remove specific data from a line (Anonymous Monk)

remove specific data from a line

created: 2006-02-02 10:07:43

perlmonks

node
author

prlmnks

node
author

Hello, I need help with Perl problem that I'm having. I just start with Perl and don't have enough knowledge to fix problem quick.
I have file with lines in that look like this

myword myobject,option2,option3 word word word

The words delimited by comma after myobject are optional. There may or may not be these optional words in each line.
myword if present always is at begin of line and if there would always be follow by at least one space then another word (and then possible optional words).
I find the lines in the file that begin with myword like this


while (my $line=) {
     if $line =~ /^myword\s+/ {
           work on line;
     }
}

I want to strip "myword myobject,option1,option2" from line but not know how do this e.g. something like

   if $line =~ /^myword\s+/ {
     $line =~ s/^myword myobject,options//g
    }

I knows what first few characters of myword will be but not what myobject will be (except it's always a word).
How does I do this ? ciao

Re: remove specific data from a line

idle

created: 2006-02-02 10:20:30

perlmonks

prlmnks

So you wanna remove the string if it start with myword? Try this:

if $line =~ /^myword.*$/ {
     $line =~ s/$&//g
    }

Re: remove specific data from a line

idle

created: 2006-02-02 10:26:29

perlmonks

prlmnks

So you wanna remove the string if it start with myword? Try this:

if $line =~ /^myword.*$/ {
     $line =~ s/$&//g
    }

Updated. Oh I see you need to remove the line until the last option. Heres the code:

if $line =~ /(^myword.*option\d+\s+).*$/ {
     $line =~ s/$1//g
    }

Re^2: remove specific data from a line

Anonymous Monk

created: 2006-02-02 10:36:50

perlmonks

prlmnks

I not explain too good sorry. Let me try eloborate. String could look like this

abc123 qwerty,asdfg,zxcvb yuio jklh

or like this

abc987 qwerty yuio jklh

or this

abcefgh qwerty, asdfg yuio jklh

So I want to end up with same string from all examples above. It should end up like this

yuio jklh

Hope this make it bit clearer, sorry for confuse my english not so good

Re^3: remove specific data from a line

graff

created: 2006-02-02 22:17:22

perlmonks

prlmnks

Given those examples, there might be a few ways to do this -- either try to match from the start of the string:

s/abc\w+\s+\w+(?:, *\w+)\s+//;  # match and remove unwanted initial content

or else just look for what you want to keep at the end:

s/.*\s(\w+\s+\w+)$/$1/;  # match desired end content and remove everything before it

And of course, if you know in advance how those last two tokens are spelled, you could even use rindex() and substr():

$_ = substr( $_, rindex( $_, 'yuio jklh' ));

Re^4: remove specific data from a line

Anonymous Monk

created: 2006-02-03 05:06:38

perlmonks

prlmnks

Thank for answer but not quite what I look for. This s/abc\w+\s+\w+(?:, *\w+)\s+//; only work correct on last line of my example.
I'm not able to use second or third option you supply because I do not know what end of line hold only what beginning of line look like. See my reponse to Roy Johnson. Thanks

Re: remove specific data from a line

Roy Johnson

created: 2006-02-02 10:36:24

perlmonks

prlmnks

I think you want:

$line =~ s/^myword \S+//; # strip myword, a space, and any stretch of non-spaces

Caution: Contents may have been coded under pressure.

Re^2: remove specific data from a line

Anonymous Monk

created: 2006-02-03 04:54:38

perlmonks

prlmnks

This one nearly work but not handle this line

abcefgh qwerty, asdfg yuio jklh

It print out

asdfg yuio jklh

I had to add extra slash to your code like this

$line =~ s/^abc\w+\s+\S+(?:,\s*\S+)*//x;

otherwise got errors. I hope this was what you mean. I try to change your code to handle last line which don't leave just yuio jklh .
How do I fix this ?
Thank you for answer so far , it show me how much flexible Perl is

Re^3: remove specific data from a line

Roy Johnson

created: 2006-02-03 13:29:04

perlmonks

prlmnks

I forgot to substitute the first word back in.

$line =~ s/^(abc)\w+\s+\S+(?:,\s*\S+)*/$1/x;

should give you what you want. You were right about the missing slash.

Caution: Contents may have been coded under pressure.