Extracting text with no consecutive whitespaces

Question

I have a sample input below and I would like to extract each individual column using regex but it cant work for multiple consecutive blank spaces. I've tried "([0-9])\s+([0-9])\s+([A-Za-z0-9- ]+)\s{2,}([A-Za-z0-9- ]+)\s+([A-Za-z0-9]+)" and it should work for each row.

Output
Module    Ports Type                                   Model            Serial No.
--------- ----- ------------------------------------   ---------------  -----------
1         2     CCS-7354 Series Supervisor Module      7354-SPP         JD546546527
2         1     Standby supervisor                     Unknown          Unknown
3         28    28-port SFP+ 10GigE Linecard           7234S-PC         FGK10449938

For the first row of the input result, I should get:

"1" for "Output Module".
"2" for "Ports"
"CCS-7354 Series Supervisor Module" for "Type".
"7354-SPP" for Model.
"JD546546527" for "Serial No."

I'm getting "CCS-7354 Series Supervisor Module 7354-SPP " for the Type which is incorrect.

Why not just count the dashes on the 3rd line and then retrieve the columns using the counts? Then you only need to trim the white-space. Then you can validate each resulting string using a regex. Divide and conquer. — Maarten Bodewes, Dec 21 '20 at 02:23
Or you can perform the following regex multiple times (using find): `(.+?)(?:\s{2,}|$)` for each line... — Maarten Bodewes, Dec 21 '20 at 02:29
@tripleee " I would like to extract each individual column using regex " -- Why have you chosen a non-regex answer for the Duplicate? — Patrick Parker, Dec 21 '20 at 06:26

Patrick Parker · Answer 1 · 2020-12-21T06:18:10.573

Your problem is that the Type column match group [A-Za-z0-9- ]+ uses a "greedy" match.

Instead you should change it to a "reluctant" match [A-Za-z0-9- ]+?

Likewise, the Model column match group after that should also be changed to a reluctant match instead of a greedy match, so that it won't preemptively eat up all its trailing spaces.

Here is the final regex -- ([0-9])\s+([0-9])\s+([A-Za-z0-9- ]+?)\s{2,}([A-Za-z0-9- ]+?)\s+([A-Za-z0-9]+)

Test here: link

Of course there are other ways you could write the regex such that you wouldn't need to use a reluctant match syntax. For example ((?:\S|\s\S)+)

This matches non-space characters separated by at most one whitespace character.

And putting it all together, it would be: ([0-9])\s+([0-9])\s+((?:\S|\s\S)+)\s+((?:\S|\s\S)+)\s+((?:\S|\s\S)+)

Writing it this way reduces the amount of potential backtracking and should thus result in a consistently fast regex, regardless of input (although with this simple input it appears to be marginally slower).

Extracting text with no consecutive whitespaces

1 Answers1