Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.python > #17195 > unrolled thread

Regexp : repeated group identification

Back to article view | Back to comp.lang.python

Contents

  Regexp : repeated group identification candide <candide@free.invalid> - 2011-12-14 12:12 +0100
    Re: Regexp : repeated group identification Vlastimil Brom <vlastimil.brom@gmail.com> - 2011-12-14 12:34 +0100
      Re: Regexp : repeated group identification candide <candide@free.invalid> - 2011-12-14 13:57 +0100
        Re: Regexp : repeated group identification Vlastimil Brom <vlastimil.brom@gmail.com> - 2011-12-14 14:38 +0100

#17195 — Regexp : repeated group identification

Consider the following code

# ----------------------------
import re

z=re.match('(Spam\d)+', 'Spam4Spam2Spam7Spam8')
print z.group(0)
print z.group(1)
# ----------------------------

outputting :

----------------------------
Spam4Spam2Spam7Spam8
Spam8
----------------------------

The '(Spam\d)+' regexp is tested against 'Spam4Spam2Spam7Spam8' and the 
regexp matches the string.

Group numbered one within the regex '(Spam\d)+' refers to Spam\d

The fours substrings

Spam4   Spam2   Spam7  and  Spam8

match the group numbered 1.

So I don't understand why z.group(1) gives the last substring (ie Spam8 
as the output shows), why not an another one, Spam4 for example ?

[toc] | [next] | [standalone]

#17196

2011/12/14 candide <candide@free.invalid>:
> Consider the following code
>
> # ----------------------------
> import re
>
> z=re.match('(Spam\d)+', 'Spam4Spam2Spam7Spam8')
> print z.group(0)
> print z.group(1)
> # ----------------------------
>
> outputting :
>
> ----------------------------
> Spam4Spam2Spam7Spam8
> Spam8
> ----------------------------
>
> The '(Spam\d)+' regexp is tested against 'Spam4Spam2Spam7Spam8' and the
> regexp matches the string.
>
> Group numbered one within the regex '(Spam\d)+' refers to Spam\d
>
> The fours substrings
>
> Spam4   Spam2   Spam7  and  Spam8
>
> match the group numbered 1.
>
> So I don't understand why z.group(1) gives the last substring (ie Spam8 as
> the output shows), why not an another one, Spam4 for example ?
> --
> http://mail.python.org/mailman/listinfo/python-list

Hi,
you may find a tiny notice in the re docs on this:
http://docs.python.org/library/re.html#re.MatchObject.group

"If a group is contained in a part of the pattern that matched
multiple times, the last match is returned."

If you need to work with the content captured in the repeated group,
you may check the new regex implementation:
http://pypi.python.org/pypi/regex

Which has a special "captures" method of the match object for this
(beyond many other improvements):

>>> import regex
>>> m=regex.match('(Spam\d)+', 'Spam4Spam2Spam7Spam8')
>>> m.captures(1)
['Spam4', 'Spam2', 'Spam7', 'Spam8']
>>>

hth,
  vbr

[toc] | [prev] | [next] | [standalone]

#17209

Le 14/12/2011 12:34, Vlastimil Brom a écrit :

> "If a group is contained in a part of the pattern that matched
> multiple times, the last match is returned."
>

I missed this point, your answer matches my question ;) thanks.


> If you need to work with the content captured in the repeated group,
> you may check the new regex implementation:
> http://pypi.python.org/pypi/regex
>
> Which has a special "captures" method of the match object for this
> (beyond many other improvements):
>
>>>> import regex
>>>> m=regex.match('(Spam\d)+', 'Spam4Spam2Spam7Spam8')
>>>> m.captures(1)
> ['Spam4', 'Spam2', 'Spam7', 'Spam8']
>>>>
>


Thanks for the reference and the example. I didn't know of this 
reimplementation, hoping it offers the Aho-Corasick algorithm allowing 
multiple keys search.

[toc] | [prev] | [next] | [standalone]

#17215

2011/12/14 candide <candide@free.invalid>:
...
>
> Thanks for the reference and the example. I didn't know of this
> reimplementation, hoping it offers the Aho-Corasick algorithm allowing
> multiple keys search.
> --
> http://mail.python.org/mailman/listinfo/python-list

Hi,
I am not sure about the underlying algorithm (it could as well be an
internal expansion of the alternatives like ...|...|...), but you can
use a list (set, actually) of alternatives to search for.
check the "named lists" feature,
\L<...>

hth,
   vbr

[toc] | [prev] | [standalone]

Back to top | Article view | comp.lang.python

csiph-web