Re: [rtcweb] Syntax issues in draft-ietf-rtcweb-sdp

[Adam Roach <[e-mail]>]

On 12/3/17 2:22 PM, Harald Alvestrand wrote:
> Great to see attention on this!
>
> Nils Ohlmeier also wrote an extraction script
> (https://github.com/nils-ohlmeier/sdpextractor - I have suggested a PR
> that changes the file name format), do you want to publish yours too?
>

What I wrote was mostly done for the purpose of validating, so 
extraction is a very small part of it. It's also a bit fragile, in that 
it heavily relies on the exact format of the -08 XML source. With those 
caveats (and the warning that it's written in perl), the relevant 
portion of my script is:

#!/usr/bin/perl

my $source = 'draft-ietf-rtcweb-sdp-08.xml';

open(S, $source) || die $!;
$xml = join ('',<S>);
close (S);
$xml =~ s/<!--.*?-->//gos;

while ($xml =~ /<texttable anchor="[^"]*" 
title="([^"]*)">(.*?)<\/texttable>/gos) {
   my $title = $1;
   my $body = $2;
   my $column = 0;
   my $sdp = '';
   while ($body =~ /<c>(.*?)<\/c>/gos) {
     if ($column == 0){
       my $line = $1;
       $line =~ s/[ \t]*[\r\n]+[ \t]*/ /gos;
       if ($line !~ /\*\*/) {
         $sdp .= "$line\n";
       }
     }
     $column = 1 - $column;
   }
   &process($title, $sdp);
}

You'll have to write your own "process" subroutine to do whatever you 
want to happen with the extracted SDP.

/a