Corrected code that re-encodes the book title.

[wxWidgets.git] / src / html / htmlpars.cpp

diff --git a/src/html/htmlpars.cpp b/src/html/htmlpars.cpp
index a41e5358d4063f218e00f1089799ed5fcf8962ea..dad3f09a427a6e554b0da5dd4b74d618c27d99ae 100644 (file)
--- a/src/html/htmlpars.cpp
+++ b/src/html/htmlpars.cpp
@@ -38,7 +38,7 @@
 // DLL options compatibility check:
 WX_CHECK_BUILD_OPTIONS("wxHTML")
 
-const wxChar *wxTRACE_HTML_DEBUG = _T("htmldebug");
+const wxChar *wxTRACE_HTML_DEBUG = wxT("htmldebug");
 
 //-----------------------------------------------------------------------------
 // wxHtmlParser helpers
@@ -227,7 +227,7 @@ void wxHtmlParser::CreateDOMSubTree(wxHtmlTag *cur,
             else
             {
                 while (i < end_pos && *i != wxT('>')) ++i;
-                textBeginning = i+1;
+                textBeginning = i < end_pos ? i+1 : i;
             }
         }
         else ++i;
@@ -250,8 +250,7 @@ void wxHtmlParser::DestroyDOMTree()
     }
     m_Tags = m_CurTag = NULL;
 
-    delete m_TextPieces;
-    m_TextPieces = NULL;
+    wxDELETE(m_TextPieces);
 }
 
 void wxHtmlParser::DoParsing()
@@ -433,7 +432,7 @@ void wxHtmlTagHandler::ParseInnerSource(const wxString& source)
 IMPLEMENT_DYNAMIC_CLASS(wxHtmlEntitiesParser,wxObject)
 
 wxHtmlEntitiesParser::wxHtmlEntitiesParser()
-#if wxUSE_WCHAR_T && !wxUSE_UNICODE
+#if !wxUSE_UNICODE
     : m_conv(NULL), m_encoding(wxFONTENCODING_SYSTEM)
 #endif
 {
@@ -441,14 +440,14 @@ wxHtmlEntitiesParser::wxHtmlEntitiesParser()
 
 wxHtmlEntitiesParser::~wxHtmlEntitiesParser()
 {
-#if wxUSE_WCHAR_T && !wxUSE_UNICODE
+#if !wxUSE_UNICODE
     delete m_conv;
 #endif
 }
 
+#if !wxUSE_UNICODE
 void wxHtmlEntitiesParser::SetEncoding(wxFontEncoding encoding)
 {
-#if wxUSE_WCHAR_T && !wxUSE_UNICODE
     if (encoding == m_encoding)
         return;
 
@@ -459,10 +458,8 @@ void wxHtmlEntitiesParser::SetEncoding(wxFontEncoding encoding)
         m_conv = NULL;
     else
         m_conv = new wxCSConv(wxFontMapper::GetEncodingName(m_encoding));
-#else
-    (void) encoding;
-#endif
 }
+#endif // !wxUSE_UNICODE
 
 wxString wxHtmlEntitiesParser::Parse(const wxString& input) const
 {
@@ -523,7 +520,6 @@ wxString wxHtmlEntitiesParser::Parse(const wxString& input) const
 #if !wxUSE_UNICODE
 wxChar wxHtmlEntitiesParser::GetCharForCode(unsigned code) const
 {
-#if wxUSE_WCHAR_T
     char buf[2];
     wchar_t wbuf[2];
     wbuf[0] = (wchar_t)code;
@@ -532,9 +528,6 @@ wxChar wxHtmlEntitiesParser::GetCharForCode(unsigned code) const
     if (conv->WC2MB(buf, wbuf, 2) == (size_t)-1)
         return '?';
     return buf[0];
-#else
-    return (code < 256) ? (wxChar)code : '?';
-#endif
 }
 #endif
 
@@ -557,6 +550,9 @@ wxChar wxHtmlEntitiesParser::GetEntityChar(const wxString& entity) const
 {
     unsigned code = 0;
 
