Mercurial > public > mercurial-scm > hg-stable
comparison mercurial/utils/stringutil.py @ 45724:ac39a8a214b1
stringutil: add function to compile stringmatcher pattern into regexp
Prepares for adding a revset predicate for "grep --diff". The grep logic
needs a regexp object instead of a match function.
| author | Yuya Nishihara <yuya@tcha.org> |
|---|---|
| date | Mon, 05 Oct 2020 20:40:39 +0900 |
| parents | edfc5820aae7 |
| children | 89a2afe31e82 |
comparison
equal
deleted
inserted
replaced
| 45723:edfc5820aae7 | 45724:ac39a8a214b1 |
|---|---|
| 374 return kind, pattern, match | 374 return kind, pattern, match |
| 375 | 375 |
| 376 raise error.ProgrammingError(b'unhandled pattern kind: %s' % kind) | 376 raise error.ProgrammingError(b'unhandled pattern kind: %s' % kind) |
| 377 | 377 |
| 378 | 378 |
| 379 def substringregexp(pattern, flags=0): | |
| 380 """Build a regexp object from a string pattern possibly starting with | |
| 381 're:' or 'literal:' prefix. | |
| 382 | |
| 383 helper for tests: | |
| 384 >>> def test(pattern, *tests): | |
| 385 ... regexp = substringregexp(pattern) | |
| 386 ... return [bool(regexp.search(t)) for t in tests] | |
| 387 >>> def itest(pattern, *tests): | |
| 388 ... regexp = substringregexp(pattern, remod.I) | |
| 389 ... return [bool(regexp.search(t)) for t in tests] | |
| 390 | |
| 391 substring matching (no prefix): | |
| 392 >>> test(b'bcde', b'abc', b'def', b'abcdefg') | |
| 393 [False, False, True] | |
| 394 | |
| 395 substring pattern should be escaped: | |
| 396 >>> substringregexp(b'.bc').pattern | |
| 397 '\\\\.bc' | |
| 398 >>> test(b'.bc', b'abc', b'def', b'abcdefg') | |
| 399 [False, False, False] | |
| 400 | |
| 401 regex matching ('re:' prefix) | |
| 402 >>> test(b're:a.+b', b'nomatch', b'fooadef', b'fooadefbar') | |
| 403 [False, False, True] | |
| 404 | |
| 405 force substring matches ('literal:' prefix) | |
| 406 >>> test(b'literal:re:foobar', b'foobar', b're:foobar') | |
| 407 [False, True] | |
| 408 | |
| 409 case insensitive literal matches | |
| 410 >>> itest(b'BCDE', b'abc', b'def', b'abcdefg') | |
| 411 [False, False, True] | |
| 412 | |
| 413 case insensitive regex matches | |
| 414 >>> itest(b're:A.+b', b'nomatch', b'fooadef', b'fooadefBar') | |
| 415 [False, False, True] | |
| 416 """ | |
| 417 kind, pattern = _splitpattern(pattern) | |
| 418 if kind == b're': | |
| 419 try: | |
| 420 return remod.compile(pattern, flags) | |
| 421 except remod.error as e: | |
| 422 raise error.ParseError( | |
| 423 _(b'invalid regular expression: %s') % forcebytestr(e) | |
| 424 ) | |
| 425 elif kind == b'literal': | |
| 426 return remod.compile(remod.escape(pattern), flags) | |
| 427 | |
| 428 raise error.ProgrammingError(b'unhandled pattern kind: %s' % kind) | |
| 429 | |
| 430 | |
| 379 def shortuser(user): | 431 def shortuser(user): |
| 380 """Return a short representation of a user name or email address.""" | 432 """Return a short representation of a user name or email address.""" |
| 381 f = user.find(b'@') | 433 f = user.find(b'@') |
| 382 if f >= 0: | 434 if f >= 0: |
| 383 user = user[:f] | 435 user = user[:f] |