Regular Expression - Matching Multiples of 3 Characters exactly.

Jeff jeffober at gmail.com
Mon Apr 28 15:22:32 CEST 2008


Regular expressions for that sort of thing can get *really* big.  The
most efficient way would be to programmatically compose the regular
expression to be as exact as possible.

import re

def permutation(lst):
	""""
	From http://labix.org/snippets/permutations/. Computes permutations
of a
	list iteratively.
	"""
	queue = [-1]
	lenlst = len(lst)
	while queue:
		i = queue[-1]+1
		if i == lenlst:
			queue.pop()
		elif i not in queue:
			queue[-1] = i
			if len(queue) == lenlst:
				yield [lst[j] for j in queue]
			queue.append(-1)
		else:
			queue[-1] = i

def segment_re(a, b):
	"""
	Creates grouped regular expression pattern to match text between all
	possibilies of three-letter sets a and b.
	"""
	def pattern(n):
		return "(%s)" % '|'.join( [''.join(grp) for grp in permutation(n)] )

	return re.compile( r'%s(\w+?)%s' % (pattern(a), pattern(b)) )

print segment_re(["a", "b", "c"], ["d", "e", "f"])

You could extend segment_re to accept an integer to limit the (\w+?)
to a definite quantifier.  This will grow the compiled expression in
memory but make matching faster (such as \w{3,n} to match from 3 to n
characters).

See http://artfulcode.net/articles/optimizing-regular-expressions/ for
specifics on optimizing regexes.



More information about the Python-list mailing list