convert.c

   1 #include "cache.h"
   2 #include "attr.h"
   3 #include "run-command.h"
   4
   5 /*
   6  * convert.c - convert a file when checking it out and checking it in.
   7  *
   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.
  12  */
  13
  14 #define CRLF_GUESS      (-1)
  15 #define CRLF_BINARY     0
  16 #define CRLF_TEXT       1
  17 #define CRLF_INPUT      2
  18
  19 struct text_stat {
  20         /* NUL, CR, LF and CRLF counts */
  21         unsigned nul, cr, lf, crlf;
  22
  23         /* These are just approximations! */
  24         unsigned printable, nonprintable;
  25 };
  26
  27 static void gather_stats(const char *buf, unsigned long size, struct text_stat *stats)
  28 {
  29         unsigned long i;
  30
  31         memset(stats, 0, sizeof(*stats));
  32
  33         for (i = 0; i < size; i++) {
  34                 unsigned char c = buf[i];
  35                 if (c == '\r') {
  36                         stats->cr++;
  37                         if (i+1 < size && buf[i+1] == '\n')
  38                                 stats->crlf++;
  39                         continue;
  40                 }
  41                 if (c == '\n') {
  42                         stats->lf++;
  43                         continue;
  44                 }
  45                 if (c == 127)
  46                         /* DEL */
  47                         stats->nonprintable++;
  48                 else if (c < 32) {
  49                         switch (c) {
  50                                 /* BS, HT, ESC and FF */
  51                         case '\b': case '\t': case '\033': case '\014':
  52                                 stats->printable++;
  53                                 break;
  54                         case 0:
  55                                 stats->nul++;
  56                                 /* fall through */
  57                         default:
  58                                 stats->nonprintable++;
  59                         }
  60                 }
  61                 else
  62                         stats->printable++;
  63         }
  64 }
  65
  66 /*
  67  * The same heuristics as diff.c::mmfile_is_binary()
  68  */
  69 static int is_binary(unsigned long size, struct text_stat *stats)
  70 {
  71
  72         if (stats->nul)
  73                 return 1;
  74         if ((stats->printable >> 7) < stats->nonprintable)
  75                 return 1;
  76         /*
  77          * Other heuristics? Average line length might be relevant,
  78          * as might LF vs CR vs CRLF counts..
  79          *
  80          * NOTE! It might be normal to have a low ratio of CRLF to LF
  81          * (somebody starts with a LF-only file and edits it with an editor
  82          * that adds CRLF only to lines that are added..). But do  we
  83          * want to support CR-only? Probably not.
  84          */
  85         return 0;
  86 }
  87
  88 static void check_safe_crlf(const char *path, int action,
  89                             struct text_stat *stats, enum safe_crlf checksafe)
  90 {
  91         if (!checksafe)
  92                 return;
  93
  94         if (action == CRLF_INPUT || auto_crlf <= 0) {
  95                 /*
  96                  * CRLFs would not be restored by checkout:
  97                  * check if we'd remove CRLFs
  98                  */
  99                 if (stats->crlf) {
 100                         if (checksafe == SAFE_CRLF_WARN)
 101                                 warning("CRLF will be replaced by LF in %s.", path);
 102                         else /* i.e. SAFE_CRLF_FAIL */
 103                                 die("CRLF would be replaced by LF in %s.", path);
 104                 }
 105         } else if (auto_crlf > 0) {
 106                 /*
 107                  * CRLFs would be added by checkout:
 108                  * check if we have "naked" LFs
 109                  */
 110                 if (stats->lf != stats->crlf) {
 111                         if (checksafe == SAFE_CRLF_WARN)
 112                                 warning("LF will be replaced by CRLF in %s", path);
 113                         else /* i.e. SAFE_CRLF_FAIL */
 114                                 die("LF would be replaced by CRLF in %s", path);
 115                 }
 116         }
 117 }
 118
 119 static int crlf_to_git(const char *path, const char *src, size_t len,
 120                        struct strbuf *buf, int action, enum safe_crlf checksafe)
 121 {
 122         struct text_stat stats;
 123         char *dst;
 124
 125         if ((action == CRLF_BINARY) || !auto_crlf || !len)
 126                 return 0;
 127
 128         gather_stats(src, len, &stats);
 129
 130         if (action == CRLF_GUESS) {
 131                 /*
 132                  * We're currently not going to even try to convert stuff
 133                  * that has bare CR characters. Does anybody do that crazy
 134                  * stuff?
