Path: csiph.com!usenet.pasdenom.info!aioe.org!news.stack.nl!newsfeed.xs4all.nl!newsfeed2a.news.xs4all.nl!xs4all!post.news.xs4all.nl!not-for-mail
MIME-Version: 1.0
In-Reply-To: <CAPV1RAAiD3qWqrAJYc4yb80CaHG4A9N4jz4aY_CsyJr_nAUx9Q@mail.gmail.com>
References: <CAPV1RAAiD3qWqrAJYc4yb80CaHG4A9N4jz4aY_CsyJr_nAUx9Q@mail.gmail.com>
From: Devin Jeanpierre <jeanpierreda@gmail.com>
Date: Mon, 20 Jan 2014 03:09:42 -0800
Subject: Re: regex multiple patterns in order
To: km <srikrishnamohan@gmail.com>
Content-Type: text/plain; charset=UTF-8
Cc: "comp.lang.python" <python-list@python.org>
Precedence: list
Newsgroups: comp.lang.python
Message-ID: <mailman.5746.1390216232.18130.python-list@python.org>
Lines: 28
NNTP-Posting-Host: 2001:888:2000:d::a6
Xref: csiph.com comp.lang.python:64356

On Mon, Jan 20, 2014 at 2:44 AM, km <srikrishnamohan@gmail.com> wrote:
> I am trying to find sub sequence patterns but constrained by the order in
> which they occur
> For example
>
>>>> p = re.compile('(CAA)+?(TCT)+?(TA)+?')
>>>> p.findall('CAACAACAATCTTCTTCTTCTTATATA')
> [('CAA', 'TCT', 'TA')]
>
> But I instead find only one instance of the CAA/TCT/TA in that order.
> How can I get 3 matches of CAA, followed by  four matches of TCT followed by
> 2 matches of TA ?
> Well these patterns (CAA/TCT/TA) can occur any number of  times and atleast
> once so I have to use + in the regex.

You want to include the '+' in the parens so that repetitions are
included in the match, but you still want to group CAA etc. together;
for that, you can use non-capturing groups.

I don't see how TA could ever match two, though. It'd match once
as-is, or thrice if you make the repetition greedy (get rid of the
?s).

>>> p = re.compile('((?:CAA)+?)((?:TCT)+?)((?:TA)+?)')
>>> p.findall('CAACAACAATCTTCTTCTTCTTATATA')
[('CAACAACAA', 'TCTTCTTCTTCT', 'TA')]

-- Devin