diff options
author | Guido van Rossum <guido@python.org> | 1997-07-17 22:39:13 (GMT) |
---|---|---|
committer | Guido van Rossum <guido@python.org> | 1997-07-17 22:39:13 (GMT) |
commit | 9e18ec7dc9e8452035085ea34d8f4dda43c8baed (patch) | |
tree | cc62ea42022b87ba6e94e98483ee7f18d29183e3 /Lib | |
parent | a4f1a78b6eaed4d5d2c609d2e3160ec64535d22a (diff) | |
download | cpython-9e18ec7dc9e8452035085ea34d8f4dda43c8baed.zip cpython-9e18ec7dc9e8452035085ea34d8f4dda43c8baed.tar.gz cpython-9e18ec7dc9e8452035085ea34d8f4dda43c8baed.tar.bz2 |
Correctly implement sub, subn, and split. Also correct and augment
the cache code.
Diffstat (limited to 'Lib')
-rw-r--r-- | Lib/re.py | 98 |
1 files changed, 88 insertions, 10 deletions
@@ -40,7 +40,7 @@ def valid_identifier(id): _cache = {} _MAXCACHE = 20 -def _cachecompile(pattern, flags): +def _cachecompile(pattern, flags=0): key = (pattern, flags) try: return _cache[key] @@ -59,13 +59,19 @@ def search(pattern, string, flags=0): return _cachecompile(pattern, flags).search(string) def sub(pattern, repl, string, count=0): - return _cachecompile(pattern).sub(repl, string, count) + if type(pattern) == type(''): + pattern = _cachecompile(pattern) + return pattern.sub(repl, string, count) def subn(pattern, repl, string, count=0): - return _cachecompile(pattern).subn(repl, string, count) + if type(pattern) == type(''): + pattern = _cachecompile(pattern) + return pattern.subn(repl, string, count) def split(pattern, string, maxsplit=0): - return _cachecompile(pattern).subn(string, maxsplit) + if type(pattern) == type(''): + pattern = _cachecompile(pattern) + return pattern.split(string, maxsplit) # # @@ -126,14 +132,86 @@ class RegexObject: regs) def sub(self, repl, string, count=0): - pass - - def subn(self, repl, string, count=0): - pass + return self.subn(repl, string, count)[0] - def split(self, string, maxsplit=0): - pass + def subn(self, repl, source, count=0): + if count < 0: raise error, "negative substibution count" + if count == 0: import sys; count = sys.maxint + if type(repl) == type(''): + if '\\' in repl: + repl = lambda m, r=repl: _expand(m, r) + else: + repl = lambda m, r=repl: r + n = 0 # Number of matches + pos = 0 # Where to start searching + lastmatch = -1 # End of last match + results = [] # Substrings making up the result + end = len(source) + while n < count and pos <= end: + m = self.search(source, pos) + if not m: break + i, j = m.span(0) + if i == j == lastmatch: + # Empty match adjacent to previous match + pos = pos+1 + results.append(source[lastmatch:pos]) + continue + if pos < i: results.append(source[pos:i]) + results.append(repl(m)) + pos = lastmatch = j + if i == j: + # Last match was empty; don't try here again + pos = pos+1 + results.append(source[lastmatch:pos]) + n = n+1 + results.append(source[pos:]) + return (string.join(results, ''), n) + def split(self, source, maxsplit=0): + if maxsplit < 0: raise error, "negative split count" + if maxsplit == 0: import sys; maxsplit = sys.maxint + n = 0 + pos = 0 + lastmatch = 0 + results = [] + end = len(source) + while n < maxsplit: + m = self.search(source, pos) + if not m: break + i, j = m.span(0) + if i == j: + # Empty match + if pos >= end: break + pos = pos+1 + continue + results.append(source[lastmatch:i]) + g = m.group() + if g: + results[len(results):] = list(g) + pos = lastmatch = j + results.append(source[lastmatch:]) + return results + +def _expand(m, repl): + results = [] + index = 0 + size = len(repl) + while index < size: + found = string.find(repl, '\\', index) + if found < 0: + results.append(repl[index:]) + break + if found > index: + results.append(repl[index:found]) + escape_type, value, index = expand_escape(repl, found+1, REPLACEMENT) + if escape_type == CHAR: + results.append(value) + elif escape_type == MEMORY_REFERENCE: + results.append(m.group(value)) + else: + raise error, "bad escape in replacement" + return string.join(results, '') + class MatchObject: def __init__(self, re, string, pos, regs): self.re = re |