 135                  */
 136                 if (stats.cr != stats.crlf)
 137                         return 0;
 138
 139                 /*
 140                  * And add some heuristics for binary vs text, of course...
 141                  */
 142                 if (is_binary(len, &stats))
 143                         return 0;
 144         }
 145
 146         check_safe_crlf(path, action, &stats, checksafe);
 147
 148         /* Optimization: No CR? Nothing to convert, regardless. */
 149         if (!stats.cr)
 150                 return 0;
 151
 152         /* only grow if not in place */
 153         if (strbuf_avail(buf) + buf->len < len)
 154                 strbuf_grow(buf, len - buf->len);
 155         dst = buf->buf;
 156         if (action == CRLF_GUESS) {
 157                 /*
 158                  * If we guessed, we already know we rejected a file with
 159                  * lone CR, and we can strip a CR without looking at what
 160                  * follow it.
 161                  */
 162                 do {
 163                         unsigned char c = *src++;
 164                         if (c != '\r')
 165                                 *dst++ = c;
 166                 } while (--len);
 167         } else {
 168                 do {
 169                         unsigned char c = *src++;
 170                         if (! (c == '\r' && (1 < len && *src == '\n')))
 171                                 *dst++ = c;
 172                 } while (--len);
 173         }
 174         strbuf_setlen(buf, dst - buf->buf);
 175         return 1;
 176 }
 177
 178 static int crlf_to_worktree(const char *path, const char *src, size_t len,
 179                             struct strbuf *buf, int action)
 180 {
 181         char *to_free = NULL;
 182         struct text_stat stats;
 183
 184         if ((action == CRLF_BINARY) || (action == CRLF_INPUT) ||
 185             auto_crlf <= 0)
 186                 return 0;
 187
 188         if (!len)
 189                 return 0;
 190
 191         gather_stats(src, len, &stats);
 192
 193         /* No LF? Nothing to convert, regardless. */
 194         if (!stats.lf)
 195                 return 0;
 196
 197         /* Was it already in CRLF format? */
 198         if (stats.lf == stats.crlf)
 199                 return 0;
 200
 201         if (action == CRLF_GUESS) {
 202                 /* If we have any bare CR characters, we're not going to touch it */
 203                 if (stats.cr != stats.crlf)
 204                         return 0;
 205
 206                 if (is_binary(len, &stats))
 207                         return 0;
 208         }
 209
 210         /* are we "faking" in place editing ? */
 211         if (src == buf->buf)
 212                 to_free = strbuf_detach(buf, NULL);
 213
 214         strbuf_grow(buf, len + stats.lf - stats.crlf);
 215         for (;;) {
 216                 const char *nl = memchr(src, '\n', len);
 217                 if (!nl)
 218                         break;
 219                 if (nl > src && nl[-1] == '\r') {
 220                         strbuf_add(buf, src, nl + 1 - src);
 221                 } else {
 222                         strbuf_add(buf, src, nl - src);
 223                         strbuf_addstr(buf, "\r\n");
 224                 }
 225                 len -= nl + 1 - src;
 226                 src  = nl + 1;
 227         }
 228         strbuf_add(buf, src, len);
 229
 230         free(to_free);
 231         return 1;
 232 }
 233
 234 struct filter_params {
 235         const char *src;
 236         unsigned long size;
 237         const char *cmd;
 238 };
 239
 240 static int filter_buffer(int fd, void *data)
 241 {
 242         /*
 243          * Spawn cmd and feed the buffer contents through its stdin.
