Python中的正则表达式：（\\ w）+的​​search（）vs findall（）

Question

I have created a regular expression as:

agentRegex = re.compile(r'Agent (\w)+')

And then I performed search() operation as:

agentRegex.search('Agent Alice told Agent Carol that Agent Eve knew Agent Bob was a double agent.').group()

I obtained 'Agent Alice' as output.

But when I performed findall() operation:

agentRegex.findall('Agent Alice told Agent Carol that Agent Eve knew Agent Bob was a double agent.') 

The output was ['e', 'l', 'e', 'b'] .

Shouldn't the output be ['Alice Agent', 'Agent Carol', 'Agent Eve', 'Agent Bob'] ?

Answer 1

re.findall() by default outputs a list of captured groups, in your case (\\w+) .

Get rid of the captured group:

Agent \w+

Example:

>>> agentRegex = re.compile(r'Agent \w+')

>>> agentRegex.findall('Agent Alice told Agent Carol that Agent Eve knew Agent Bob was a double agent.') 
['Agent Alice', 'Agent Carol', 'Agent Eve', 'Agent Bob']

Answer 2

Your regex:

'Agent (\w)+'

It will keep matching and capturing single \\w characters after 'Agent ' and will keep overwriting the matched group with the next match. Thats how you get ['e', 'l', 'e', 'b'] which are the last characters of ['Alice', 'Carol', 'Eve', 'Bob']

You got correct answer in .search().group() because group() defaults to group(0) which contains everything that got matched, but if you do .search().group(1) you will get ['e'] .

What you are looking for is capture the Agent as well as next word. So yo u can try like heemayl and Dietrich suggested.

Answer 3

You could do this too:

import re
agentRegex = re.compile(r'Agent\s+[^\s]+')
print agentRegex.findall('Agent Alice told Agent Carol that Agent Eve knew Agent Bob was a double agent.') 
# ['Agent Alice', 'Agent Carol', 'Agent Eve', 'Agent Bob']