Remove comma from last record in a csv 1

[epatton]

I have a comma-separated values text file, consisting of just numbers and commas:

192123623,192123628,192123635,192123637,192123637,
192123642,192123652,192123652,192123655,192123704,
192123714,192123714,192123722,192123728,192123628,
192123749,192123824,

I want to remove the comma from the last record only (after 192123824) because this is the end of the file.

Thanks for any help you can provide,

~ Eric.

 

[feherke]

Hi

This can be solved much easier with [tt]sed[/tt] :

sed '$s/,$//' /input/file

Tested with GNU [tt]sed[/tt], but should work with other versions too.

Feherke.

http://rootshell.be/~feherke/

 

[feherke]

Hi

With [tt]awk[/tt] is abit uglier because it can not tell if a line is last.

awk 'l{print l}{l=$0}END{sub(/,$/,"",l);print l}' /input/file

Feherke.

http://rootshell.be/~feherke/

 

[epatton]

Thanks! That works fine. I guess I could have avoided creating a csv file with a trailing comma in the first place if I knew of a way to convert a file like this

192123623
192123628
192123635
192123637
192123637
...

into a comma-delimited file with no comma after the last record.

 

[feherke]

Hi

1 column -> 5 columns separated with ,

awk '{printf (NR>1?",":"")(NR%5==1?"\n":"")$0}' /input/file

Feherke.

http://rootshell.be/~feherke/

 

[epatton]

Suppose that the number of comma-delimited records in the input file is variable (usually in the hundreds). How would your script work in a general case such as this?

Thanks for your feedback!

~ Eric.

 

[epatton]

I should also mention that the input file is just one long string of comma-separated numbers, dozens of values long; I placed 5 values on one line in the above example so that they would fit on the screen.

~ Eric.

 

[feherke]

Hi

Sorry, now I am confused regarding the number of input and output columns.

How would your script work in a general case such as this?

Your input was 1 value in 1 line, so the [tt]awk[/tt] counted the lines :

awk '{
 printf 
  (NR>1?",":"")      [gray]# for lines after line 1 put a comma[/gray]
  (NR%5==1?"\n":"")  [gray]# for each 5th line starting from 1 put a newline[/gray]
  $0                 [gray]# put the value[/gray]
}' /input/file

But that code has a small bug : puts a newline ( \n ) before the first line. Corrected :

awk '{printf (NR>1?","(NR%5==1?"\n":""):"")$0}' /input/file

Feherke.

http://rootshell.be/~feherke/