Understanding group and or in regular expression

I'm looking for words like "one year", "two years", "2-3 years" or "3 - 4 years" in a long string. I've tried to do it using regular expressions. But I'm not sure that I got it when groups are involved.

Let's see what I mean:

import re

text = 'one year, honey 2-5 year, dressed six, ten'
pattern = r'(one|two|three|four|five|six|seven|eight|nine|ten| \
                  eleven|twelve|thirteen|fourteen|fifteen|sixteen|seventeen| \
                  eighteen|nineteen|twenty|[0-9]+[- ]*[0-9]*)[+ ]*year?'

re.findall(pattern, text)  # ['one', '2-5']

My problem is that I want ['one year', '2-5 years']. I'm not sure how to do it. If I forgot about the numbers in words:

pattern = r'[0-9]+[- ]*[0-9]*[\+ ]*year?'
re.findall(pattern, text)  # ['2-5 years']

Why I got years in the second and not in the first? How I can modify it to got years in the first one?

Thanks in advance,

Solution

You need to fix the pattern to match the numbers first. Here is an example:

>>> pattern = r'''(?x)\b(?:[0-9]+(?:[- ]*[0-9]+)?|one|two|three|four|five|six|seven|eight|nine|ten
|eleven|twelve|thirteen|fourteen|fifteen|sixteen|seventeen|eighteen|nineteen|twenty)
[+ ]*years?\b'''
>>> re.findall(pattern, text)
['one year', '2-5 year']

See the Python demo and the regex demo.

Details

(?x) - re.X / re.VERBOSE inline modifier
\b - a word boundary
(?: - start of a non-capturing group
- [0-9]+(?:[- ]*[0-9]+)? - one or more digits followed with zero or more spaces or - and then one or more digits
- |one|two|three|four|five|six|seven|eight|nine|ten|eleven|twelve|thirteen|fourteen|fifteen|sixteen|seventeen|eighteen|nineteen|twenty - one of the words in the alternation list
) - end of the non-capturing group
[+ ]* - zero or more + or spaces
years? - year or years
\b - a word boundary.