = strdup(value);
343 static void setup_convert_check(struct git_attr_check *check)
345 static struct git_attr *attr_crlf;
346 static struct git_attr *attr_ident;
347 static struct git_attr *attr_filter;
350 attr_crlf = git_attr("crlf", 4);
351 attr_ident = git_attr("ident", 5);
352 attr_filter = git_attr("filter", 6);
353 user_convert_tail = &user_convert;
354 git_config(read_convert_config);
356 check[0].attr = attr_crlf;
357 check[1].attr = attr_ident;
358 check[2].attr = attr_filter;
361 static int count_ident(const char *cp, unsigned long size)
364 * "$Id: 0000000000000000000000000000000000000000 $" <=> "$Id$"
376 if (memcmp("Id", cp, 2))
387 * "$Id: ... "; scan up to the closing dollar sign and discard.
401 static int ident_to_git(const char *path, const char *src, size_t len,
402 struct strbuf *buf, int ident)
406 if (!ident || !count_ident(src, len))
409 /* only grow if not in place */
410 if (strbuf_avail(buf) + buf->len < len)
411 strbuf_grow(buf, len - buf->len);
414 dollar = memchr(src, '$', len);
417 memcpy(dst, src, dollar + 1 - src);
418 dst += dollar + 1 - src;
419 len -= dollar + 1 - src;
422 if (len > 3 && !memcmp(src, "Id:", 3)) {
423 dollar = memchr(src + 3, '$', len - 3);
426 memcpy(dst, "Id$", 3);
428 len -= dollar + 1 - src;
432 memcpy(dst, src, len);
433 strbuf_setlen(buf, dst + len - buf->buf);
437 static int ident_to_worktree(const char *path, const char *src, size_t len,
438 struct strbuf *buf, int ident)
440 unsigned char sha1[20];
441 char *to_free = NULL, *dollar;
447 cnt = count_ident(src, len);
451 /* are we "faking" in place editing ? */
453 to_free = strbuf_detach(buf, NULL);
454 hash_sha1_file(src, len, "blob", sha1);
456 strbuf_grow(buf, len + cnt * 43);
458 /* step 1: run to the next '$' */
459 dollar = memchr(src, '$', len);
462 strbuf_add(buf, src, dollar + 1 - src);
463 len -= dollar + 1 - src;
466 /* step 2: does it looks like a bit like Id:xxx$ or Id$ ? */
467 if (len < 3 || memcmp("Id", src, 2))
470 /* step 3: skip over Id$ or Id:xxxxx$ */
474 } else if (src[2] == ':') {
476 * It's possible that an expanded Id has crept its way into the
477 * repository, we cope with that by stripping the expansion out
479 dollar = memchr(src + 3, '$', len - 3);
481 /* incomplete keyword, no more '$', so just quit the loop */
485 len -= dollar + 1 - src;
488 /* it wasn't a "Id$" or "Id:xxxx$" */
492 /* step 4: substitute */
493 strbuf_addstr(buf, "Id: ");
494 strbuf_add(buf, sha1_to_hex(sha1), 40);
495 strbuf_addstr(buf, " $");
497 strbuf_add(buf, src, len);
503 static int git_path_check_crlf(const char *path, struct git_attr_check *check)
505 const char *value = check->value;
507 if (ATTR_TRUE(value))
509 else if (ATTR_FALSE(value))
511 else if (ATTR_UNSET(value))
513 else if (!strcmp(value, "input"))
518 static struct convert_driver *git_path_check_convert(const char *path,
519 struct git_attr_check *check)
521 const char *value = check->value;
522 struct convert_driver *drv;
524 if (ATTR_TRUE(value) || ATTR_FALSE(value) || ATTR_UNSET(value))
526 for (drv = user_convert; drv; drv = drv->next)
527 if (!strcmp(value, drv->name))
532 static int git_path_check_ident(const char *path, struct git_attr_check *check)
534 const char *value = check->value;
536 return !!ATTR_TRUE(value);
539 int convert_to_git(const char *path, const char *src, size_t len, struct strbuf *dst)
541 struct git_attr_check check[3];
542 int crlf = CRLF_GUESS;
543 int ident = 0, ret = 0;
546 setup_convert_check(check);
547 if (!git_checkattr(path, ARRAY_SIZE(check), check)) {
548 struct convert_driver *drv;
549 crlf = git_path_check_crlf(path, check + 0);
550 ident = git_path_check_ident(path, check + 1);
551 drv = git_path_check_convert(path, check + 2);
552 if (drv && drv->clean)
556 ret |= apply_filter(path, src, len, dst, filter);
561 ret |= crlf_to_git(path, src, len, dst, crlf);
566 return ret | ident_to_git(path, src, len, dst, ident);
569 int convert_to_working_tree(const char *path, const char *src, size_t len, struct strbuf *dst)
571 struct git_attr_check check[3];
572 int crlf = CRLF_GUESS;
573 int ident = 0, ret = 0;
576 setup_convert_check(check);
577 if (!git_checkattr(path, ARRAY_SIZE(check), check)) {
578 struct convert_driver *drv;
579 crlf = git_path_check_crlf(path, check + 0);
580 ident = git_path_check_ident(path, check + 1);
581 drv = git_path_check_convert(path, check + 2);
582 if (drv && drv->smudge)
583 filter = drv->smudge;
586 ret |= ident_to_worktree(path, src, len, dst, ident);
591 ret |= crlf_to_worktree(path, src, len, dst, crlf);
596 return ret | apply_filter(path, src, len, dst, filter);