Question

I would like to do something like:

temp=a.split()
#do some stuff with this new list
b=" ".join(temp)

where a is the original string, and b is after it has been modified. The problem is that when performing such methods, the newlines are removed from the new string. So how can I do this without removing newlines?

Answer 1

I assume in your third line you mean join(temp), not join(a).

To split and yet keep the exact “splitters”, you need the re.split function (or split method of RE objects) with a capturing group:

>>> import re
>>> f='tanto va\nla gatta al lardo'
>>> re.split(r'(\s+)', f)
['tanto', ' ', 'va', '\n', 'la', ' ', 'gatta', ' ', 'al', ' ', 'lardo']

The pieces you’d get from just re.split are at index 0, 2, 4, … while the odd indices have the “separators” — the exact sequences of whitespace that you’ll use to re-join the list at the end (with ''.join) to get the same whitespace the original string had.

You can either work directly on the even-spaced items, or you can first extract them:

>>> x = re.split(r'(\s+)', f)
>>> y = x[::2]
>>> y
['tanto', 'va', 'la', 'gatta', 'al', 'lardo']

then alter y as you will, e.g.:

>>> y[:] = [z+z for z in y]
>>> y
['tantotanto', 'vava', 'lala', 'gattagatta', 'alal', 'lardolardo']

then reinsert and join up:

>>> x[::2] = y
>>> ''.join(x)
'tantotanto vava\nlala gattagatta alal lardolardo'

Note that the \n is exactly in the position equivalent to where it was in the original, as desired.