pathspec.c 16.6 KB
Newer Older
1
/*
Edward Thomson committed
2
 * Copyright (C) the libgit2 contributors. All rights reserved.
3 4 5 6 7
 *
 * This file is part of libgit2, distributed under the GNU GPL v2 with
 * a Linking Exception. For full terms see the included COPYING file.
 */

8
#include "git2/pathspec.h"
9
#include "git2/diff.h"
10
#include "pathspec.h"
11
#include "buf_text.h"
12
#include "attr_file.h"
13 14 15
#include "iterator.h"
#include "repository.h"
#include "index.h"
16 17
#include "bitvec.h"
#include "diff.h"
18 19 20 21 22 23 24 25

/* what is the common non-wildcard prefix for all items in the pathspec */
char *git_pathspec_prefix(const git_strarray *pathspec)
{
	git_buf prefix = GIT_BUF_INIT;
	const char *scan;

	if (!pathspec || !pathspec->count ||
26
		git_buf_text_common_prefix(&prefix, pathspec) < 0)
27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
		return NULL;

	/* diff prefix will only be leading non-wildcards */
	for (scan = prefix.ptr; *scan; ++scan) {
		if (git__iswildcard(*scan) &&
			(scan == prefix.ptr || (*(scan - 1) != '\\')))
			break;
	}
	git_buf_truncate(&prefix, scan - prefix.ptr);

	if (prefix.size <= 0) {
		git_buf_free(&prefix);
		return NULL;
	}

42
	git_buf_text_unescape(&prefix);
43 44 45 46 47

	return git_buf_detach(&prefix);
}

/* is there anything in the spec that needs to be filtered on */
48
bool git_pathspec_is_empty(const git_strarray *pathspec)
49
{
50
	size_t i;
51

52
	if (pathspec == NULL)
53 54
		return true;

55 56 57 58 59 60 61
	for (i = 0; i < pathspec->count; ++i) {
		const char *str = pathspec->strings[i];

		if (str && str[0])
			return false;
	}

62 63 64 65
	return true;
}

/* build a vector of fnmatch patterns to evaluate efficiently */
66
int git_pathspec__vinit(
67 68 69 70 71 72
	git_vector *vspec, const git_strarray *strspec, git_pool *strpool)
{
	size_t i;

	memset(vspec, 0, sizeof(*vspec));

73
	if (git_pathspec_is_empty(strspec))
74 75 76 77 78 79 80 81 82 83 84 85
		return 0;

	if (git_vector_init(vspec, strspec->count, NULL) < 0)
		return -1;

	for (i = 0; i < strspec->count; ++i) {
		int ret;
		const char *pattern = strspec->strings[i];
		git_attr_fnmatch *match = git__calloc(1, sizeof(git_attr_fnmatch));
		if (!match)
			return -1;

86
		match->flags = GIT_ATTR_FNMATCH_ALLOWSPACE | GIT_ATTR_FNMATCH_ALLOWNEG;
87

88
		ret = git_attr_fnmatch__parse(match, strpool, NULL, &pattern);
89 90 91 92 93 94 95 96 97 98 99 100 101 102
		if (ret == GIT_ENOTFOUND) {
			git__free(match);
			continue;
		} else if (ret < 0)
			return ret;

		if (git_vector_insert(vspec, match) < 0)
			return -1;
	}

	return 0;
}

/* free data from the pathspec vector */
103
void git_pathspec__vfree(git_vector *vspec)
104 105 106 107 108 109 110 111 112 113 114 115
{
	git_attr_fnmatch *match;
	unsigned int i;

	git_vector_foreach(vspec, i, match) {
		git__free(match);
		vspec->contents[i] = NULL;
	}

	git_vector_free(vspec);
}

116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162
struct pathspec_match_context {
	int fnmatch_flags;
	int (*strcomp)(const char *, const char *);
	int (*strncomp)(const char *, const char *, size_t);
};

