Regex in Shell script

Question

I want to get rid of texts like "uc010obu.2:" in each line. I tried to write the following, but the pattern is not recognizable.

%s/uc(\d\+)(\w\+).(\d\+)://

Could someone help with the right pattern? thanks

each line is something like this:

CPEB2 uc010obu.2:c.1070_1071insCGG:p.G357delinsGG

ps: I am trying to do in vim editor.

Can you also give us the text you are trying to perform the regexp on? it gives us something to reference and test on. — Mic1780
– Mic1780, Commented Sep 17, 2014 at 19:46
Is that being used as an argument to sed? vim? Something else? Why are you grouping those matches if you are throwing it all away? — Etan Reisner
– Etan Reisner, Commented Sep 17, 2014 at 19:47
Is the text you are removing always two letters followed by three numbers three letters a period a number and ends with a colon? — Mic1780
– Mic1780, Commented Sep 17, 2014 at 19:51

Jason Hu · Accepted Answer · 2014-09-17 19:50:56Z

1

do not group your regex.

:%s/uc\d\+\w\+\.\d\+://

if you want to group, use \( and \) instead. if you are using perl-compatible tool, it's ( and ).

answered Sep 17, 2014 at 19:50

Jason Hu

6,3631 gold badge22 silver badges43 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

anubhava · Accepted Answer · 2014-09-17 19:48:39Z

1

You can use this regex in vim/sed:

%s/uc\d\+\w\+\.\d\+://

Or else:

g/uc\d\+\w\+\.\d\+:/d

answered Sep 17, 2014 at 19:48

anubhava

790k67 gold badges603 silver badges671 bronze badges

but uc010obu.2 can be ucxxxxxx.x

Sorry, bad copy/paste fixed now.

Cristian Calabro · Accepted Answer · 2014-09-17 19:51:54Z

1

If you want to match text with only the varying numbers you can try this pattern:

uc(\d+)(\w+).(\d+):

To try, you can use the website http://rubular.com/

answered Sep 17, 2014 at 19:51

Cristian Calabro

917 bronze badges