Re: regex pattern match problem

It looks to me like you need to either preserve the original sentence and run a global substitution on it once for the sentence #1 output, and then run again with a non-global substitution for each of the other sentences, addressing each match one at a time; or else put a function in the replacement side of the substitution and do your processing within that function.

Blessings,

~Polyglot~

Comment on Re: regex pattern match problem

Replies are listed 'Best First'.
Re^2: regex pattern match problem by newbio (Beadle) on Aug 05, 2009 at 20:49 UTC
Yes, Polyglot, that's what I want. Could you please elaborate on your suggestion. How do I perform the non-global substitutions for their corresponding positions? It is possible that exactly the same pattern may get repeated in a sentence, so their positions need to be taken into account, possibly using substr function. Still, I cannot figure out how that works. I tried to do something like this but it does not work. $string=$line="Input sentence"; if ($line =~ /\\([^\]+)\\\s(kinase\|isoform\|protein\|peptide\|li +gand)\s\$\$([^\$]+)\$\$\s[\(\,]\s\\([^\]+)\\\s[\)\,]/) { print "yes"; while ($line =~ /(\\([^\]+)\\\s(kinase\|isoform\|protein\|pe +ptide\|ligand)\s\$\$([^\$]+)\$\$\s[\(\,]\s\\([^\]+)\\\s[\)\,])/g) + { $pattern=$1; $string =~ s/(\\([^\]+)\\\s(kinase\|isoform\|protein\|pe +ptide\|ligand)\s\$\$([^\$]+)\$\$\s[\(\,]\s\\([^\]+)\\\s[\)\,])/pq +rstuv/; $string =~ s/\\//g; $string =~ s/\$\$//g; $string =~ s/p +qrstuv/$pattern/; print WF "$string\n"; $string=$line; } } [download]	[reply] [d/l]

Replies are listed 'Best First'.

Re^2: regex pattern match problem
by newbio (Beadle) on Aug 05, 2009 at 20:49 UTC

$string=$line="Input sentence";
    if ($line =~ /\*\*([^\*]+)\*\*\s(kinase|isoform|protein|peptide|li
+gand)\s\$\$([^\$]+)\$\$\s[\(\,]\s\*\*([^\*]+)\*\*\s[\)\,]/) {
        print "yes";
        while ($line =~ /(\*\*([^\*]+)\*\*\s(kinase|isoform|protein|pe
+ptide|ligand)\s\$\$([^\$]+)\$\$\s[\(\,]\s\*\*([^\*]+)\*\*\s[\)\,])/g)
+ {
            $pattern=$1;
            $string =~ s/(\*\*([^\*]+)\*\*\s(kinase|isoform|protein|pe
+ptide|ligand)\s\$\$([^\$]+)\$\$\s[\(\,]\s\*\*([^\*]+)\*\*\s[\)\,])/pq
+rstuv/;
            $string =~ s/\*\*//g; $string =~ s/\$\$//g; $string =~ s/p
+qrstuv/$pattern/;
            print WF "$string\n";
            $string=$line;
        }
    }
[download]

[reply]
[d/l]