static void pathspec_match_context_init(
	struct pathspec_match_context *ctxt,
	bool disable_fnmatch,
	bool casefold)
{
	if (disable_fnmatch)
		ctxt->fnmatch_flags = -1;
	else if (casefold)
		ctxt->fnmatch_flags = FNM_CASEFOLD;
	else
		ctxt->fnmatch_flags = 0;

	if (casefold) {
		ctxt->strcomp  = git__strcasecmp;
		ctxt->strncomp = git__strncasecmp;
	} else {
		ctxt->strcomp  = git__strcmp;
		ctxt->strncomp = git__strncmp;
	}
}

static int pathspec_match_one(
	const git_attr_fnmatch *match,
	struct pathspec_match_context *ctxt,
	const char *path)
{
	int result = (match->flags & GIT_ATTR_FNMATCH_MATCH_ALL) ? 0 : FNM_NOMATCH;

	if (result == FNM_NOMATCH)
		result = ctxt->strcomp(match->pattern, path) ? FNM_NOMATCH : 0;

	if (ctxt->fnmatch_flags >= 0 && result == FNM_NOMATCH)
		result = p_fnmatch(match->pattern, path, ctxt->fnmatch_flags);

	/* if we didn't match, look for exact dirname prefix match */
	if (result == FNM_NOMATCH &&
		(match->flags & GIT_ATTR_FNMATCH_HASWILD) == 0 &&
		ctxt->strncomp(path, match->pattern, match->length) == 0 &&
		path[match->length] == '/')
		result = 0;

163 164 165 166 167 168
	/* if we didn't match and this is a negative match, check for exact
	 * match of filename with leading '!'
	 */
	if (result == FNM_NOMATCH &&
		(match->flags & GIT_ATTR_FNMATCH_NEGATIVE) != 0 &&
		*path == '!' &&
169 170
		ctxt->strncomp(path + 1, match->pattern, match->length) == 0 &&
		(!path[match->length + 1] || path[match->length + 1] == '/'))
171 172
		return 1;

173 174 175 176 177
	if (result == 0)
		return (match->flags & GIT_ATTR_FNMATCH_NEGATIVE) ? 0 : 1;
	return -1;
}

178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199
static int git_pathspec__match_at(
	size_t *matched_at,
	const git_vector *vspec,
	struct pathspec_match_context *ctxt,
	const char *path0,
	const char *path1)
{
	int result = GIT_ENOTFOUND;
	size_t i = 0;
	const git_attr_fnmatch *match;

	git_vector_foreach(vspec, i, match) {
		if (path0 && (result = pathspec_match_one(match, ctxt, path0)) >= 0)
			break;
		if (path1 && (result = pathspec_match_one(match, ctxt, path1)) >= 0)
			break;
	}

	*matched_at = i;
	return result;
}

200
/* match a path against the vectorized pathspec */
201 202
bool git_pathspec__match(
	const git_vector *vspec,
203 204 205
	const char *path,
	bool disable_fnmatch,
	bool casefold,
206 207
	const char **matched_pathspec,
	size_t *matched_at)
208
{
209 210
	int result;
	size_t pos;
211
	struct pathspec_match_context ctxt;
212

213 214
	if (matched_pathspec)
		*matched_pathspec = NULL;
215 216
	if (matched_at)
		*matched_at = GIT_PATHSPEC_NOMATCH;
217

218 219 220
	if (!vspec || !vspec->length)
		return true;

221
	pathspec_match_context_init(&ctxt, disable_fnmatch, casefold);
222

223 224 225 226 227
	result = git_pathspec__match_at(&pos, vspec, &ctxt, path, NULL);
	if (result >= 0) {
		if (matched_pathspec) {
			const git_attr_fnmatch *match = git_vector_get(vspec, pos);
			*matched_pathspec = match->pattern;
228
		}
229 230 231

		if (matched_at)
			*matched_at = pos;
232 233
	}

234
	return (result > 0);
235
}
nulltoken committed
236

237

238 239 240
int git_pathspec__init(git_pathspec *ps, const git_strarray *paths)
{
	int error = 0;
241

