Matching an exact amount of characters?

[metaperl]

Hi Paul, 2 recent pypi libraries are designed to make regular expressions easier to construct:

• pregex
• humre

I was looking at both of them and asking myself why I would ever use either if I have PyParsing. After examining the PyParsing docs, one thing that I did not see was the ability to specify an exact amount of a particular character set. Under expression subclasses we see an example of a us phone number:

us_phone_number = Word(nums) + ('-' + Word(nums)) * (1,2)

But first of all, a number doesnt typically include letters. And second of all, the amount of characters in the various elements of a phone number are specific.

So, keeping to the topic of this post, how does one specify an exact number of characters for an element of a PyParsing expression?

[ptmcg]

To specifically answer your question, Word can take min, max, and exact arguments to have more control over the number of matched characters.

We don't see this in us_phone_number because this is an ambiguous expression. It could match 123-4567 or 012-345-6789. To make this a little less ambiguous, we could write it as:

local_phone_number = Word(nums, exact=3) + '-' + Word(nums, exact=4)
us_phone_number = '(' + Word(nums, exact=3) + ')' + local_phone_number
phone_number = us_phone_number | local_phone_number

phone_number.run_tests("""\
(212)555-1212
234-5678
""")