Extract Word Templates from String

Question

Is there a quick way to extract the argument-value pairs from below string in Python?

s = '[ARG0: elephants] were [ARGM-TMP: first] [V: walk] [ARGM-DIR: up]'

Output is in the form below:

ARG0 = elephants
ARGM-TMP = first
ARGM-DIR = up

Answer 1

You could find all arg/value pairs using re.findall and build a dictionary by splitting the strings:

import re
s = '[ARG0: elephants] were [ARGM-TMP: first] [V: walk] [ARGM-DIR: up]'

d = re.findall(r'\[(ARG.*?\:*?)]', s)

dict([i.split(': ') for i in d])
# {'ARG0': 'elephants', 'ARGM-TMP': 'first', 'ARGM-DIR': 'up'}

Answer 2

Read up on regex in python. And you can use a regex string like the following.

\[(?P<key>[a-zA-Z0-9-_]+): (?P<value>[a-z]+)\]