242
	memset(ps, 0, sizeof(*ps));
243

244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288
	ps->prefix = git_pathspec_prefix(paths);

	if ((error = git_pool_init(&ps->pool, 1, 0)) < 0 ||
		(error = git_pathspec__vinit(&ps->pathspec, paths, &ps->pool)) < 0)
		git_pathspec__clear(ps);

	return error;
}

void git_pathspec__clear(git_pathspec *ps)
{
	git__free(ps->prefix);
	git_pathspec__vfree(&ps->pathspec);
	git_pool_clear(&ps->pool);
	memset(ps, 0, sizeof(*ps));
}

int git_pathspec_new(git_pathspec **out, const git_strarray *pathspec)
{
	int error = 0;
	git_pathspec *ps = git__malloc(sizeof(git_pathspec));
	GITERR_CHECK_ALLOC(ps);

	if ((error = git_pathspec__init(ps, pathspec)) < 0) {
		git__free(ps);
		return error;
	}

	GIT_REFCOUNT_INC(ps);
	*out = ps;
	return 0;
}

static void pathspec_free(git_pathspec *ps)
{
	git_pathspec__clear(ps);
	git__free(ps);
}

void git_pathspec_free(git_pathspec *ps)
{
	if (!ps)
		return;
	GIT_REFCOUNT_DEC(ps, pathspec_free);
}
289

290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312
int git_pathspec_matches_path(
	const git_pathspec *ps, uint32_t flags, const char *path)
{
	bool no_fnmatch = (flags & GIT_PATHSPEC_NO_GLOB) != 0;
	bool casefold =  (flags & GIT_PATHSPEC_IGNORE_CASE) != 0;

	assert(ps && path);

	return (0 != git_pathspec__match(
		&ps->pathspec, path, no_fnmatch, casefold, NULL, NULL));
}

static void pathspec_match_free(git_pathspec_match_list *m)
{
	git_pathspec_free(m->pathspec);
	m->pathspec = NULL;

	git_array_clear(m->matches);
	git_array_clear(m->failures);
	git_pool_clear(&m->pool);
	git__free(m);
}

313 314
static git_pathspec_match_list *pathspec_match_alloc(
	git_pathspec *ps, int datatype)
315 316 317 318 319 320 321 322 323 324 325 326 327 328
{
	git_pathspec_match_list *m = git__calloc(1, sizeof(git_pathspec_match_list));

	if (m != NULL && git_pool_init(&m->pool, 1, 0) < 0) {
		pathspec_match_free(m);
		m = NULL;
	}

	/* need to keep reference to pathspec and increment refcount because
	 * failures array stores pointers to the pattern strings of the
	 * pathspec that had no matches
	 */
	GIT_REFCOUNT_INC(ps);
	m->pathspec = ps;
329
	m->datatype = datatype;
330 331 332 333

	return m;
}

334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376
GIT_INLINE(size_t) pathspec_mark_pattern(git_bitvec *used, size_t pos)
{
	if (!git_bitvec_get(used, pos)) {
		git_bitvec_set(used, pos, true);
		return 1;
	}

	return 0;
}

static size_t pathspec_mark_remaining(
	git_bitvec *used,
	git_vector *patterns,
	struct pathspec_match_context *ctxt,
	size_t start,
	const char *path0,
	const char *path1)
{
	size_t count = 0;

	if (path1 == path0)
		path1 = NULL;

	for (; start < patterns->length; ++start) {
		const git_attr_fnmatch *pat = git_vector_get(patterns, start);

		if (git_bitvec_get(used, start))
			continue;

		if (path0 && pathspec_match_one(pat, ctxt, path0) > 0)
			count += pathspec_mark_pattern(used, start);
		else if (path1 && pathspec_match_one(pat, ctxt, path1) > 0)
			count += pathspec_mark_pattern(used, start);
	}

