From 464037b2c29b37ce95d82c5cdefef90a2d31d0c3 Mon Sep 17 00:00:00 2001
From: Yorhel <git@yorhel.nl>
Date: Fri, 14 Nov 2008 21:21:01 +0100
Subject: Rewrote & fixed bb2html()

The general concept and method is really the same, this version just
does a few things different to ensure correct parsing and HTML output.
---
 lib/VNDB/Func.pm | 168 +++++++++++++++++++++----------------------------------
 1 file changed, 65 insertions(+), 103 deletions(-)

(limited to 'lib/VNDB')

diff --git a/lib/VNDB/Func.pm b/lib/VNDB/Func.pm
index 92f31616..bc6cbc87 100644
--- a/lib/VNDB/Func.pm
+++ b/lib/VNDB/Func.pm
@@ -45,120 +45,82 @@ sub datestr {
 }
 
 
-# Parses BBCode-enhanced strings into the correct HTML variant with an optional maximum length
-sub bb2html { # input, length
-  return $_[2] || ' ' if !$_[0];  # No clue what this is for, but it is included from the previous verison for posterity...
-  my $raw = $_[0];
-  my $maxLength;
-  if (defined($_[1])) {$maxLength = $_[1]}
-  else {$maxLength = length $_[0]}
-
-  my $result = '';
-  my $length = 0;
-  my $inRaw = 0;
-  my $inSpoiler = 0;
-  my $inUrl = 0;
-
-  # Split the input string into segments
-  foreach (split /(\s|\[.+?\])/, $raw)
-  {
-    if (!defined($_)) {next}
-
-    if (!$inRaw)
-    {
-      # Cases for BBCode tags
-      if    ($_ eq '[raw]')      {$inRaw = 1; next}
-      elsif ($_ eq '[spoiler]')  {$inSpoiler = 1; next}
-      elsif ($_ eq '[/spoiler]') {$inSpoiler = 0; next}
-      elsif ($_ eq '[/url]')
-      {
-        $result .= '</a>';
-        $inUrl = 0;
+# Arguments: input, and optionally the maximum length
+# Parses:
+#  [url=..] [/url]
+#  [raw] .. [/raw]
+#  [spoiler] .. [/spoiler]
+#  v+,  v+.+
+#  http://../
+sub bb2html { 
+  my $raw = shift;
+  my $maxlength = shift;
+  $raw =~ s/\r//g;
+  return '' if !$raw && $raw != 0;
+
+  my($result, $length, @open) = ('', 0, 'first');
+
+  my $e = sub {
+    local $_ = shift;
+    tr/A-Za-z/N-ZA-Mn-za-m/ if !@_ && grep /spoiler/, @open;
+    s/&/&amp;/g;
+    s/>/&gt;/g;
+    s/</&lt;/g;
+    s/\n/<br \/>/g if !$maxlength;
+    s/\n/ /g if $maxlength;
+    return $_;
+  };
+
+  for (split /(\s|\n|\[[^\]]+\])/, $raw) {
+    next if !defined $_;
+
+    my $lit = $_;
+    if($open[$#open] ne 'raw') {
+      if    ($_ eq '[raw]')      { push @open, 'raw'; next }
+      elsif ($_ eq '[spoiler]')  { push @open, 'spoiler'; next }
+      elsif ($_ eq '[/spoiler]') { pop @open if $open[$#open] eq 'spoiler'; next }
+      elsif ($_ eq '[/url]')     {
+        if($open[$#open] eq 'url') {
+          $result .= '</a>';
+          pop @open;
+        }
         next;
-      }
-      # Process [url=.+] tags
-      if (s/\[url=((https?:\/\/|\/)[^\]>]+)\]/<a href="$1" rel="nofollow">/i)
-      {
+      } elsif(s{\[url=((https?://|/)[^\]>]+)\]}{<a href="$1" rel="nofollow">}i) {
         $result .= $_;
-        $inUrl = 1;
+        push @open, 'url';
         next;
