Mercurial > cpython-lru_cache-weakref
comparison Lib/difflib.py @ 29721:09e42f79f830 2.2
Fix SF bug #763023, difflib.py: ratio() zero division not caught
| author | Neal Norwitz <nnorwitz@gmail.com> |
|---|---|
| date | Tue, 01 Jul 2003 15:12:33 +0000 |
| parents | fc6a19f69ac7 |
| children |
comparison
equal
deleted
inserted
replaced
| 29709:1f3547d4939f | 29721:09e42f79f830 |
|---|---|
| 21 For producing human-readable deltas from sequences of lines of text. | 21 For producing human-readable deltas from sequences of lines of text. |
| 22 """ | 22 """ |
| 23 | 23 |
| 24 __all__ = ['get_close_matches', 'ndiff', 'restore', 'SequenceMatcher', | 24 __all__ = ['get_close_matches', 'ndiff', 'restore', 'SequenceMatcher', |
| 25 'Differ', 'IS_CHARACTER_JUNK', 'IS_LINE_JUNK'] | 25 'Differ', 'IS_CHARACTER_JUNK', 'IS_LINE_JUNK'] |
| 26 | |
| 27 def _calculate_ratio(matches, length): | |
| 28 if length: | |
| 29 return 2.0 * matches / length | |
| 30 return 1.0 | |
| 26 | 31 |
| 27 class SequenceMatcher: | 32 class SequenceMatcher: |
| 28 | 33 |
| 29 """ | 34 """ |
| 30 SequenceMatcher is a flexible class for comparing pairs of sequences of | 35 SequenceMatcher is a flexible class for comparing pairs of sequences of |
| 523 1.0 | 528 1.0 |
| 524 """ | 529 """ |
| 525 | 530 |
| 526 matches = reduce(lambda sum, triple: sum + triple[-1], | 531 matches = reduce(lambda sum, triple: sum + triple[-1], |
| 527 self.get_matching_blocks(), 0) | 532 self.get_matching_blocks(), 0) |
| 528 return 2.0 * matches / (len(self.a) + len(self.b)) | 533 return _calculate_ratio(matches, len(self.a) + len(self.b)) |
| 529 | 534 |
| 530 def quick_ratio(self): | 535 def quick_ratio(self): |
| 531 """Return an upper bound on ratio() relatively quickly. | 536 """Return an upper bound on ratio() relatively quickly. |
| 532 | 537 |
| 533 This isn't defined beyond that it is an upper bound on .ratio(), and | 538 This isn't defined beyond that it is an upper bound on .ratio(), and |
| 552 else: | 557 else: |
| 553 numb = fullbcount.get(elt, 0) | 558 numb = fullbcount.get(elt, 0) |
| 554 avail[elt] = numb - 1 | 559 avail[elt] = numb - 1 |
| 555 if numb > 0: | 560 if numb > 0: |
| 556 matches = matches + 1 | 561 matches = matches + 1 |
| 557 return 2.0 * matches / (len(self.a) + len(self.b)) | 562 return _calculate_ratio(matches, len(self.a) + len(self.b)) |
| 558 | 563 |
| 559 def real_quick_ratio(self): | 564 def real_quick_ratio(self): |
| 560 """Return an upper bound on ratio() very quickly. | 565 """Return an upper bound on ratio() very quickly. |
| 561 | 566 |
| 562 This isn't defined beyond that it is an upper bound on .ratio(), and | 567 This isn't defined beyond that it is an upper bound on .ratio(), and |
| 564 """ | 569 """ |
| 565 | 570 |
| 566 la, lb = len(self.a), len(self.b) | 571 la, lb = len(self.a), len(self.b) |
| 567 # can't have more matches than the number of elements in the | 572 # can't have more matches than the number of elements in the |
| 568 # shorter sequence | 573 # shorter sequence |
| 569 return 2.0 * min(la, lb) / (la + lb) | 574 return _calculate_ratio(min(la, lb), la + lb) |
| 570 | 575 |
| 571 def get_close_matches(word, possibilities, n=3, cutoff=0.6): | 576 def get_close_matches(word, possibilities, n=3, cutoff=0.6): |
| 572 """Use SequenceMatcher to return list of the best "good enough" matches. | 577 """Use SequenceMatcher to return list of the best "good enough" matches. |
| 573 | 578 |
| 574 word is a sequence for which close matches are desired (typically a | 579 word is a sequence for which close matches are desired (typically a |