	return count;
}

static int pathspec_build_failure_array(
	git_pathspec_string_array_t *failures,
	git_vector *patterns,
	git_bitvec *used,
	git_pool *pool)
377
{
378 379 380 381 382 383 384 385 386 387 388 389 390 391 392
	size_t pos;
	char **failed;
	const git_attr_fnmatch *pat;

	for (pos = 0; pos < patterns->length; ++pos) {
		if (git_bitvec_get(used, pos))
			continue;

		if ((failed = git_array_alloc(*failures)) == NULL)
			return -1;

		pat = git_vector_get(patterns, pos);

		if ((*failed = git_pool_strdup(pool, pat->pattern)) == NULL)
			return -1;
393
	}
394 395

	return 0;
396 397 398 399 400 401 402 403 404
}

static int pathspec_match_from_iterator(
	git_pathspec_match_list **out,
	git_iterator *iter,
	uint32_t flags,
	git_pathspec *ps)
{
	int error = 0;
405
	git_pathspec_match_list *m = NULL;
406 407 408
	const git_index_entry *entry = NULL;
	struct pathspec_match_context ctxt;
	git_vector *patterns = &ps->pathspec;
409 410
	bool find_failures = out && (flags & GIT_PATHSPEC_FIND_FAILURES) != 0;
	bool failures_only = !out || (flags & GIT_PATHSPEC_FAILURES_ONLY) != 0;
411 412
	size_t pos, used_ct = 0, found_files = 0;
	git_index *index = NULL;
413
	git_bitvec used_patterns;
414 415
	char **file;

416 417 418
	if (git_bitvec_init(&used_patterns, patterns->length) < 0)
		return -1;

419
	if (out) {
420
		*out = m = pathspec_match_alloc(ps, PATHSPEC_DATATYPE_STRINGS);
421 422
		GITERR_CHECK_ALLOC(m);
	}
423 424 425 426 427 428 429 430 431

	if ((error = git_iterator_reset(iter, ps->prefix, ps->prefix)) < 0)
		goto done;

	if (git_iterator_type(iter) == GIT_ITERATOR_TYPE_WORKDIR &&
		(error = git_repository_index__weakptr(
			&index, git_iterator_owner(iter))) < 0)
		goto done;

432 433 434
	pathspec_match_context_init(
		&ctxt, (flags & GIT_PATHSPEC_NO_GLOB) != 0,
		git_iterator_ignore_case(iter));
435 436

	while (!(error = git_iterator_advance(&entry, iter))) {
437 438 439
		/* search for match with entry->path */
		int result = git_pathspec__match_at(
			&pos, patterns, &ctxt, entry->path, NULL);
440 441 442 443 444 445 446

		/* no matches for this path */
		if (result < 0)
			continue;

		/* if result was a negative pattern match, then don't list file */
		if (!result) {
447
			used_ct += pathspec_mark_pattern(&used_patterns, pos);
448 449 450
			continue;
		}

451
		/* check if path is ignored and untracked */
452 453 454 455 456 457
		if (index != NULL &&
			git_iterator_current_is_ignored(iter) &&
			git_index__find(NULL, index, entry->path, GIT_INDEX_STAGE_ANY) < 0)
			continue;

		/* mark the matched pattern as used */
458
		used_ct += pathspec_mark_pattern(&used_patterns, pos);
459 460 461
		++found_files;

		/* if find_failures is on, check if any later patterns also match */
462 463 464
		if (find_failures && used_ct < patterns->length)
			used_ct += pathspec_mark_remaining(
				&used_patterns, patterns, &ctxt, pos + 1, entry->path, NULL);
465 466

		/* if only looking at failures, exit early or just continue */
467
		if (failures_only || !out) {
468 469 470 471 472 473
			if (used_ct == patterns->length)
				break;
			continue;
		}

