regex question

[Robert Mckennon]

Hey.. I'm looking for some regex help.

I have a large csv file like:

name,v1,v2,v3
s1,1.0,2.0,3.0
s2,1.1,2.1,3.1
s3,Error: blah3,2.0,3.1
s4,1.0,2.0,Error: blah4
s5,1.0,Error: blah5,3.0

I'm looking to remove all the "Error"(s).

So I start with: 

rob@Galadriel:~/test$ sed -e 's/Error.*,//g' test.csv
name,v1,v2,v3
s1,1.0,2.0,3.0
s2,1.1,2.1,3.1
s3,3.1
s4,1.0,2.0,Error: blah2
s5,1.0,3.0

And this sort of works (ok, not really).  I think I need to tell it "up TO the next comma" because it's removing the comma.

Then there is the case on line s4 where it's the EOL.  So I need some sort of OR statement there.  (comma OR EOL).

Can some regex guru throw me a bone here please? 

Rob.

[timhol...@gmail.com]

Maybe something like "/Error:,*[^\,$]/" might do. Should match on the "Error and anything up to, but not including comma or end-of-line if I've got it right.

   Tim

--
You received this message because you are subscribed to the Google Groups "JaxLUG" group.
To unsubscribe from this group and stop receiving emails from it, send an email to jaxlug+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/jaxlug/CAErk9YCnYAL8QF2b%3DG67tUK-Ogxcyt8A%3DehWxz%3Do1bzt1SX6WA%40mail.gmail.com.

[Robert Mckennon]

Hi Tim!

Thank you for the quick reply!

That's closer... it covers the 'up to the comma', and the EOL problem,   but it leaves the rest of the error messages:

rob@skylan:~/test$ sed -e 's/Error:,*[^\,$]//g' test.csv

name,v1,v2,v3

s1,1.0,2.0,3.0

s2,1.1,2.1,3.1

s3,blah3,2.0,3.1

s4,1.0,2.0,blah4

s5,1.0,blah5,3.0

I have an assortment of Error messages....

Error:  blah1

Error:  blah2

Error:  foo

etc... and they can be lengthy, but NOT contain any commas.

Rob.

To view this discussion on the web visit https://groups.google.com/d/msgid/jaxlug/b36844db6b254351ddbeabe4c730b4c80b741fba.camel%40gmail.com.

[timhol...@gmail.com]

I did a sloppy copy. Not "Error,.*[", but "Error:.*[". The comma following the ":" should not be there.

[Rajesh Verma]

Do you accept Python? Worked for the example errors.

Rajesh Verma

To view this discussion on the web visit https://groups.google.com/d/msgid/jaxlug/fa1f3042e030935fb662015349c458a9177f72e7.camel%40gmail.com.

[Robert Mckennon]

I did think the comma was out of place, so I tried it with a period,and it's removing to the EOL:

rob@skylan:~/test$ sed -e 's/Error:.*[^\,$]//g' test.csv

name,v1,v2,v3

s1,1.0,2.0,3.0

s2,1.1,2.1,3.1

s3,

s4,1.0,2.0,

s5,1.0,

I'll try more tomorrow... thank you!

Rob.

On Sun, Oct 1, 2023 at 10:06 PM <timhol...@gmail.com> wrote:

To view this discussion on the web visit https://groups.google.com/d/msgid/jaxlug/fa1f3042e030935fb662015349c458a9177f72e7.camel%40gmail.com.

[Robert Mckennon]

Yes, I do accept python... I'll try this out tomorrow!

Thanx!

Rob.

To view this discussion on the web visit https://groups.google.com/d/msgid/jaxlug/CAKQOrS7%2BS_226cuSDSU86ooJ0ixo%3Dn8TWHgwmkV6QTW2jKF0jg%40mail.gmail.com.

[Rajesh Verma]

Great, let me know if you need any help with changes.

Rajesh Verma

To view this discussion on the web visit https://groups.google.com/d/msgid/jaxlug/CAErk9YBqLUL0Svs3bTiXZuGFf5UNpGd-rnzytiG7E7hdF1ofHw%40mail.gmail.com.