Python regex to extract number of processors

Question

I have a string which contains the number of processors:

SQLDB_GP_Gen5_2

The number is after _Gen and before _ (the number 5). How can I extract this using python and regular expressions?

I am trying to do it like this but don't get a match:

re.match('_Gen(.*?)_', 'SQLDB_GP_Gen5_2')

I was also trying this using pandas:

x['SLO'].extract(pat = '(?<=_Gen).*?(?:(?!_).)')

But this also wasn't working. (x is a Series)

Can someone please also point me to a book/tutorial site where I can learn regex and how to use with Pandas.

Thanks,

Mick

You can use lookaround [`(?<=_Gen)[^_]+(?=_)`](https://regex101.com/r/BlrhaB/2/) — Code Maniac, Jan 28 '20 at 11:16

Rúben · Accepted Answer · 2020-01-28T11:20:22.987

2

re.match searches from the beginning of the string. Use re.search instead, and retrieve the first capturing group:

>>> re.search(r'_Gen(\d+)_', 'SQLDB_GP_Gen5_2').group(1)
'5'

edited Jan 28 '20 at 11:20

answered Jan 28 '20 at 11:17

Rúben

1

You could make the search better by only looking for numbers instead of `.*?` – James Jan 28 '20 at 11:19
1

I think I prefer `'_Gen(\d+?)_'` just for being explicit, and making sure we at least get 1 digit – DeepSpace Jan 28 '20 at 11:19
Thanks for the suggestions, I initially went with OP's search pattern as it also works – Rúben Jan 28 '20 at 11:21
In Pandas, there is no need to call `re.search` explicitly, there are dedicated vectorized methods for that. – Wiktor Stribiżew Jan 28 '20 at 11:24

Wiktor Stribiżew · Answer 2 · 2020-01-28T11:26:08.287

2

You need to use Series.str.extract with a pattern containing a capturing group:

x['SLO'].str.extract(r'_Gen(.*?)_', expand=False)
        ^^^^           ^^^^^^^^^^^

To only match a number, use r'_Gen(\d+)_'.

NOTES:

With Series.str.extract, you need to use a capturing group, the method only returns any value if it is captured
r'_Gen(.*?)_' will match _Gen, then will capture any 0+ chars other than line break chars as few as possible, and then match _. If you use \d+, it will only match 1+ digits.

edited Jan 28 '20 at 11:26

answered Jan 28 '20 at 11:19

Wiktor Stribiżew

1

Yes, that works. Thanks. This is the best solution is using Pandas. – Mick Jan 28 '20 at 11:35

score 0 · Answer 3 · answered Jan 28 '20 at 11:20

0

Using re :

re.findall(r'Gen(.*)_',text)[0]

answered Jan 28 '20 at 11:20

Vikas Periyadath

3 Answers3