Difference between re.search() and re.findall()

Question

The following code is very strange:

 >>> words = "4324324 blahblah"
 >>> print re.findall(r'(\s)\w+', words)
 [' ']
 >>> print re.search(r'(\s)\w+', words).group()
 blahblah

The () operator seems to behave poorly with findall. Why is this? I need it for a csv file.

Edit for clarity: I want to display blahblah using findall.

I discovered that re.findall(r'\s(\w+)', words) does what I want, but have no idea why findall treats groups in this way.

Answer 1

One character off:

>>> print re.search(r'(\s)\w+', words).groups()
(' ',)
>>> print re.search(r'(\s)\w+', words).group(1)
' '

findall returns a list of all groups captured. You're getting a space back because that's what you capture. Stop capturing, and it works fine:

>>> print re.findall(r'\s\w+', words)
[' blahblah']

Use the csv module

Answer 2

If you prefer to keep the capturing groups in your regex, but you still want to find the entire contents of each match instead of the groups, you can use the following:

[m.group() for m in re.finditer(r'(\s)\w+', words)]

For example:

>>> [m.group() for m in re.finditer(r'(\s)\w+', '4324324 blahblah')]
[' blahblah']