(old)Puppy Linux Discussion Forum

Posted: **Thu 06 Dec 2012, 03:52**

Hi.

I have a list like this one:

LP2_AbiWord-2.8.6.sfs
LP2_AFI-1.0.3.sfs
...
LP2_Aqualung09b11.sfs
LP2_Ardour286.sfs
...
LP2_AsUnder_2.1.sfs
LP2_AutoAdjustPhoto_GUI-1.0.1.sfs
...
...
...

I want to remove the "LP2_" prefix and everything except the name, like the "09b11.sfs" from Aqualung, the "-1.0.1.sfs" from AutoAdjustPhoto, the "_2.1.sfs" from AsUnder and the "-2.8.6.sfs" from AbiWord.

Could this be done using a single command like sed?

If, how?

Or will the different endings:" -1.0.1.sfs" - "_2.1.sfs" - "09b11.sfs" cause problems?

Thanks

RSH

Edit: I want to do this using bash.

Posted: **Thu 06 Dec 2012, 11:41**

The LP2 prefix is easy.

Assuming that stuff is stored in a file called input.txt (one for each line), you can do either:

Code: Select all

#!/bin/sh
while read p; do
   echo ${p#LP2_}
done < input.txt > output.txt

or

Code: Select all

sed 's/^LP2_//' input.txt > output.txt

The suffix, however, is difficult, because there is no way to know a certain word is part of the suffix to be erased, or it is part of the name you want to keep, so you'll have to use a marker that is unique to separate between the words you want to keep and the suffix you want to delete. It can the the last underscore (_), last minus (-), or whatever.

I will assume here that you've decided that all the suffix will be preceded by an underscore; that is, everything after the last underscore will be deleted.

Code: Select all

#!/bin/sh
while read p; do
   echo ${p%_*}
done < input.txt > output.txt

or

Code: Select all

sed 's/_[^_]*$//' input.txt > output.txt

Posted: **Thu 06 Dec 2012, 11:49**

RSH wrote:Hi.

I have a list like this one:
Code: Select all
LP2_AbiWord-2.8.6.sfs
LP2_AFI-1.0.3.sfs
...
LP2_Aqualung09b11.sfs
LP2_Ardour286.sfs
...
LP2_AsUnder_2.1.sfs
LP2_AutoAdjustPhoto_GUI-1.0.1.sfs
...
...
...
I want to remove the "LP2_" prefix and everything except the name, like the "09b11.sfs" from Aqualung, the "-1.0.1.sfs" from AutoAdjustPhoto, the "_2.1.sfs" from AsUnder and the "-2.8.6.sfs" from AbiWord.

Could this be done using a single command like sed?

If, how?

Or will the different endings:" -1.0.1.sfs" - "_2.1.sfs" - "09b11.sfs" cause problems?

Thanks

RSH

Edit: I want to do this using bash.

Will something like this help

Code: Select all

#!/bin/bash
sed -r 's/[_-0123456789]+/ /g' ./list1 > ./list2
awk '{ print $2 }' < ./list2 > ./list3
sed '/^ *$/d' < ./list3 > ./list4

input list

Code: Select all

LP2_AbiWord-2.8.6.sfs
LP2_AFI-1.0.3.sfs
...
LP2_Aqualung09b11.sfs
LP2_Ardour286.sfs
...
LP2_AsUnder_2.1.sfs
LP2_AutoAdjustPhoto_GUI-1.0.1.sfs
...
...
...

output list

Code: Select all

AbiWord
AFI
Aqualung
Ardour
AsUnder
AutoAdjustPhoto

Posted: **Thu 06 Dec 2012, 12:29**

Thanks to both of you, jamesbond and Geoffrey!

Will something like this help

Code: Select all

#!/bin/bash
sed -r 's/[_-0123456789]+/ /g' ./list1 > ./list2
awk '{ print $2 }' < ./list2 > ./list3
sed '/^ *$/d' < ./list3 > ./list4

Yes, this is great help! Thanks.

It doesn't work on some special names like LP2_SAG_CAD-0.9.14.sfs, which returns SAG only instead of SAG_CAD. Never mind, I can refine these special cases manually.

It is much better than to refine the complete list (272 lines) manually (several times).

Thanks

(old)Puppy Linux Discussion Forum

How to use Bash to bulk-edit filenames?

How to use Bash to bulk-edit filenames?

Re: How to remove parts from filename?