Hi!

I've got a file that is comma delimited, such as;


134,145,245,678,5888,10000,24323,


The problem I'm facing is I want to remove just the last comma such that the file retains all number and commas minus the last one. I am brain farting after GOOGLING all night .... I saw much reference to 'chomp' and 'chop' using PERL, but I'm trying to keep this very simple.

Is there a command in Linux that can be used or included in a simple script to remove this last character (the comma) from the file?

Any information or pointers are greatly appreciated!!!

-lc

try :

If comma is always the last character, you can try something like
Basically, it reverses each line, cuts from the 2nd character to the end of line and reversing it back.

HTH

Thank you both kindly. Each worked. Very happy. Now going to sleep!

-lc

The sed solution is preferable, since it's a single process, which is quicker to run than 3(!), and is more obvious (for maintenance purposes).

Hi,

I have following lines in my text file. File contains more than 20 million lines. like

/home/apache/ht/images/5/1/200904021159351.jpg
/home/apache/ht/images/5/1/20090401113951.jpg
/home/apache/ht/images/5/1/200904041728451.jpg
/home/apache/ht/images/9/5/2561195/2561195-20090402015241.xls
/home/apache/ht/images/attachments/9/5/2561195/2561195-20090407011002.xls
/home/apache/ht/images/attachments/9/5/2561195/2561195-20090406090748.xls


Now i want to remove filename from entire file. File should look like this


/home/apache/ht/images/5/1/
/home/apache/ht/images/5/1/
/home/apache/ht/images/5/1/
/home/apache/ht/images/9/5/2561195/
/home/apache/ht/images/attachments/9/5/2561195/
/home/apache/ht/images/attachments/9/5/2561195/


can any one let me know the command please. It would be highly appreciated.

you should start a new thread. anyway, you can use this

Hi ghost,

It works for me. Thanks for your kind help. Can you explain that syntax for me please.

awk -F"/" '{$NF=""}1' OFS="/" file

what is $NF and OFS ?


Another question i am running find command on the directory. I am getting list of files created or modified in last 24 hours through following command.

find testdir -mtime -1 -type f -print >> images.log

where testdir have 20 subdirectories and each subdirectory have more than 1 directories beneath. So my command took lot of time consuming resources extensively because it has to search in every directory and beneath. I have two questions in mind.

1) What should be the command to search for files created or modified in last 12 hours.
2) What Should be the less expensive command to search means can i search through timestamp for files in last 24 hours ? Like

find . -name -mtime mm hh date mm:ss type -f > test.txt


Thanks in advance.

-F"/" -> set field separator as "/"
NF -> number of field
$NF -> value of the last field
OFS -> output field separator.

all these are explained in the man page or awk. or you can take a look at effective awk.

Another solution using sed

Hi,

can you explain the command please

find . -name -mtime mm hh date mm:ss type -f

Let suppose how could i find today(Apr 10) files through this command ?

The sed command was just another solution to the previous issue. To find files more recent than a specific date, you can create a dummy file using the touch command with the -t option, e.g:
then use the test -newer in the find command:

Hi,

My question is find should not work on entire directory as it will take lot of time and increases I/O. Is there any command that directly work on month. Like find search files which are created or modified in month April
directly and escape other files without looking into them.

Nope. The find command must look at every single item under the search path to see if it matches the search criteria. If you have billions of files under the search path, it will look at them all. No chance to skip some check in my opinion.

would basename work
man basename

then maybe you could sed s/`the return of basename`/""/ in a loop ?