Shell Programming and Scripting

Hi,

I have file as follows:

Input:
Output:
Required output:
Code:

Code:

 
tr '\n' '|' < ${TEMPDIR}/in.txt > ${TEMPDIR}/out.txt

Help is appreciated to fix code to remove the last pipe.

If you are running ksh, you could do the following:-

Code:

tmpvar=`tr '\n' '|' < ${TEMPDIR}/in.txt`
echo "${tmpvar%\|}" > $TEMPDIR/out.txt

That will chop off the last bit. There might be a neater way with awk though.



I hope that this helps,
Robin
Liverpool/Blackburn
UK

Using awk program:

Code:

awk ' {
        A[++r] = $0
}
END {
        for (i = 1; i <= r; i++) {
                printf i == r ? A[i] : (A[i] "|")
        }
        printf "\n"
} ' file

Try:

Code:

paste -sd \| in.txt

A much simpler awk approach:

Code:

awk '$1=$1' RS= OFS="|" file

@yoda, that would be problematic if there is any kind of whitespace other than the newlines. Even if you set FS to a newline, the RS= would compress multiple newlines and create multiple records.

As a work-around one could choose a character that will not occur in the text for example:

Code:

awk '{$1=$1}1' FS='\n' RS=§ OFS="|"

Note that there would need to be curly braces around $1=$1 otherwise it would not print in case of 0 or whitespace in $1

But this would leave a trailing pipe symbol after the last record, so that still would not work as desired...

Also there might be record length limitations..

Is that all there is to the input file? If so, doesn't seem like a real-world problem. What comes after those three lines, if anything?