From 0da58ad502ef3315f70369b0eacf8914b4466578 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?V=C3=A1clav=20Slav=C3=ADk?= <vslavik@fastmail.fm>
Date: Mon, 7 Apr 2003 22:11:53 +0000
Subject: [PATCH] backported correct CDATA and <script> handling

git-svn-id: https://svn.wxwidgets.org/svn/wx/wxWidgets/branches/WX_2_4_BRANCH@20065 c3d73ce0-8a6f-49c7-b76d-6d57e0e08775
---
 src/html/htmlpars.cpp | 12 +++++++++++
 src/html/htmltag.cpp  | 47 +++++++++++++++++++++++++++++++++++++++++++
 src/html/m_layout.cpp |  9 +++++++++
 3 files changed, 68 insertions(+)

diff --git a/src/html/htmlpars.cpp b/src/html/htmlpars.cpp
index 9ed716fe81..ac2fc44880 100644
--- a/src/html/htmlpars.cpp
+++ b/src/html/htmlpars.cpp
@@ -127,6 +127,8 @@ void wxHtmlParser::CreateDOMTree()
     m_CurTextPiece = 0;
 }
 
+extern bool wxIsCDATAElement(const wxChar *tag);
+
 void wxHtmlParser::CreateDOMSubTree(wxHtmlTag *cur,
                                     int begin_pos, int end_pos,
                                     wxHtmlTagsCache *cache)
@@ -137,6 +139,15 @@ void wxHtmlParser::CreateDOMSubTree(wxHtmlTag *cur,
     int i = begin_pos;
     int textBeginning = begin_pos;
 
+    // If the tag contains CDATA text, we include the text between beginning
+    // and ending tag verbosely. Setting i=end_pos will skip to the very
+    // end of this function where text piece is added, bypassing any child
+    // tags parsing (CDATA element can't have child elements by definition):
+    if (cur != NULL && wxIsCDATAElement(cur->GetName().c_str()))
+    {
+        i = end_pos;
+    }
+
     while (i < end_pos)
     {
         c = m_Source.GetChar(i);
@@ -209,6 +220,7 @@ void wxHtmlParser::CreateDOMSubTree(wxHtmlTag *cur,
                 }
                 else
                     i = chd->GetBeginPos();
+                
                 textBeginning = i;
             }
 
diff --git a/src/html/htmltag.cpp b/src/html/htmltag.cpp
index 47ac16a2e4..0b0b894f5b 100644
--- a/src/html/htmltag.cpp
+++ b/src/html/htmltag.cpp
@@ -57,6 +57,12 @@ IMPLEMENT_CLASS(wxHtmlTagsCache,wxObject)
 
 #define CACHE_INCREMENT  64
 
+bool wxIsCDATAElement(const wxChar *tag)
+{
+    return (wxStrcmp(tag, _T("SCRIPT")) == 0) ||
+           (wxStrcmp(tag, _T("STYLE")) == 0);
+}
+
 wxHtmlTagsCache::wxHtmlTagsCache(const wxString& source)
 {
     const wxChar *src = source.c_str();
@@ -108,6 +114,47 @@ wxHtmlTagsCache::wxHtmlTagsCache(const wxString& source)
             else
             {
                 m_Cache[tg].End1 = m_Cache[tg].End2 = -1;
+
+                if (wxIsCDATAElement(tagBuffer))
+                {
+                    // find next matching tag
+                    int tag_len = wxStrlen(tagBuffer);
+                    while (pos < lng)
+                    {
+                        // find the ending tag
+                        while (pos + 1 < lng &&
+                               (src[pos] != '<' || src[pos+1] != '/'))
+                            ++pos;
+                        if (src[pos] == '<')
+                            ++pos;
+                        
+                        // see if it matches
+                        int match_pos = 0;
+                        while (pos < lng && match_pos < tag_len && src[pos] != '>' && src[pos] != '<') {
+                            if (wxToupper(src[pos]) == tagBuffer[match_pos]) {
+                                ++match_pos;
+                            }  
+                            else if (src[pos] == wxT(' ') || src[pos] == wxT('\n') ||
+                                src[pos] == wxT('\r') || src[pos] == wxT('\t')) {
+                                // need to skip over these
+                            }
+                            else {
+                                match_pos = 0;
+                            }
+                            ++pos;
+                        }
+
+                        // found a match
+                        if (match_pos == tag_len) {
+                            pos = pos - tag_len - 3;
+                            stpos = pos;
+                            break;
+                        }
+                        else {
+                            ++pos;
+                        }
+                    }
+                }
             }
         }
 
diff --git a/src/html/m_layout.cpp b/src/html/m_layout.cpp
index aeb0478d49..e042da6e6a 100644
--- a/src/html/m_layout.cpp
+++ b/src/html/m_layout.cpp
@@ -224,6 +224,14 @@ TAG_HANDLER_END(BLOCKQUOTE)
 
 
 
+// Tag handler for tags that we have to ignore, otherwise non-text data
+// would show up as text:
+TAG_HANDLER_BEGIN(DoNothing, "SCRIPT")
+    TAG_HANDLER_PROC(tag)
+    {
+        return true;
+    }
+TAG_HANDLER_END(DoNothing)
 
 
 
@@ -236,6 +244,7 @@ TAGS_MODULE_BEGIN(Layout)
     TAGS_MODULE_ADD(TITLE)
     TAGS_MODULE_ADD(BODY)
     TAGS_MODULE_ADD(BLOCKQUOTE)
+    TAGS_MODULE_ADD(DoNothing)
 
 TAGS_MODULE_END(Layout)