getting things synced up for a release

[catagits/HTML-Zoom.git] / lib / HTML / Zoom / Parser / BuiltIn.pm

diff --git a/lib/HTML/Zoom/Parser/BuiltIn.pm b/lib/HTML/Zoom/Parser/BuiltIn.pm
index cfe21de..1e87919 100644 (file)
--- a/lib/HTML/Zoom/Parser/BuiltIn.pm
+++ b/lib/HTML/Zoom/Parser/BuiltIn.pm
@@ -1,7 +1,6 @@
 package HTML::Zoom::Parser::BuiltIn;
 
-use strict;
-use warnings FATAL => 'all';
+use strictures 1;
 use base qw(HTML::Zoom::SubObject);
 
 sub html_to_events {
@@ -17,14 +16,27 @@ sub html_to_stream {
               ->stream_from_array(@{$self->html_to_events($text)});
 }
 
+# DO NOT BE AFRAID.
+#
+# Well, ok. Be afraid. A little. But this is lexing HTML with a regexp,
+# not really parsing (since the structure nesting isn't handled here) so
+# it's relatively not dangerous.
+#
+# Certainly it's not really any more or any less heinous than anything else
+# I could do in a handful of lines of pure perl.
+
 sub _hacky_tag_parser {
   my ($text, $handler) = @_;
+  $text =~ m{^([^<]*)}g;
+  if ( length $1 ) { # leading PCDATA
+      $handler->({ type => 'TEXT', raw => $1 });
+  }
   while (
     $text =~ m{
       (
         (?:[^<]*) < (?:
             ( / )? ( [^/!<>\s"'=]+ )
-            ( (?:"[^"]*"|'[^']*'|[^"'<>])+? )?
+            ( (?:"[^"]*"|'[^']*'|[^/"'<>])+? )?
         |   
             (!-- .*? -- | ![^\-] .*? )
         ) (\s*/\s*)? >