void wx28HtmlParser::CreateDOMSubTree(wx28HtmlTag *cur,
int begin_pos, int end_pos,
wx28HtmlTagsCache *cache)
{
if (end_pos <= begin_pos) return;
wxChar c;
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);
if (c == wxT('<'))
{
// add text to m_TextPieces:
if (i - textBeginning > 0)
m_TextPieces->Add(
wx28HtmlTextPiece(textBeginning, i - textBeginning));
// if it is a comment, skip it:
if (i < end_pos-6 && m_Source.GetChar(i+1) == wxT('!') &&
m_Source.GetChar(i+2) == wxT('-') &&
m_Source.GetChar(i+3) == wxT('-'))
{
// Comments begin with "<!--" and end with "--[ \t\r\n]*>"
// according to HTML 4.0
int dashes = 0;
i += 4;
while (i < end_pos)
{
c = m_Source.GetChar(i++);
if ((c == wxT(' ') || c == wxT('\n') ||
c == wxT('\r') || c == wxT('\t')) && dashes >= 2) {}
else if (c == wxT('>') && dashes >= 2)
{
textBeginning = i;
break;
}
else if (c == wxT('-'))
dashes++;
else
dashes = 0;
}
}
// add another tag to the tree:
else if (i < end_pos-1 && m_Source.GetChar(i+1) != wxT('/'))
{
wx28HtmlTag *chd;
if (cur)
chd = new wx28HtmlTag(cur, m_Source,
i, end_pos, cache, m_entitiesParser);
else
{
chd = new wx28HtmlTag(NULL, m_Source,
i, end_pos, cache, m_entitiesParser);
if (!m_Tags)
{
// if this is the first tag to be created make the root
// m_Tags point to it:
m_Tags = chd;
}
else
{
// if there is already a root tag add this tag as
// the last sibling:
chd->m_Prev = m_Tags->GetLastSibling();
chd->m_Prev->m_Next = chd;
}
}
if (chd->HasEnding())
{
CreateDOMSubTree(chd,
chd->GetBeginPos(), chd->GetEndPos1(),
cache);
i = chd->GetEndPos2();
}
else
i = chd->GetBeginPos();
textBeginning = i;
}
// ... or skip ending tag:
else
{
while (i < end_pos && m_Source.GetChar(i) != wxT('>')) i++;
textBeginning = i+1;
}
}
else i++;
}
// add remaining text to m_TextPieces:
if (end_pos - textBeginning > 0)
m_TextPieces->Add(
wx28HtmlTextPiece(textBeginning, end_pos - textBeginning));
}
void wxHtmlParser::CreateDOMSubTree(wxHtmlTag *cur,
const wxString::const_iterator& begin_pos,
const wxString::const_iterator& end_pos,
wxHtmlTagsCache *cache)
{
if (end_pos <= begin_pos)
return;
wxChar c;
wxString::const_iterator i = begin_pos;
wxString::const_iterator 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()))
{
i = end_pos;
}
while (i < end_pos)
{
c = *i;
if (c == wxT('<'))
{
// add text to m_TextPieces:
if (i > textBeginning)
m_TextPieces->push_back(wxHtmlTextPiece(textBeginning, i));
// if it is a comment, skip it:
if ( SkipCommentTag(i, m_Source->end()) )
{
textBeginning = i = i + 1; // skip closing '>' too
}
// add another tag to the tree:
else if (i < end_pos-1 && *(i+1) != wxT('/'))
{
wxHtmlTag *chd;
if (cur)
chd = new wxHtmlTag(cur, m_Source,
i, end_pos, cache, m_entitiesParser);
else
{
chd = new wxHtmlTag(NULL, m_Source,
i, end_pos, cache, m_entitiesParser);
if (!m_Tags)
{
// if this is the first tag to be created make the root
// m_Tags point to it:
m_Tags = chd;
}
else
{
// if there is already a root tag add this tag as
// the last sibling:
chd->m_Prev = m_Tags->GetLastSibling();
chd->m_Prev->m_Next = chd;
}
}
if (chd->HasEnding())
{
CreateDOMSubTree(chd,
chd->GetBeginIter(), chd->GetEndIter1(),
cache);
i = chd->GetEndIter2();
}
else
i = chd->GetBeginIter();
textBeginning = i;
}
// ... or skip ending tag:
else
{
while (i < end_pos && *i != wxT('>')) ++i;
textBeginning = i < end_pos ? i+1 : i;
}
}
else ++i;
}
// add remaining text to m_TextPieces:
if (end_pos > textBeginning)
m_TextPieces->push_back(wxHtmlTextPiece(textBeginning, end_pos));
}
wxHtmlTagsCache::wxHtmlTagsCache(const wxString& source)
{
m_Cache = new wxHtmlTagsCacheData;
m_CachePos = 0;
wxChar tagBuffer[256];
const wxString::const_iterator end = source.end();
for ( wxString::const_iterator pos = source.begin(); pos < end; ++pos )
{
if (*pos != wxT('<'))
continue;
// possible tag start found:
// don't cache comment tags
if ( wxHtmlParser::SkipCommentTag(pos, end) )
continue;
// Remember the starting tag position.