+    if (entity.empty())
+      return 0; // invalid entity reference
+
     if (entity[0] == wxT('#'))
     {
         // NB: parsed value is a number, so it's OK to use wx_str(), internal
@@ -652,6 +648,7 @@ wxChar wxHtmlEntitiesParser::GetEntityChar(const wxString& entity) const
             ENTITY("amp", 38),
             ENTITY("and", 8743),
             ENTITY("ang", 8736),
+            ENTITY("apos", 39),
             ENTITY("aring", 229),
             ENTITY("asymp", 8776),
             ENTITY("atilde", 227),
@@ -843,9 +840,10 @@ wxChar wxHtmlEntitiesParser::GetEntityChar(const wxString& entity) const
             while (substitutions[substitutions_cnt].code != 0)
                 substitutions_cnt++;
 
-        wxHtmlEntityInfo *info = NULL;
+        wxHtmlEntityInfo *info;
 #ifdef __WXWINCE__
         // bsearch crashes under WinCE for some reason
+        info = NULL;
         size_t i;
         for (i = 0; i < substitutions_cnt; i++)
         {
@@ -871,10 +869,14 @@ wxChar wxHtmlEntitiesParser::GetEntityChar(const wxString& entity) const
         return GetCharForCode(code);
 }
 
-wxFSFile *wxHtmlParser::OpenURL(wxHtmlURLType WXUNUSED(type),
+wxFSFile *wxHtmlParser::OpenURL(wxHtmlURLType type,
                                 const wxString& url) const
 {
-    return m_FS ? m_FS->OpenFile(url) : NULL;
+    int flags = wxFS_READ;
+    if (type == wxHTML_URL_IMAGE)
+        flags |= wxFS_SEEKABLE;
+
+    return m_FS ? m_FS->OpenFile(url, flags) : NULL;
 
 }
 
@@ -893,7 +895,7 @@ public:
 protected:
     virtual void AddText(const wxString& WXUNUSED(txt)) {}
 
-    DECLARE_NO_COPY_CLASS(wxMetaTagParser)
+    wxDECLARE_NO_COPY_CLASS(wxMetaTagParser);
 };
 
 class wxMetaTagHandler : public wxHtmlTagHandler
@@ -906,23 +908,23 @@ public:
 private:
     wxString *m_retval;
 
-    DECLARE_NO_COPY_CLASS(wxMetaTagHandler)
+    wxDECLARE_NO_COPY_CLASS(wxMetaTagHandler);
 };
 
 bool wxMetaTagHandler::HandleTag(const wxHtmlTag& tag)
 {
-    if (tag.GetName() == _T("BODY"))
+    if (tag.GetName() == wxT("BODY"))
     {
         m_Parser->StopParsing();
         return false;
     }
 
-    if (tag.HasParam(_T("HTTP-EQUIV")) &&
-        tag.GetParam(_T("HTTP-EQUIV")).IsSameAs(_T("Content-Type"), false) &&
-        tag.HasParam(_T("CONTENT")))
+    if (tag.HasParam(wxT("HTTP-EQUIV")) &&
+        tag.GetParam(wxT("HTTP-EQUIV")).IsSameAs(wxT("Content-Type"), false) &&
+        tag.HasParam(wxT("CONTENT")))
     {
-        wxString content = tag.GetParam(_T("CONTENT")).Lower();
-        if (content.Left(19) == _T("text/html; charset="))
+        wxString content = tag.GetParam(wxT("CONTENT")).Lower();
+        if (content.Left(19) == wxT("text/html; charset="))
         {
             *m_retval = content.Mid(19);
             m_Parser->StopParsing();
@@ -951,16 +953,18 @@ bool
 wxHtmlParser::SkipCommentTag(wxString::const_iterator& start,
                              wxString::const_iterator end)
 {
-    wxASSERT_MSG( *start == '<', _T("should be called on the tag start") );
+    wxASSERT_MSG( *start == '<', wxT("should be called on the tag start") );
 
     wxString::const_iterator p = start;
 
-    // comments begin with "<!--" in HTML 4.0
-    if ( p > end - 3 || *++p != '!' || *++p != '-' || *++p != '-' )
-    {
-        // not a comment at all
-        return false;
-    }
+    // Comments begin with "<!--" in HTML 4.0; anything shorter or not containing
+    // these characters is not a comment and we're not going to skip it.
+    if ( ++p == end || *p != '!' )
+      return false;
+    if ( ++p == end || *p != '-' )
+      return false;
+    if ( ++p == end || *p != '-' )
+      return false;
 
     // skip the start of the comment tag in any case, if we don't find the
     // closing tag we should ignore broken markup