正则表达式：组函数

Question

I want to extract newsletter_ and _mon_gallery from the phrase

002c2833d0-newsletter_20131028_mon_gallery

I tried with ([^\\d-_]+){3,}

002c2833d0-newsletter_20131028_mon_gallery

So i can check in http://www.regexpal.com/ . it is visually separating the two entities newsletter_ and _mon_gallery .

But the problem is i am not able to retrieve the matched values from group function.

import re
string='002c2833d0-newsletter_20131028_mon_enamour'
m=re.search('([^\d-]+){3,}',string)
print m.group()

i just get

newsletter_

Answer 1

re.search() is designed to return the first match. You want

m = re.findall('[^\d-]{3,}',string)

Note that I've edited your regex to remove the nested quantifiers (can you say " catastrophic backtracking "?) and the unnecessary (and harmful if repeated) capturing group.

Answer 2

You can use findall , but you have to slightly change the regular expression from this:

([^\d-]+){3,}

to this:

([^\d-]{3,})

(In general, there's no need to have both + and {3,} together as the latter implies the first.)

>>> re.findall('[^\d-]{3,}', string)
['newsletter_', '_mon_enamour']