Shell Programming and Scripting

hi,
I want to remove comments in somany files. coments started from /* and ends with */. so i wnt to remove words /* to */
how to remove words between /* and */ in several lines of file

EX :
cat haha.txt
"HI",/*hk*/ob1,raju,/*hjh*/boju,beeju
output should be : "HI",ob1,raju,boju,beeju

I tried like below way :

cat haha.txt | sed -e 's:/*.*/::'

but output is coming : boju,beeju

Please help me ...

Try this:

Code:

sed '
s@/\*[^/\*]*\*/@@g
\@/\*@{
  N
  s@ *\n@ @
  s@/\*[^/\*]*\*/@@g
}' haha.txt

But it won't remove comments, spaning more than two lines...

cat haha.txt | perl -e 'while(<>){ s/\/\*.*?\*\///g; print;}'

"HI",ob1,raju,boju,beeju

This is done with the assumption that all comments are in a single line. If comments span multiple lines then the codes need to be tweaked

Cheers

Hi Sapfeer,

I tried ur code in my machine. Its working fine . But i couldnt understand how it works. Im very new to shell scripting.. even i tried with tutorials.....
Please explain me
1. s@/\*[^/\*]*\*/@@g why [^/\*] this is required?
2. how it checks for multiple lines... becaz i saw in one of the tutorial...label is required to check multiple lines ...



Please help me...

Thanks in advance

Code:

awk 'BEGIN{ FS="*/"}
{  
    for(i=1;i<=NF;i++){
        gsub(/\/\*.*/,"",$i)
    }
}1' file

spc432
Sure!

Lets start from the first line of the script.

Code:

s@/\*[^/\*]*\*/@@g

This removes comments on one line. '[^/\*]' needs to find all symbols, but '/' '*' to the closing comment. Actually, I've missed that in comments could be symbols '/' and '*'. It won't work in this circumstances. '[^/\*]' should be removed...

Code:

\@/\*@{

This line matches opening comment and start section, which is applied to all line, containing '/*' symbols.

Code:

N

This command reads next line and puts it in pattern space.

Code:

s@ *\n@

This line is not necessary here

Code:

@   s@/\*[^/\*]*\*/@@g}

This substitutes all symbols between '/*' and '*/' to ''.

And it doesn't need label to check two lines. For checking in more than two it needs.

So, final script should looks like this

Code:

s@/\*.*\*/@@g
\@/\*@{
  N
  s@/\*.*\*/@@g
}

And finally, try to find book Sed & Awk - it explains all sed stuff

Code:

echo '"HI",/*hk*/ob1,raju,/*hjh*/boju,beeju' | sed 's/\/\*[^*]*\*\///g'