/////////////////////////////////////////////////////////////////////////////
-// Name: htmltag.cpp
+// Name: src/html/htmltag.cpp
// Purpose: wxHtmlTag class (represents single tag)
// Author: Vaclav Slavik
// RCS-ID: $Id$
// Copyright: (c) 1999 Vaclav Slavik
-// Licence: wxWindows Licence
+// Licence: wxWindows licence
/////////////////////////////////////////////////////////////////////////////
+#include "wx/wxprec.h"
-#ifdef __GNUG__
-#pragma implementation
+#ifdef __BORLANDC__
+ #pragma hdrstop
#endif
-#include "wx/wxprec.h"
-
-#include "wx/defs.h"
#if wxUSE_HTML
-#ifdef __BORDLANDC__
-#pragma hdrstop
-#endif
+#include "wx/html/htmltag.h"
-#ifndef WXPRECOMP
+#ifndef WX_PRECOMP
+ #include "wx/colour.h"
+ #include "wx/wxcrtvararg.h"
#endif
-#include "wx/html/htmltag.h"
#include "wx/html/htmlpars.h"
-#include "wx/colour.h"
#include <stdio.h> // for vsscanf
#include <stdarg.h>
-
//-----------------------------------------------------------------------------
// wxHtmlTagsCache
//-----------------------------------------------------------------------------
#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();
- int tg, stpos;
- int lng = source.Length();
+ int lng = source.length();
wxChar tagBuffer[256];
m_Cache = NULL;
m_CacheSize = 0;
m_CachePos = 0;
- int pos = 0;
- while (pos < lng)
+ for ( int pos = 0; pos < lng; pos++ )
{
if (src[pos] == wxT('<')) // tag found:
{
+ // don't cache comment tags
+ wxString::const_iterator iter = source.begin() + pos;
+ if ( wxHtmlParser::SkipCommentTag(iter, source.end()) )
+ {
+ pos = iter - source.begin();
+ continue;
+ }
+
if (m_CacheSize % CACHE_INCREMENT == 0)
m_Cache = (wxHtmlCacheItem*) realloc(m_Cache, (m_CacheSize + CACHE_INCREMENT) * sizeof(wxHtmlCacheItem));
- tg = m_CacheSize++;
- m_Cache[tg].Key = stpos = pos++;
+ int tg = m_CacheSize++;
+ int stpos = pos++;
+ m_Cache[tg].Key = stpos;
- size_t i;
+ int i;
for ( i = 0;
- pos < lng && i < WXSIZEOF(tagBuffer) - 1 &&
+ pos < lng && i < (int)WXSIZEOF(tagBuffer) - 1 &&
src[pos] != wxT('>') && !wxIsspace(src[pos]);
i++, pos++ )
{
- tagBuffer[i] = wxToupper(src[pos]);
+ tagBuffer[i] = (wxChar)wxToupper(src[pos]);
}
tagBuffer[i] = _T('\0');
else
{
m_Cache[tg].End1 = m_Cache[tg].End2 = -1;
+
+ if (wxIsCDATAElement(tagBuffer))
+ {
+ // store the orig pos in case we are missing the closing
+ // tag (see below)
+ wxInt32 old_pos = pos;
+ bool foundCloseTag = false;
+
+ // 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] != '<') {
+ // cast to wxChar needed to suppress warning in
+ // Unicode build
+ if ((wxChar)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;
+ foundCloseTag = true;
+ break;
+ }
+ else // keep looking for the closing tag
+ {
+ ++pos;
+ }
+ }
+ if (!foundCloseTag)
+ {
+ // we didn't find closing tag; this means the markup
+ // is incorrect and the best thing we can do is to
+ // ignore the unclosed tag and continue parsing as if
+ // it didn't exist:
+ pos = old_pos;
+ }
+ }
}
}
-
- pos++;
}
// ok, we're done, now we'll free .Name members of cache - we don't need it anymore:
int delta = (at < m_Cache[m_CachePos].Key) ? -1 : 1;
do
{
