Split Filename and extension in KSH 2

[davism]

It is possible to split the filename and extension in KSH? If so, how?

I have a script called test.ksh

in the script I have:

echo $0

The $0 returns "test.ksh"

But how can I get just the item, as a variable, to the left of the "." while retaining the contents of the item to the right of the "." in a variable.

In Perl, it can be done with:

($filename, $file_ext) = split/\./,$original_file;

But how could/would I do this in KSH?

Any assistance would greatly be appreciated.

 

[KenCunningham]

Something like:

echo $0 | cut -f1 -d'.'

?

HTH

 

[davism]

Worked out GREAT! But can you do me a favor and explain that cut statement?

Would be greatly appreciated.

Thanks

 

[KenCunningham]

Glad to help. The cut takes the first 'field' (-f1) of the text and prints everything before the first defined delimiter (-d'.')

Have a good weekend!

 

[davism]

Great! Thanks for the explanation and very helpful!

 

[thutmose]

OK,


Here is another method


echo $0 | awk -F . '{print $1}'

 

[davism]

Thutmose,

Appreciate the additional info.

Can you provide an explanation on this one as well?

Would be greatly appreciated.

Thanks

 

[hvn]

echo $0 | awk -F . '{print $1}'

awk parses the line test.ksh, while the "-F ." tells awk to use "." as a field seperator, which then effectively splits the "test.ksh" into "test" and "ksh" for awk, which reads "test" as "$1" and "ksh" as "$2".

'{print $}' tells awk to print the value of "$1" which is "test"

 

[nwardez]

Hi,

And another one (only works with ksh):

filename=${original_file%%.*}
(search the pattern matching ".*" at the end of the variable's value, delete the longuest part of it and return the rest)

file_ext=${original_file##*.}
(the same from the beginning)

If someone could be nice enough to give the solution with sed ... ;-)

 

[danielhozac]

A sed solution shouldn't be hard.

filename=`echo $0 | sed 's/^\(.*\)\.\([^.]*\)/\1/'`
fileext=`echo $0 | sed 's/^\(.*\)\.\([^.]*\)/\2/'`

//Daniel

 

[davism]

KenCunningham and Thutmose,

how would I assign the results of your solutions to a variable?

chk_val = $0 | cut -f1 -d'.'

or

chk_val = $0 | awk -F. '{print $1}'

when I try that I can't get anything that I can echo to confirm (i.e. echo chk_val)

Any information on that?

 

[davism]

KenCunningham and Thutmose,

how would I assign the results of your solutions to a variable?

chk_val = $0 | cut -f1 -d'.'

or

chk_val = $0 | awk -F. '{print $1}'

when I try that I can't get anything that I can echo to confirm (i.e. echo chk_val)

Any information on that?

 

[danielhozac]

Put it in [tt]$([/tt] and [tt])[/tt] //Daniel

 

[davism]

Daniel,

Can you elaborate a little more? Please.

Thanks

 

[danielhozac]

chk_val = $(echo $0 | awk -F. '{print $1}')

//Daniel

 

[davism]

I tried that as below:

chk_val = $(echo $0 | awk -F. '{print $1}')

It's giving me a:

chk_val: {kshname} {line #} not found

Where {kshname} and {line #} are dependant upon what I am using for the name and the line # of where things are at.

What am I doing wrong?

 

[davism]

The problem is/was a space before and after the "=". I didn't realize that having those would cause an issue.

Thanks for the assistance and the information you provided.

Very helpful.