Hi,
I'm having some trouble reading a file that was 'cat' through a while loop. Can anyone suggest alternatives? what i do is

cat filename|grep [23]*.stuff
while read line
do
echo $line
... and other commands
done

The cat,grep line seems to work correctly, but the script hangs when i add in the while loop. echo line deosn't seem to work either.

I've been fiddling around, and sometimes I get a messagefor 'a broken pipe' and a syntax error, unexpected token near done? Help please! Been trying for hours. Thanks!

cat filename|grep [23]*.stuff
while read line
do
echo $line
... and other commands
done


you are just missing "|" at the end of grep, hence no input is going to while loop...

cat filename|grep [23]*.stuff |

moreover, grep command also don't look proper..
can you tell me what you are trying to grep actually if you are not getting the expected output

Hi,
Thanks for replying. That wa at least part of my problem as so obvious!!
filename contains a list of file names with different extensions. i just want files with a particular extension, 'stuff' in the example to do processing on. itmust have been looping infinitely.

thanks heaps. do you have suggestions for comparing the contents of two very similar files, when i need to be able to output a true/false value against a particular line? it's like an AND operation, with 2 different sets, not the same size.

is bash enough, or would perl be better? Thans again for helping.

Yes you can use perl, that would be better than bash..

But if there is any extra line in middle on any one file, then your script will show that all lines after that extra line are not matching though the data might match, because you are comparing data against exact line numbers. You need to put extra efforts here to match this kind of stuff like what diff unix command does, diff unix command won't match based on the line numbers...

Why can't you use unix diff command ?? here is an example


$ more datafile
abcdefghijklmnopqrstuvwxyz.
123456789009876543211234567
This line to be removed.
zyxwvutsrqponmlkjihgfedcba.

$ more datafile1
abcdefghijklmnopqrstuvwxyz.
1234567890098765432112345672
hello how are you
This line to be removed.
zyxwvutsrqponmlkjihgfedcba.4

$ $ diff datafile datafile1
2c2,3
< 123456789009876543211234567
---
> 1234567890098765432112345672
> hello how are you
4c5
< zyxwvutsrqponmlkjihgfedcba.
---
> zyxwvutsrqponmlkjihgfedcba.4



in the above diff, you can see (highlighted in bold)

<number>c<number>,<number>.......

i.e first number before character "c" is the line number from first file and numbers after "c" are numbers from 2nd file in diff command.

2c2,3 mean line number 2 in datafile is different from line numbers 2 and 3 in datafile1. Here you can see that line 3 in datafile1 is an extra line that is completly missing in datafile, so diff ignores this line and starts comparing datafile line 3 against datafile1 line 4.... not line 3 in datafile against line 3 in datafile 4.

You can actually egrep "^[0-9]*c[0-9]*[,0-9]*" or egrep -v "<|>|-" from the diff output, to get the lines which differ and the make use of it... based on the line numbers you can set the true or false..

not sure there is a simple way.. may be i can take a look if i find time