script for removing certain text

[Anonymous Monk]

Dear Monks, I would like to remove certain section of text from a file specified by begin and end keyswords. Does anyone of you have a perl script which could do this (given begin and end word). Thanks very much J.

[thpfft]

The good news is that perl is exactly the right way to do what you describe, and the task you describe is a good way to start learning perl.

The bad news: even if people were inclined to just provide the solution to your problem, without the very small amount of perl knowledge you would need to start solving it yourself, the chances of applying someone else's solution successfully are very slim.

So you'll have to do a bit more work, i'm afraid. but we can start things off. the sequence of events you need, at its simplest:

• read a file into memory
• remove whatever falls between your markers
• write what remains of the file back to disk

So you need to learn about file input and output. The s/// operator will probably be the way to remove those parts you don't want, and you'll need to know a little about string matching and regular expressions to make that work. Then you'll just print what remains back to the same file (or another one).

If you need to learn about the basics of perl and the structure of a simple script, you really need the camel book. Failing that, there are tutorials here and elsewhere.

[dooberwah]

If I understand your question correctly a regular-expression something like this would work:

s/hello.*goodbye//g;

Where the begin word is "hello" and the end word is "goodbye". If you in *NIX (this might work in Windoze also but I havn't tested it) you could give this as a command line argument using the -p -i and -e switchs:

perl -pi -e 's/hello.*goodbye//g' [FILENAME]

-Ben Jacobs (dooberwah)
one thing i can tell you is you got to be free
http://dooberwah.perlmonk.org

[boo_radley]

perlmonks is more oriented to technical discussion of perl, and not entirely as a trading post for code samples or entire scripts... You may find something useful in snippets or craft, but I'd recommend going to a site geared towards being a repository, such as the perl archive. Good luck.

[bikeNomad]

If you want to remove the beginning and end lines as well, you can do this as a one-liner, editing the fields in place:

perl -nie "print unless /begin/../end/" file1 file2

Of course, the details of quoting, etc. are up to your shell (like for instance, you'd use " instead of ' under Windows shells).

If you want backup files to be made, do this:

perl -ni.bak -e "print unless /begin/../end/" file1 file2

This will make backup files that are named file1.bak and so on.

[Anonymous Monk]

Hi Nomad, I tried your command, but it says "Can't open perl script "print unless /specify/../endspecify/": No such file or directory"..... Any guess. Thanks J

[thraxil]

# $text is the main text you want to process
# $begin is the begin keyword
# $end is the end keyword

$text =~ s/\Q$begin\E.*\Q$end\E//s;

the '\Q' and '\E' around '$begin' and '$end' quote them so that even if they contain regexp characters, they won't do unexpected things. see the perlre manpage. of course, i'm sure someone will point out that dotstar is bad, but the question was pretty vague.

anders pearson