UNIX for Dummies Questions & Answers

Hi guys,

I want to replace certain values with the number 1.

But it is also replacing other values which contain the value I want to replace. e.g.:


I want to replace ID-INTERNAL with 1, that's no problem but it will also replace ID-INTERNAL-NON-REM with 1-NON-REM

I don't want to replace that.
I tought of doing grep -w ID-INTERNAL but that still returns both any ideas ?

You can use sed itself.

Try this out

Code:

sed -e 's/ID-INTERNAL /1 /g'

Please note the extra white-space after ID-INTERNAL and 1. This assumes you have ID-INTERNAL as a standalone word.

Vino

no didnt work,

because I might have an instance of ID-INTERNAL,

Try this...

sed -e 's/ID-INTERNAL /1 /g' -e 's/ID-INTERNAL[^-]/1/g' -e 's/ID-INTERNAL$/1/g' $filename

Thanks for that it works now, any chance you could explain what you exactly did please ?

Here goes the explaination...

sed -e 's/ID-INTERNAL /1 /g' -e 's/ID-INTERNAL[^-]/1/g' -e 's/ID-INTERNAL$/1/g' $filename


the first replace will make replace only those where space is the next character then we replace all other then '-' characters then comes provision for if the next character being newline assuming your variable ends the line it is in.

hope its clear, if not pls lemme know.

This was required to do so..as we need to prioritze and replace, there could be a better way to achieve this say with only one expression..but this is what i could come up at first instance and posted...

if any better suggestions with anyone...do share...

Thanks

Thanks RishiPahuja,

That make things clear. You might have noticed a new topic I have posted which describes a different solution to this problem.

The reason why is that ID-INTERNAL, will be replaced with 1 but also the comma will be removed and that is wrong. It doesn' necessarily have to be a comma it can be a brackes or plus sign,...


examples:
ID-INTERNAL,
ID-INTERNAL)
ID-INTERNAL+

Any ideas ?