 244          */
 245         struct child_process child_process;
 246         struct filter_params *params = (struct filter_params *)data;
 247         int write_err, status;
 248         const char *argv[] = { "sh", "-c", params->cmd, NULL };
 249
 250         memset(&child_process, 0, sizeof(child_process));
 251         child_process.argv = argv;
 252         child_process.in = -1;
 253         child_process.out = fd;
 254
 255         if (start_command(&child_process))
 256                 return error("cannot fork to run external filter %s", params->cmd);
 257
 258         write_err = (write_in_full(child_process.in, params->src, params->size) < 0);
 259         if (close(child_process.in))
 260                 write_err = 1;
 261         if (write_err)
 262                 error("cannot feed the input to external filter %s", params->cmd);
 263
 264         status = finish_command(&child_process);
 265         if (status)
 266                 error("external filter %s failed %d", params->cmd, -status);
 267         return (write_err || status);
 268 }
 269
 270 static int apply_filter(const char *path, const char *src, size_t len,
 271                         struct strbuf *dst, const char *cmd)
 272 {
 273         /*
 274          * Create a pipeline to have the command filter the buffer's
 275          * contents.
 276          *
 277          * (child --> cmd) --> us
 278          */
 279         int ret = 1;
 280         struct strbuf nbuf;
 281         struct async async;
 282         struct filter_params params;
 283
 284         if (!cmd)
 285                 return 0;
 286
 287         memset(&async, 0, sizeof(async));
 288         async.proc = filter_buffer;
 289         async.data = &params;
 290         params.src = src;
 291         params.size = len;
 292         params.cmd = cmd;
 293
 294         fflush(NULL);
 295         if (start_async(&async))
 296                 return 0;       /* error was already reported */
 297
 298         strbuf_init(&nbuf, 0);
 299         if (strbuf_read(&nbuf, async.out, len) < 0) {
 300                 error("read from external filter %s failed", cmd);
 301                 ret = 0;
 302         }
 303         if (close(async.out)) {
 304                 error("read from external filter %s failed", cmd);
 305                 ret = 0;
 306         }
 307         if (finish_async(&async)) {
 308                 error("external filter %s failed", cmd);
 309                 ret = 0;
 310         }
 311
 312         if (ret) {
 313                 strbuf_swap(dst, &nbuf);
 314         }
 315         strbuf_release(&nbuf);
 316         return ret;
 317 }
 318
 319 static struct convert_driver {
 320         const char *name;
 321         struct convert_driver *next;
 322         const char *smudge;
 323         const char *clean;
 324 } *user_convert, **user_convert_tail;
 325
 326 static int read_convert_config(const char *var, const char *value, void *cb)
 327 {
 328         const char *ep, *name;
 329         int namelen;
 330         struct convert_driver *drv;
 331
 332         /*
 333          * External conversion drivers are configured using
 334          * "filter.<name>.variable".
 335          */
 336         if (prefixcmp(var, "filter.") || (ep = strrchr(var, '.')) == var + 6)
 337                 return 0;
 338         name = var + 7;
 339         namelen = ep - name;
 340         for (drv = user_convert; drv; drv = drv->next)
 341                 if (!strncmp(drv->name, name, namelen) && !drv->name[namelen])
 342                         break;
 343         if (!drv) {
 344                 drv = xcalloc(1, sizeof(struct convert_driver));
 345                 drv->name = xmemdupz(name, namelen);
 346                 *user_convert_tail = drv;
 347                 user_convert_tail = &(drv->next);
 348         }
 349
 350         ep++;
 351
 352         /*
 353          * filter.<name>.smudge and filter.<name>.clean specifies
 354          * the command line:
 355          *
 356          *      command-line
 357          *
 358          * The command-line will not be interpolated in any way.
 359          */
 360
 361         if (!strcmp("smudge", ep))
 362                 return git_config_string(&drv->smudge, var, value);
 363
 364         if (!strcmp("clean", ep))
 365                 return git_config_string(&drv->clean, var, value);
 366
 367         return 0;
 368 }
 369
 370 static void setup_convert_check(struct git_attr_check *check)
 371 {
 372         static struct git_attr *attr_crlf;
 373         static struct git_attr *attr_ident;
 374         static struct git_attr *attr_filter;
 375
 376         if (!attr_crlf) {
 377                 attr_crlf = git_attr("crlf", 4);
 378                 attr_ident = git_attr("ident", 5);
 379                 attr_filter = git_attr("filter", 6);
 380                 user_convert_tail = &user_convert;
 381                 git_config(read_convert_config, NULL);
 382         }
 383         check[0].attr = attr_crlf;
 384         check[1].attr = attr_ident;
 385         check[2].attr = attr_filter;
 386 }
 387
 388 static int count_ident(const char *cp, unsigned long size)
 389 {
 390         /*
 391          * "$Id: 0000000000000000000000000000000000000000 $" <=> "$Id$"
 392          */
 393         int cnt = 0;
 394         char ch;
 395
 396         while (size) {
 397                 ch = *cp++;
 398                 size--;
 399                 if (ch != '$')
 400                         continue;
 401                 if (size < 3)
 402                         break;
 403                 if (memcmp("Id", cp, 2))
 404                         continue;
 405                 ch = cp[2];
 406                 cp += 3;
 407                 size -= 3;
 408                 if (ch == '$')
 409                         cnt++; /* $Id$ */
 410                 if (ch != ':')
 411                         continue;
 412
 413                 /*
 414                  * "$Id: ... "; scan up to the closing dollar sign and discard.
