UNIX schell scripting, find string, replace at the start...

[Koing]

How do I find a string within a file say loadRuleLatest and check if it has a # at the start and if it does, can it take out the # and leave the line as is? That can be one script

And also the reverse?
How can I check that if it it finds loadRuleLatest and then takes out # at the start of the line if it is present?

Currently I we have to manually edit the pair of config files and there are 8 pairs of files. This would speed up the progress dramatically for me!

I can use the SED command but not sure how to get it to replace at the start of the file

e.g
Finds:
#/server/box1/algo/loadRuleLatest pathA/ruleAAB

changed to
/server/box1/algo/loadRuleLatest pathA/ruleAAB


Finds
/server/box1/algo/loadRuleLatest pathA/ruleAAB
#/server/box1/algo/loadRuleLatest pathA/ruleAAB

Every occurrence in file: :%s/OLD/NEW/g
I've used SED but ion this case the path after /box1 is different

Cheers guys. I've been searching online how to do it but need more help (n00b). Cheers!

Koing

 

[degibson]

sed -i "s/^#//" myFile.txt

 

[Koing]

Originally posted by: degibson
sed -i "s/^#//" myFile.txt

Cheers, mate but unfortunately that isn't going to work in my case as I have other # in other lines that I need to stay.

This is why I need to find the lines with the string 'loadRuleLatest and only replace # if its present or add it if it is not depending which file I want to change.

But cheers mate, appreciate the helpl!

Koing

 

[Koing]

sed '/loadRuleLatest/s/^#//g'

This will work for when I want to take out the # at the start of the line when I have 'loadRuleLatest' in the line. Now to bodge the other one to work...

Koing

 

[Onund]

Originally posted by: Koing
sed '/loadRuleLatest/s/^#//g'

This will work for when I want to take out the # at the start of the line when I have 'loadRuleLatest' in the line. Now to bodge the other one to work...

Koing

well, if that syntax works then you should be able to do (or very similar to):

sed '/^(^#).*loadRuleLatest/s/^/#/g'

 

[Koing]

Originally posted by: Onund

Originally posted by: Koing
sed '/loadRuleLatest/s/^#//g'

This will work for when I want to take out the # at the start of the line when I have 'loadRuleLatest' in the line. Now to bodge the other one to work...

Koing

well, if that syntax works then you should be able to do (or very similar to):

sed '/^(^#).*loadRuleLatest/s/^/#/g'

/server/box1/algo/loadRuleLatest pathA/ruleAAB
#/server/box1/algo/loadRuleLatest pathA/ruleAAB

That didn't seem to work on it?

box01 kTEST $ sed '/^(^#).*loadRuleLatest/s/^/#/g' A.txt
/server/box1/algo/loadRuleLatest pathA/ruleAAB
#/server/box1/algo/loadRuleLatest pathA/ruleAAB

/server/box1/algo/loadRuleLatest pathA/ruleAAB
#/server/box1/algo/loadRuleLatest pathA/ruleAAB

/server/box1/algo/loadRuleLatest pathA/ruleAAB
#/server/box1/algo/loadRuleLatest pathA/ruleAAB

/server/box1/algo/loadRuleLatest pathA/ruleAAB
#/server/box1/algo/loadRuleLatest pathA/ruleAAB

BTW what does the ^ mean? I understand your script should work:

sed '/^(^#).*loadRuleLatest/s/^/#/g' A.txt

Sorry mate I was being a f0cking n00b (no change here...) and that you meant ^(^#) as an OPTION and not both of them doh! Cheers mate. MUCH APPRECIATED!

sed '/^.*loadRuleLatest/s/^/#/g' A.txt add #
sed '/^#.*loadRuleLatest/s/^#//g' A.txt minus #

Cheers!

Kind Regards,
Koing

 

[Onund]

Originally posted by: Koing

Originally posted by: Onund

Originally posted by: Koing
sed '/loadRuleLatest/s/^#//g'

This will work for when I want to take out the # at the start of the line when I have 'loadRuleLatest' in the line. Now to bodge the other one to work...

Koing

well, if that syntax works then you should be able to do (or very similar to):

sed '/^(^#).*loadRuleLatest/s/^/#/g'

/server/box1/algo/loadRuleLatest pathA/ruleAAB
#/server/box1/algo/loadRuleLatest pathA/ruleAAB

That didn't seem to work on it?

box01 kTEST $ sed '/^(^#).*loadRuleLatest/s/^/#/g' A.txt
/server/box1/algo/loadRuleLatest pathA/ruleAAB
#/server/box1/algo/loadRuleLatest pathA/ruleAAB

/server/box1/algo/loadRuleLatest pathA/ruleAAB
#/server/box1/algo/loadRuleLatest pathA/ruleAAB

/server/box1/algo/loadRuleLatest pathA/ruleAAB
#/server/box1/algo/loadRuleLatest pathA/ruleAAB

/server/box1/algo/loadRuleLatest pathA/ruleAAB
#/server/box1/algo/loadRuleLatest pathA/ruleAAB

BTW what does the ^ mean? I understand your script should work:

sed '/^(^#).*loadRuleLatest/s/^/#/g' A.txt

Sorry mate I was being a f0cking n00b (no change here...) and that you meant ^(^#) as an OPTION and not both of them doh! Cheers mate. MUCH APPRECIATED!

sed '/^.*loadRuleLatest/s/^/#/g' A.txt add #
sed '/^#.*loadRuleLatest/s/^#//g' A.txt minus #

Cheers!

Kind Regards,
Koing

no... i did mean ^(^#) in the expression (but I'm wrong, see below).

If you do this:
sed '/^.*loadRuleLatest/s/^/#/g' A.txt add #

and you fed it this:
/server/box1/algo/loadRuleLatest pathA/ruleAAB
#/server/box1/algo/loadRuleLatest pathA/ruleAAB

the ouput would be:
#/server/box1/algo/loadRuleLatest pathA/ruleAAB
##/server/box1/algo/loadRuleLatest pathA/ruleAAB

the ^ means different things depending on where it is. In /^a/ the ^ is position and means 'at the start of the string' so this will find 'a' at the start of a string.

If... ah crap... I think I should have told you [^#]... so you can define groups of characters right? [abc] means match any of 'a' 'b' or 'c'. If you do [^abc] then you're doing the compliment or, match anything but 'a', 'b' or 'c'. I told you (^#). when you put stuff in parenthesis you're telling regex to match that sequence so (abc) means match 'abc'. In this case, it's probably looking for '^#'. My bad.


This rule below says "find any zero or more characters at the start of the string, followed by 'loadruleLatest'. Change the start of the string into a #"
sed '/^.*loadRuleLatest/s/^/#/g' A.txt add #

This rule below says "find a line that does NOT start with #, follow by any zero or more characters, followed by loadRuleLatest. change the start of that string into a #"
sed '/^[^#].*loadRuleLatest/s/^/#/g' A.txt add #

 

[Koing]

Onund got you mate! I appreciate man. Works a treat man

sed '/^[^#].*loadRuleLatest/s/^/#/g' A.txt add #
sed '/^#.*loadRuleLatest/s/^#//g' A.txt remove #

I have a few scripts that I can use to create a backup of a file and the then I can get it to read in a filename and then make the changes on to it

Again cheers!

Koing