1 #!/usr/bin/perl
   2 
   3 
   4 # Converts a "temporary TEI" segmentation (output of Anotatornia) to target TEI. Minor changes, mostly xlink:href->corresp.
   5 
   6 # Usage:
   7 #	perl modify-tei-segmentation.pl <input file> <output file>
   8 #
   9 
  10 open INTEXT, "<$ARGV[0]";
  11 open OUTTEXT, ">$ARGV[1]";
  12 
  13 
  14 while (<INTEXT>) {
  15 
  16     s#<teiCorpus.*>#<teiCorpus xmlns:xi="http://www.w3.org/2001/XInclude" xmlns="http://www.tei-c.org/ns/1.0" xmlns:nkjp="http://www.nkjp.pl/ns/1.0">#;
  17     s/<text xml:lang="pl" xml:id="segm_text">/<text xml:id="segm_text" xml:lang="pl">/;
  18     s/<p xlink:href=(".*?")/<p corresp=$1/;
  19     s/<nkjp:paren nkjp:rejected="true">/<nkjp:paren>/;
  20     
  21     if (/<seg (.*?)xml:id="(segm_.*?)"/) {
  22 	$segpars = $1;
  23 	$segid = $2; 
  24     } elsif (/<xi:include href="(.*?)\.xml" xpointer="(.*?)"\/>/) {
  25 	$segdesc="<seg corresp=\"$1.xml#$2\" $segpars"."xml:id=\"$segid\"\/>";
  26     } elsif (/(\s*)<!--.*?-->/) {
  27 	print OUTTEXT;
  28 	print OUTTEXT "$1$segdesc\n" unless $segdesc eq '';
  29     } elsif (/<\/s>/) {
  30 	$segdesc="";
  31 	print OUTTEXT;
  32     } elsif (/<\/seg>/) {
  33     } elsif (/^<\?o/) {
  34     } else {
  35 	print OUTTEXT;
  36     }
  37     
  38 }
  39 
  40 close INTEXT;
  41 close OUTTEXT;