 415                  */
 416                 while (size) {
 417                         ch = *cp++;
 418                         size--;
 419                         if (ch == '$') {
 420                                 cnt++;
 421                                 break;
 422                         }
 423                 }
 424         }
 425         return cnt;
 426 }
 427
 428 static int ident_to_git(const char *path, const char *src, size_t len,
 429                         struct strbuf *buf, int ident)
 430 {
 431         char *dst, *dollar;
 432
 433         if (!ident || !count_ident(src, len))
 434                 return 0;
 435
 436         /* only grow if not in place */
 437         if (strbuf_avail(buf) + buf->len < len)
 438                 strbuf_grow(buf, len - buf->len);
 439         dst = buf->buf;
 440         for (;;) {
 441                 dollar = memchr(src, '$', len);
 442                 if (!dollar)
 443                         break;
 444                 memcpy(dst, src, dollar + 1 - src);
 445                 dst += dollar + 1 - src;
 446                 len -= dollar + 1 - src;
 447                 src  = dollar + 1;
 448
 449                 if (len > 3 && !memcmp(src, "Id:", 3)) {
 450                         dollar = memchr(src + 3, '$', len - 3);
 451                         if (!dollar)
 452                                 break;
 453                         memcpy(dst, "Id$", 3);
 454                         dst += 3;
 455                         len -= dollar + 1 - src;
 456                         src  = dollar + 1;
 457                 }
 458         }
 459         memcpy(dst, src, len);
 460         strbuf_setlen(buf, dst + len - buf->buf);
 461         return 1;
 462 }
 463
 464 static int ident_to_worktree(const char *path, const char *src, size_t len,
 465                              struct strbuf *buf, int ident)
 466 {
 467         unsigned char sha1[20];
 468         char *to_free = NULL, *dollar;
 469         int cnt;
 470
 471         if (!ident)
 472                 return 0;
 473
 474         cnt = count_ident(src, len);
 475         if (!cnt)
 476                 return 0;
 477
 478         /* are we "faking" in place editing ? */
 479         if (src == buf->buf)
 480                 to_free = strbuf_detach(buf, NULL);
 481         hash_sha1_file(src, len, "blob", sha1);
 482
 483         strbuf_grow(buf, len + cnt * 43);
 484         for (;;) {
 485                 /* step 1: run to the next '$' */
 486                 dollar = memchr(src, '$', len);
 487                 if (!dollar)
 488                         break;
 489                 strbuf_add(buf, src, dollar + 1 - src);
 490                 len -= dollar + 1 - src;
 491                 src  = dollar + 1;
 492
 493                 /* step 2: does it looks like a bit like Id:xxx$ or Id$ ? */
 494                 if (len < 3 || memcmp("Id", src, 2))
 495                         continue;
 496
 497                 /* step 3: skip over Id$ or Id:xxxxx$ */
 498                 if (src[2] == '$') {
 499                         src += 3;
 500                         len -= 3;
 501                 } else if (src[2] == ':') {
 502                         /*
 503                          * It's possible that an expanded Id has crept its way into the
 504                          * repository, we cope with that by stripping the expansion out
 505                          */
 506                         dollar = memchr(src + 3, '$', len - 3);
 507                         if (!dollar) {
 508                                 /* incomplete keyword, no more '$', so just quit the loop */
 509                                 break;
 510                         }
 511
 512                         len -= dollar + 1 - src;
 513                         src  = dollar + 1;
 514                 } else {
 515                         /* it wasn't a "Id$" or "Id:xxxx$" */
 516                         continue;
 517                 }
