带有特殊字符的Python正则表达式

Question

Having trouble writing a robust regular expression to grab information out of a string.

$ string1 = 'A_XYZ_THESE_WORDS'
$ string2 = 'A_ABC_THOSE_WORDS'

I would like a robust solution that pulls out from string1 or string2 respectfully 'THESE_WORDS' or 'THOSE_WORDS'.

Basically, I need something that removes everything before the first two underscores (_), but the text before them will vary.

$ get_text = re.search('(?<=A_)\w+(_)',string1)
$ print get_text.group()
$ 'XYZ_THESE_'

Answer 1

Based on your problem statement:

I need something that removes everything before the first two underscores

you don't necessarily need a regular expression:

>>> string1 = 'A_XYZ_THESE_WORDS'
>>> string1.split("_", 2)[2]
'THESE_WORDS'

The second argument to str.split is the maximum number of times to split. This will split on the first two '_' s, then take the third item (the rest of the string) from the resulting list.

This will throw an IndexError if there are fewer than two underscores in the string - this lets you know that the string is not in a format you expect, but if this behaviour is not desirable, consider:

>>> string1 = 'A_XYZ_THESE_WORDS'
>>> string1.split("_", 2)[-1]
'THESE_WORDS'  

Which takes the last item in the list from str.split , rather than assuming that there will be three. Comparison:

>>> "JUST_ONE".split("_", 2)[2]
Traceback (most recent call last):
  File "<pyshell#3>", line 1, in <module>
    "JUST_ONE".split("_", 2)[2]
IndexError: list index out of range

>>> "JUST_ONE".split("_", 2)[-1]
'ONE'

Answer 2

The below regex will print the texts which was just after to the second underscore(_),

>>> import re
>>> string1 = 'A_XYZ_THESE_WORDS'
>>> string2 = 'A_ABC_THOSE_WORDS'
>>> m = re.search(r'^[^_]*_[^_]*_(.*)$', string1)
>>> m.group(1)
'THESE_WORDS'
>>> m = re.search(r'^[^_]*_[^_]*_(.*)$', string2)
>>> m.group(1)
'THOSE_WORDS'

Answer 3

In [21]: regex = re.compile(r'^([a-zA-Z]+_){2}(.*)$')

In [22]: m = regex.search(string1)

In [23]: m.groups()
Out[23]: ('XYZ_', 'THESE_WORDS')

In [24]: m = regex.search(string2)

In [25]: m.groups()
Out[25]: ('ABC_', 'THOSE_WORDS')