-      }
-    }
-
-    my $lit = $_;   # Literal version of the segment to refence in case we link-ify the original
-
-    if ($_ eq '[/raw]')
-    {
-      # Special case for leaving raw mode
-      $inRaw = 0;
-    }
-    elsif ($_ =~ m/\n/)
-    {
-      # Parse line breaks
-      $result .= (defined($_[1])?'':'<br />');
-    }
-    elsif (!$inRaw && !$inUrl && s/(http|https):\/\/(.+[0-9a-zA-Z=\/])/<a href="$1:\/\/$2" rel="nofollow">/)
-    {
-      # Parse automatic links
-      $length += 4;
-      if ($length <= $maxLength)
-      {
-        $lit = 'link';
-
-        # ROT-13 of 'link'
-        $lit = 'yvax' if $inSpoiler;
-
-        $result .= $_ . $lit . '</a>';
-      }
-    }
-    elsif (!$inRaw && !$inUrl && (s/^(.*[^\w]|)([tdvpr][1-9][0-9]*)\.([1-9][0-9]*)([^\w].*|)$/"$1<a href=\"\/$2.$3\">". ($inSpoiler?rot13("$2.$3"):"$2.$3") ."<\/a>$4"/e ||
-        s/^(.*[^\w]|)([tduvpr][1-9][0-9]*)([^\w].*|)$/"$1<a href=\"\/$2\">". ($inSpoiler?rot13($2):$2) ."<\/a>$3"/e))
-    {
-      # Parse VNDBID
-      $length += length $lit;
-      if ($length <= $maxLength)
-      {
+      } elsif(!grep(/url/, @open) &&
+           s{(.*)(http|https)://(.+[0-9a-zA-Z=/])(.*)}
+            {$e->($1).qq|<a href="$2://|.$e->($3, 1).'" rel="nofollow">'.$e->('link').'</a>'.$e->($4)}e) {
+        $length += 4;
+        last if $maxlength && $length > $maxlength;
         $result .= $_;
-      }
-    }
-    else
-    {
-      # Normal text processing
-      $length += length $_;
-      if ($length <= $maxLength)
-      {
-        # ROT-13
-        tr/A-Za-z/N-ZA-Mn-za-m/ if $inSpoiler;
-
-        # Character escaping
-        s/\&/&amp;/g;
-        s/>/&gt;/g;
-        s/</&lt;/g;
-
+        next;
+      } elsif(!grep(/url/, @open) && (
+          s{^(.*[^\w]|)([tdvpr][1-9][0-9]*)\.([1-9][0-9]*)([^\w].*|)$}{$e->($1).qq|<a href="/$2.$3">$2.$3</a>|.$e->($4)}e ||
+          s{^(.*[^\w]|)([tduvpr][1-9][0-9]*)([^\w].*|)$}{$e->($1).qq|<a href="/$2">$2</a>|.$e->($3)}e)) {
+        $length += length $lit;
+        last if $maxlength && $length > $maxlength;
         $result .= $_;
+        next;
       }
-    }
-
-    # End if we've reached the maximum length/string end
-    if ($length >= $maxLength)
-    {
-      # Tidy up the end of the string
-      $result =~ s/\s+$//;
-      $result .= '...';
-      last;
-    }
+    } elsif($_ eq '[/raw]') {
+      pop @open if $open[$#open] eq 'raw';
+      next;
+    } 
+    
+    # normal text processing
+    $length += length $_;
+    last if $maxlength && $length > $maxlength;
+    $result .= $e->($_);
   }
 
-  # Close any un-terminated url tags
-  $result .= '</a>' if $inUrl;
+  $result .= '</a>' 
+    while((local $_ = pop @open) ne 'first');
+  $result .= '...' if $maxlength && $length > $maxlength;
 
   return $result;
 }
- 
-# Performs a ROT-13 cypher (used by bb2html)
-sub rot13 {
-  return tr/A-Za-z/N-ZA-Mn-za-m/;
-}
 
 
 1;
-- 
cgit v1.2.3