 518
 519                 /* step 4: substitute */
 520                 strbuf_addstr(buf, "Id: ");
 521                 strbuf_add(buf, sha1_to_hex(sha1), 40);
 522                 strbuf_addstr(buf, " $");
 523         }
 524         strbuf_add(buf, src, len);
 525
 526         free(to_free);
 527         return 1;
 528 }
 529
 530 static int git_path_check_crlf(const char *path, struct git_attr_check *check)
 531 {
 532         const char *value = check->value;
 533
 534         if (ATTR_TRUE(value))
 535                 return CRLF_TEXT;
 536         else if (ATTR_FALSE(value))
 537                 return CRLF_BINARY;
 538         else if (ATTR_UNSET(value))
 539                 ;
 540         else if (!strcmp(value, "input"))
 541                 return CRLF_INPUT;
 542         return CRLF_GUESS;
 543 }
 544
 545 static struct convert_driver *git_path_check_convert(const char *path,
 546                                              struct git_attr_check *check)
 547 {
 548         const char *value = check->value;
 549         struct convert_driver *drv;
 550
 551         if (ATTR_TRUE(value) || ATTR_FALSE(value) || ATTR_UNSET(value))
 552                 return NULL;
 553         for (drv = user_convert; drv; drv = drv->next)
 554                 if (!strcmp(value, drv->name))
 555                         return drv;
 556         return NULL;
 557 }
 558
 559 static int git_path_check_ident(const char *path, struct git_attr_check *check)
 560 {
 561         const char *value = check->value;
 562
 563         return !!ATTR_TRUE(value);
 564 }
 565
 566 int convert_to_git(const char *path, const char *src, size_t len,
 567                    struct strbuf *dst, enum safe_crlf checksafe)
 568 {
 569         struct git_attr_check check[3];
 570         int crlf = CRLF_GUESS;
 571         int ident = 0, ret = 0;
 572         const char *filter = NULL;
 573
 574         setup_convert_check(check);
 575         if (!git_checkattr(path, ARRAY_SIZE(check), check)) {
 576                 struct convert_driver *drv;
 577                 crlf = git_path_check_crlf(path, check + 0);
 578                 ident = git_path_check_ident(path, check + 1);
 579                 drv = git_path_check_convert(path, check + 2);
 580                 if (drv && drv->clean)
 581                         filter = drv->clean;
 582         }
 583
 584         ret |= apply_filter(path, src, len, dst, filter);
 585         if (ret) {
 586                 src = dst->buf;
 587                 len = dst->len;
 588         }
 589         ret |= crlf_to_git(path, src, len, dst, crlf, checksafe);
 590         if (ret) {
 591                 src = dst->buf;
 592                 len = dst->len;
 593         }
 594         return ret | ident_to_git(path, src, len, dst, ident);
 595 }
 596
 597 int convert_to_working_tree(const char *path, const char *src, size_t len, struct strbuf *dst)
 598 {
 599         struct git_attr_check check[3];
 600         int crlf = CRLF_GUESS;
 601         int ident = 0, ret = 0;
 602         const char *filter = NULL;
 603
 604         setup_convert_check(check);
 605         if (!git_checkattr(path, ARRAY_SIZE(check), check)) {
 606                 struct convert_driver *drv;
 607                 crlf = git_path_check_crlf(path, check + 0);
 608                 ident = git_path_check_ident(path, check + 1);
 609                 drv = git_path_check_convert(path, check + 2);
 610                 if (drv && drv->smudge)
 611                         filter = drv->smudge;
 612         }
 613
 614         ret |= ident_to_worktree(path, src, len, dst, ident);
 615         if (ret) {
 616                 src = dst->buf;
 617                 len = dst->len;
 618         }
 619         ret |= crlf_to_worktree(path, src, len, dst, crlf);
 620         if (ret) {
 621                 src = dst->buf;
 622                 len = dst->len;
 623         }
 624         return ret | apply_filter(path, src, len, dst, filter);
 625 }