+ if ( m_CachePos < 0 || m_CachePos == m_CacheSize )
+ {
+ // something is very wrong with HTML, give up by returning an
+ // impossibly large value which is going to be ignored by the
+ // caller
+ *end1 =
+ *end2 = INT_MAX;
+ return;
+ }
+
m_CachePos += delta;
}
while (m_Cache[m_CachePos].Key != at);
}
#undef IS_WHITE
- }
- m_Begin = i;
+ }
+ m_Begin = i;
- cache->QueryTag(pos, &m_End1, &m_End2);
- if (m_End1 > end_pos) m_End1 = end_pos;
- if (m_End2 > end_pos) m_End2 = end_pos;
+ cache->QueryTag(pos, &m_End1, &m_End2);
+ if (m_End1 > end_pos) m_End1 = end_pos;
+ if (m_End2 > end_pos) m_End2 = end_pos;
}
wxHtmlTag::~wxHtmlTag()
bool wxHtmlTag::HasParam(const wxString& par) const
{
- return (m_ParamNames.Index(par, FALSE) != wxNOT_FOUND);
+ return (m_ParamNames.Index(par, false) != wxNOT_FOUND);
}
wxString wxHtmlTag::GetParam(const wxString& par, bool with_commas) const
{
- int index = m_ParamNames.Index(par, FALSE);
+ int index = m_ParamNames.Index(par, false);
if (index == wxNOT_FOUND)
return wxEmptyString;
if (with_commas)
}
int wxHtmlTag::ScanParam(const wxString& par,
- const wxChar *format,
+ const char *format,
+ void *param) const
+{
+ wxString parval = GetParam(par);
+ return wxSscanf(parval, format, param);
+}
+
+int wxHtmlTag::ScanParam(const wxString& par,
+ const wchar_t *format,
void *param) const
{
wxString parval = GetParam(par);
bool wxHtmlTag::GetParamAsColour(const wxString& par, wxColour *clr) const
{
+ wxCHECK_MSG( clr, false, _T("invalid colour argument") );
+
wxString str = GetParam(par);
- if (str.IsEmpty()) return FALSE;
- if (str.GetChar(0) == wxT('#'))
+ // handle colours defined in HTML 4.0 first:
+ if (str.length() > 1 && str[0] != _T('#'))
{
- unsigned long tmp;
- if (ScanParam(par, wxT("#%lX"), &tmp) != 1)
- return FALSE;
- *clr = wxColour((unsigned char)((tmp & 0xFF0000) >> 16),
- (unsigned char)((tmp & 0x00FF00) >> 8),
- (unsigned char)(tmp & 0x0000FF));
- return TRUE;
- }
- else
- {
- // Handle colours defined in HTML 4.0:
- #define HTML_COLOUR(name,r,g,b) \
- if (str.IsSameAs(wxT(name), FALSE)) \
- { *clr = wxColour(r,g,b); return TRUE; }
+ #define HTML_COLOUR(name, r, g, b) \
+ if (str.IsSameAs(wxT(name), false)) \
+ { clr->Set(r, g, b); return true; }
HTML_COLOUR("black", 0x00,0x00,0x00)
HTML_COLOUR("silver", 0xC0,0xC0,0xC0)
HTML_COLOUR("gray", 0x80,0x80,0x80)
#undef HTML_COLOUR
}
- return FALSE;
+ // then try to parse #rrggbb representations or set from other well
+ // known names (note that this doesn't strictly conform to HTML spec,
+ // but it doesn't do real harm -- but it *must* be done after the standard
+ // colors are handled above):
+ if (clr->Set(str))
+ return true;
+
+ return false;
}
bool wxHtmlTag::GetParamAsInt(const wxString& par, int *clr) const
{
- if (!HasParam(par)) return FALSE;
+ if (!HasParam(par)) return false;
long i;
bool succ = GetParam(par).ToLong(&i);
*clr = (int)i;
wxString wxHtmlTag::GetAllParams() const
{
- // VS: this function is for backward compatiblity only,
+ // VS: this function is for backward compatibility only,
// never used by wxHTML
wxString s;
size_t cnt = m_ParamNames.GetCount();