a regular expression question

[brassbin]

hi,

i'm processing a netgroup file, trying to delete entries from the netgroup file, here's the issue i have

so say i have this pattern in the netgroup file

ng1 (-,bob,) (-,dick,) (,jane,) (,bob,)

so i want to match the entries "bob" and delete it, here's what i use

s/\(.*,\s*bob\s*,.*\)//g

i do want to use the "g" flag to match multiple instances, i thought it would match the shortest possible patterns, but it does the opposite, it matches the widest pattern, from the first instance of "bob" all the way to the end of the line, matching the ")" at the end of the line

what am i doing wrong?

Thank you very much for your help!

[toolic]

Be less greedy:

s/\(.?,\s*bob\s*,.?\)//g;

• perlre

[brassbin]

i know the netgroup file i'm processing probably doesn't have 3 spaces before the first comma, what if it does?

i thought regular expression matches the short possible pattern, not the longest, am i wrong?

[roboticus]

brassbin:

As toolic mentioned, the .* construct is greedy and tries to get the longest match. If you turn it into .*?, it then looks for the shortest match. If you read the link he provided, you could verify that yourself.

...roboticus

When your only tool is a hammer, all problems look like your thumb.