substring in shell

mystical dervish · 06-29-2007, 07:22 AM

hello
I'm trying to examin each word fom list in text file (one word per line)
if the condition is true it has to remove the fisrt two chars from accepted word
the problem is in // line = ${line:2:$len-1} which it had to get substring from the word

Code:

while read line; do 
len=${#line}
if $len > 3 && $len < 5; then
    echo $line    
    line = ${line:2:$len-1}
    echo $line;
done < wrdstxt

can someone help
thanks

Hobbletoe · 06-29-2007, 08:08 AM

What you want is the following. Note also that I took out the spaces around the '='. Bash doesn't like those. Since you want to do some math, you need to use the $(( )) construct in bash.

Code:

line=${line:2:$((${len}-1))}

Note that you could forgo the $len variable entirely.

Code:

line=${line:2:$((${#line}-1))}

Actually, since it looks like you want from the second character to the end of the line, you could just use

Code:

line=${line:2}

cfaj · 06-30-2007, 08:40 PM

Quote:

Originally Posted by mystical dervish

hello
I'm trying to examin each word fom list in text file (one word per line)
if the condition is true it has to remove the fisrt two chars from accepted word
the problem is in // line = ${line:2:$len-1} which it had to get substring from the word

Code:

while read line; do 
len=${#line}
if $len > 3 && $len < 5; then
    echo $line    
    line = ${line:2:$len-1}
    echo $line;
done < wrdstxt

There are several mistakes in that code. This does what you want using
only POSIX syntax:

Code:

while read line
do 
   len=${#line}
   ## if [ $len -gt 3 ] && [ $len -lt 5 ]  ##  Why?
   if [ $len -eq 4 ]
   then
      printf "%s\n" "$line"
      line=${line#??}
      printf "%s\n" "$line"
   fi
done < wrdstxt

Or:

Code:

while read line
do 
   len=${#line}
   case $len in
      4) printf "%s\n" "$line"
         line=${line#??}
         printf "%s\n" "$line"
         ;;
   esac
done < wrdstxt

mystical dervish · 07-03-2007, 05:09 AM

Quote:

Originally Posted by cfaj

There are several mistakes in that code. This does what you want using
only POSIX syntax:
Code:
   ## if [ $len -gt 3 ] && [ $len -lt 5 ]  ##  Why?

I wanted 3 and 5 to be changable through tow variables

thanks for your reply

mystical dervish · 07-09-2007, 07:24 AM

is there any problem in my code here

Code:

if [ grep -w "$line" /lexic.wlst] | [ grep -w "$line" /wordofcorpus/wrd.txt]; then

because I execute my script , it results out a problem refering to that line saying " [: missing `]'"

cfaj · 07-09-2007, 08:53 AM

Quote:

Originally Posted by mystical dervish

is there any problem in my code here

Code:

if [ grep -w "$line" /lexic.wlst] | [ grep -w "$line" /wordofcorpus/wrd.txt]; then

because I execute my script , it results out a problem refering to that line saying " [: missing `]'"

You need spaces before the closing brackets:
Code:
if [ grep -w "$line" /lexic.wlst ] |
 [ grep -w "$line" /wordofcorpus/wrd.txt ]; then

mystical dervish · 07-09-2007, 09:40 AM

Quote:

Originally Posted by cfaj

You need spaces before the closing brackets:
Code:
if [ grep -w "$line" /lexic.wlst ] |
 [ grep -w "$line" /wordofcorpus/wrd.txt ]; then

I did so,but it gives another error,

Code:

if [ grep -w "$nline" /lexic.wlst ] | [ grep -w "$nline" /wordofcorpus/wrd.txt ]; then

[: too many arguments

cfaj · 07-09-2007, 09:50 AM

Quote:

Originally Posted by mystical dervish

I did so,but it gives another error,

Code:

if [ grep -w "$nline" /lexic.wlst ] | [ grep -w "$nline" /wordofcorpus/wrd.txt ]; then

[: too many arguments

Sorry; I shouldn't post before I have by first shot of caffeine.

That should be:
Code:
if grep -w "$nline" /lexic.wlst |
 grep -w "$nline" /wordofcorpus/wrd.txt ; then
But that doesn't make sense; you are giving the second grep two inputs, stdin and a file. Do one or the other.

Perhaps you mean:
Code:
if grep -w "$nline" /lexic.wlst ||
 grep -w "$nline" /wordofcorpus/wrd.txt ; then
Or (better):
Code:
if grep -w "$nline" /lexic.wlst /wordofcorpus/wrd.txt ; then

mystical dervish · 07-09-2007, 11:37 AM

Quote:

Originally Posted by cfaj

Sorry; I shouldn't post before I have by first shot of caffeine.

That should be:
Code:
if grep -w "$nline" /lexic.wlst |
 grep -w "$nline" /wordofcorpus/wrd.txt ; then
But that doesn't make sense; you are giving the second grep two inputs, stdin and a file. Do one or the other.

Perhaps you mean:
Code:
if grep -w "$nline" /lexic.wlst ||
 grep -w "$nline" /wordofcorpus/wrd.txt ; then
Or (better):
Code:
if grep -w "$nline" /lexic.wlst /wordofcorpus/wrd.txt ; then

thanx a lot and merci bien

ghostdog74 · 07-09-2007, 08:22 PM

Quote:

Originally Posted by mystical dervish

hello
I'm trying to examin each word fom list in text file (one word per line)
if the condition is true it has to remove the fisrt two chars from accepted word
the problem is in // line = ${line:2:$len-1} which it had to get substring from the word

Code:

while read line; do 
len=${#line}
if $len > 3 && $len < 5; then
    echo $line    
    line = ${line:2:$len-1}
    echo $line;
done < wrdstxt

can someone help
thanks

in awk:

Code:

awk 'length > 3 && length<5{
       print substr($0,3)
}' "file"