		/* insert matched path into matches array */
474
		if ((file = (char **)git_array_alloc(m->matches)) == NULL ||
475 476 477
			(*file = git_pool_strdup(&m->pool, entry->path)) == NULL) {
			error = -1;
			goto done;
478
		}
479 480
	}

481 482 483 484 485
	if (error < 0 && error != GIT_ITEROVER)
		goto done;
	error = 0;

	/* insert patterns that had no matches into failures array */
486 487 488 489
	if (find_failures && used_ct < patterns->length &&
		(error = pathspec_build_failure_array(
			&m->failures, patterns, &used_patterns, &m->pool)) < 0)
		goto done;
490 491 492 493 494 495 496 497

	/* if every pattern failed to match, then we have failed */
	if ((flags & GIT_PATHSPEC_NO_MATCH_ERROR) != 0 && !found_files) {
		giterr_set(GITERR_INVALID, "No matching files were found");
		error = GIT_ENOTFOUND;
	}

done:
498
	git_bitvec_free(&used_patterns);
499 500 501

	if (error < 0) {
		pathspec_match_free(m);
502
		if (out) *out = NULL;
503 504 505
	}

	return error;
506 507
}

508 509 510 511 512 513 514 515
static git_iterator_flag_t pathspec_match_iter_flags(uint32_t flags)
{
	git_iterator_flag_t f = 0;

	if ((flags & GIT_PATHSPEC_IGNORE_CASE) != 0)
		f |= GIT_ITERATOR_IGNORE_CASE;
	else if ((flags & GIT_PATHSPEC_USE_CASE) != 0)
		f |= GIT_ITERATOR_DONT_IGNORE_CASE;
516

517 518 519 520 521 522 523 524
	return f;
}

int git_pathspec_match_workdir(
	git_pathspec_match_list **out,
	git_repository *repo,
	uint32_t flags,
	git_pathspec *ps)
525 526
{
	int error = 0;
527 528
	git_iterator *iter;

529
	assert(repo);
530

531 532
	if (!(error = git_iterator_for_workdir(
			&iter, repo, pathspec_match_iter_flags(flags), NULL, NULL))) {
533

534
		error = pathspec_match_from_iterator(out, iter, flags, ps);
535

536 537
		git_iterator_free(iter);
	}
538 539 540 541

	return error;
}

542 543 544 545 546 547 548 549 550
int git_pathspec_match_index(
	git_pathspec_match_list **out,
	git_index *index,
	uint32_t flags,
	git_pathspec *ps)
{
	int error = 0;
	git_iterator *iter;

551
	assert(index);
552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572

	if (!(error = git_iterator_for_index(
			&iter, index, pathspec_match_iter_flags(flags), NULL, NULL))) {

		error = pathspec_match_from_iterator(out, iter, flags, ps);

		git_iterator_free(iter);
	}

	return error;
}

int git_pathspec_match_tree(
	git_pathspec_match_list **out,
	git_tree *tree,
	uint32_t flags,
	git_pathspec *ps)
{
	int error = 0;
	git_iterator *iter;

573
	assert(tree);
574 575 576 577 578 579 580 581 582 583 584 585

	if (!(error = git_iterator_for_tree(
			&iter, tree, pathspec_match_iter_flags(flags), NULL, NULL))) {

		error = pathspec_match_from_iterator(out, iter, flags, ps);

		git_iterator_free(iter);
	}

	return error;
}

586 587
int git_pathspec_match_diff(
	git_pathspec_match_list **out,
588
	git_diff *diff,
589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678
	uint32_t flags,
	git_pathspec *ps)
{
	int error = 0;
	git_pathspec_match_list *m = NULL;
	struct pathspec_match_context ctxt;
	git_vector *patterns = &ps->pathspec;
	bool find_failures = out && (flags & GIT_PATHSPEC_FIND_FAILURES) != 0;
	bool failures_only = !out || (flags & GIT_PATHSPEC_FAILURES_ONLY) != 0;
	size_t i, pos, used_ct = 0, found_deltas = 0;
	const git_diff_delta *delta, **match;
	git_bitvec used_patterns;

