Python使用正则表达式从字符串中删除最后一个字符_

Question

I know there are a bunch of other regex questions, but I was hoping someone could point out what is wrong with my regex. I have done some research into it and it looks like it should work. I used rubular to test it, yes I know that is regex for ruby, but the same rules I used should apply to python from what it looks like in the python docs

Currently I have

a = ["SDFSD_SFSDF234234","SDFSDF_SDFSDF_234324","TSFSD_SDF_213123"]
c = [re.sub(r'[A-Z]+', "", x) for x in a]

which returns

['SDFSD_SFSDF', 'SDFSDF_SDFSDF_', 'TSFSD_SDF_']

But I want it to return

['SDFSD_SFSDF', 'SDFSDF_SDFSDF', 'TSFSD_SDF']

I try to use this regex

c = [re.sub(r'$?_[^A-Z_]+', "", x) for x in a]

but I am getting this error

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/lib64/python2.6/re.py", line 151, in sub
    return _compile(pattern, 0).sub(repl, string, count)
  File "/usr/lib64/python2.6/re.py", line 245, in _compile
    raise error, v # invalid expression

Can anyone help me figure out what I am doing wrong?

Answer 1

import re

a = ["SDFSD_SFSDF234234","SDFSDF_SDFSDF_234324","TSFSD_SDF_213123"]
c = [re.match(r'[A-Z_]+[A-Z]', x).group() for x in a]

print c

Results:

['SDFSD_SFSDF', 'SDFSDF_SDFSDF', 'TSFSD_SDF']

Please note, that "re.sub" which you use in your example is a regex replace command, not a search. Your regex seems to be matching for what you're asking for, not what you're trying to get rid of to get what you're asking for.

Answer 2

You could insert 'lookahead' into your regexp. Written as (?=...) your regexp will match only text followed by whatever you put in the … . So in your case you could choose to ignore the underscore unless it is followed by [AZ] . Your reg exp will look like this: r'[AZ]+_(?[AZ])' so an underscore not followed by letters will be ignored.

Answer 3

Without regex using rstrip :

a = ["ends_with_underscore_", "does_not", "multiple_____"]
b = [ x.rstrip("_") for x in a]
print b

>> ['ends_with_underscore', 'does_not', 'multiple']

Answer 4

>>> import re
>>> a = ["SDFSD_SFSDF234234","SDFSDF_SDFSDF_234324","TSFSD_SDF_213123"]
>>> c = [re.sub('_?\d+','',x) for x in a]
>>> c
['SDFSD_SFSDF', 'SDFSDF_SDFSDF', 'TSFSD_SDF']
>>>

It's short and simple. Basically, it's saying "replace everything that is a stream of digits or a stream of digits preceded by an _".

Answer 5

The error in:

c = [re.sub(r'$?_[^A-Z_]+', "", x) for x in a]

Is caused by the ? , it is not preceded by any characters so it doesn't know what to match 0 or 1 times. If you change it to:

>>> [re.sub(r'_?[^A-Z_]+$', "", x) for x in a]
['SDFSD_SFSDF', 'SDFSDF_SDFSDF', 'TSFSD_SDF']

It works as you expect.

Another thing, $ is used to detonate the end of the line, so it probably shouldn't be the first character.