Whitespace Ignored when reading variables

[mjm22]

Hi

I use the below commands in a script to extract the header line from a file, remove the first couple of zeros and then write the value to another file.

hdrLine=`head -1 $REF_FILE`
echo $hdrLine | sed s/"00"//g | read COMP_LINE
echo $COMP_LINE >> $COMPLETE_FILE

The problem I have is that in the header record there are a number of fields of 8 characters that are left justified and padded with spaces (where the value is less the 8 chars). I need to preserve this fixed length but the above drops a lot of the whitespace.

I partly corrected this by quoting the variables in the echo command but the final field in the record still had a probelm and was being truncated. Eventually I used the following...

hdrLine=`head -1 $REF_FILE`
echo "'$hdrLine'" | sed s/"00"//g | read COMP_LINE
echo "$COMP_LINE" | sed s/"'"// >> $COMPLETE_FILE

This works but I am wondering if there is a better and less clunky way?

cheers

Mike

 

[aigles]

Try this :

head -1 $REF_FILE | sed s/"00"//g >> $COMPLETE_FILE

Jean Pierre.

 

[mjm22]

Ah well, actually I missed some lines out from the above post. In between the reading of the header line and the writing to the other file I also do some additional verification and validation on the contents of COMP_LINE before writing it to the file. Otherwise yes, I could have done it like that.

Incidentally there is a typo in the original post. The g switch of the sed command should be on the second instance of sed rather than the first one.

Mike

 

[aigles]

'read' removes leading and trailing spaces because space is a field delimitor.

Try this:

head -1 $REF_FILE | sed s/"00"// | IFS= read COMP_LINE
echo "$COMP_LINE" >> $COMPLETE_FILE

Jean Pierre.