finditer

[Jason Friedman]

On Mon, Jul 7, 2014 at 1:19 AM, gintare <g.statkute at gmail.com> wrote:
> If smbd has time, maybe you could advice how to accomplish this task in faster way.
>
> I have a text = """ word{vb}
> wordtransl {vb}
>
> sent1.
>
> sent1trans.
>
> sent2
>
> sent2trans... """
>
> I need to match once wordtransl, and than many times repeating patterns consisting of sent and senttrans.

You might try itertools.groupby
(https://docs.python.org/3/library/itertools.html#module-itertools).

text = """ word{vb}
wordtransl {vb}

sent1

sent1trans

sent2

sent2trans
"""

import itertools
import re
result_list = list()
lines = text.split("\n")

for line in lines[:]:
    if line.startswith("sent"):
        break
    lines.pop(0)

def is_start(x):
    pattern = re.compile(r"sent\d+$")
    if re.search(pattern, x):
        return True

for key, mygroup in itertools.groupby(lines, is_start):
    result_list.append(list(mygroup))

print(result_list)