	assert(diff);

	if (git_bitvec_init(&used_patterns, patterns->length) < 0)
		return -1;

	if (out) {
		*out = m = pathspec_match_alloc(ps, PATHSPEC_DATATYPE_DIFF);
		GITERR_CHECK_ALLOC(m);
	}

	pathspec_match_context_init(
		&ctxt, (flags & GIT_PATHSPEC_NO_GLOB) != 0,
		git_diff_is_sorted_icase(diff));

	git_vector_foreach(&diff->deltas, i, delta) {
		/* search for match with delta */
		int result = git_pathspec__match_at(
			&pos, patterns, &ctxt, delta->old_file.path, delta->new_file.path);

		/* no matches for this path */
		if (result < 0)
			continue;

		/* mark the matched pattern as used */
		used_ct += pathspec_mark_pattern(&used_patterns, pos);

		/* if result was a negative pattern match, then don't list file */
		if (!result)
			continue;

		++found_deltas;

		/* if find_failures is on, check if any later patterns also match */
		if (find_failures && used_ct < patterns->length)
			used_ct += pathspec_mark_remaining(
				&used_patterns, patterns, &ctxt, pos + 1,
				delta->old_file.path, delta->new_file.path);

		/* if only looking at failures, exit early or just continue */
		if (failures_only || !out) {
			if (used_ct == patterns->length)
				break;
			continue;
		}

		/* insert matched delta into matches array */
		if (!(match = (const git_diff_delta **)git_array_alloc(m->matches))) {
			error = -1;
			goto done;
		} else {
			*match = delta;
		}
	}

	/* insert patterns that had no matches into failures array */
	if (find_failures && used_ct < patterns->length &&
		(error = pathspec_build_failure_array(
			&m->failures, patterns, &used_patterns, &m->pool)) < 0)
		goto done;

	/* if every pattern failed to match, then we have failed */
	if ((flags & GIT_PATHSPEC_NO_MATCH_ERROR) != 0 && !found_deltas) {
		giterr_set(GITERR_INVALID, "No matching deltas were found");
		error = GIT_ENOTFOUND;
	}

done:
	git_bitvec_free(&used_patterns);

	if (error < 0) {
		pathspec_match_free(m);
		if (out) *out = NULL;
	}

	return error;
}

679 680
void git_pathspec_match_list_free(git_pathspec_match_list *m)
{
681 682
	if (m)
		pathspec_match_free(m);
683 684 685 686 687
}

size_t git_pathspec_match_list_entrycount(
	const git_pathspec_match_list *m)
{
688
	return m ? git_array_size(m->matches) : 0;
689 690 691 692 693
}

const char *git_pathspec_match_list_entry(
	const git_pathspec_match_list *m, size_t pos)
{
694 695 696 697 698 699 700 701 702 703 704 705 706 707 708
	if (!m || m->datatype != PATHSPEC_DATATYPE_STRINGS ||
		!git_array_valid_index(m->matches, pos))
		return NULL;

	return *((const char **)git_array_get(m->matches, pos));
}

const git_diff_delta *git_pathspec_match_list_diff_entry(
	const git_pathspec_match_list *m, size_t pos)
{
	if (!m || m->datatype != PATHSPEC_DATATYPE_DIFF ||
		!git_array_valid_index(m->matches, pos))
		return NULL;

	return *((const git_diff_delta **)git_array_get(m->matches, pos));
709 710 711 712 713
}

size_t git_pathspec_match_list_failed_entrycount(
	const git_pathspec_match_list *m)
{
714
	return m ? git_array_size(m->failures) : 0;
715 716 717 718
}

const char * git_pathspec_match_list_failed_entry(
	const git_pathspec_match_list *m, size_t pos)
719
{
720 721
	char **entry = m ? git_array_get(m->failures, pos) : NULL;

722
	return entry ? *entry : NULL;
723
}
724