wxString::const_iterator stpos = pos++;
// And look for the ending one.
int i;
for ( i = 0;
pos < end && i < (int)WXSIZEOF(tagBuffer) - 1 &&
*pos != wxT('>') && !wxIsspace(*pos);
++i, ++pos )
{
tagBuffer[i] = (wxChar)wxToupper(*pos);
}
tagBuffer[i] = wxT('\0');
while (pos < end && *pos != wxT('>'))
++pos;
if ( pos == end )
{
// We didn't find a closing bracket, this is not a valid tag after
// all. Notice that we need to roll back pos to avoid creating an
// invalid iterator when "++pos" is done in the loop statement.
--pos;
continue;
}
// We have a valid tag, add it to the cache.
size_t tg = Cache().size();
Cache().push_back(wxHtmlCacheItem());
Cache()[tg].Key = stpos;
Cache()[tg].Name = new wxChar[i+1];
memcpy(Cache()[tg].Name, tagBuffer, (i+1)*sizeof(wxChar));
if ((stpos+1) < end && *(stpos+1) == wxT('/')) // ending tag:
{
Cache()[tg].type = wxHtmlCacheItem::Type_EndingTag;
// find matching begin tag:
for (i = tg; i >= 0; i--)
{
if ((Cache()[i].type == wxHtmlCacheItem::Type_NoMatchingEndingTag) && (wxStrcmp(Cache()[i].Name, tagBuffer+1) == 0))
{
Cache()[i].type = wxHtmlCacheItem::Type_Normal;
Cache()[i].End1 = stpos;
Cache()[i].End2 = pos + 1;
break;
}
}
}
else
{
Cache()[tg].type = wxHtmlCacheItem::Type_NoMatchingEndingTag;
if (wxIsCDATAElement(tagBuffer))
{
// store the orig pos in case we are missing the closing
// tag (see below)
const wxString::const_iterator old_pos = pos;
bool foundCloseTag = false;
// find next matching tag
int tag_len = wxStrlen(tagBuffer);
while (pos < end)
{
// find the ending tag
while (pos + 1 < end &&
(*pos != '<' || *(pos+1) != '/'))
++pos;
if (*pos == '<')
++pos;
// see if it matches
int match_pos = 0;
while (pos < end && match_pos < tag_len )
{
wxChar c = *pos;
if ( c == '>' || c == '<' )
break;
// cast to wxChar needed to suppress warning in
// Unicode build
if ((wxChar)wxToupper(c) == tagBuffer[match_pos])
{
++match_pos;
}
else if (c == wxT(' ') || c == wxT('\n') ||
c == wxT('\r') || c == 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;
}
}
}
}
// ok, we're done, now we'll free .Name members of cache - we don't need it anymore:
for ( wxHtmlTagsCacheData::iterator i = Cache().begin();
i != Cache().end(); ++i )
{
wxDELETEA(i->Name);
}
}
wx28HtmlTagsCache::wx28HtmlTagsCache(const wxString& source)
{
const wxChar *src = source.c_str();
int lng = source.length();
wxChar tagBuffer[256];
m_Cache = NULL;
m_CacheSize = 0;
m_CachePos = 0;
int pos = 0;
while (pos < lng)
{
if (src[pos] == wxT('<')) // tag found:
{
if (m_CacheSize % CACHE_INCREMENT == 0)
m_Cache = (wx28HtmlCacheItem*) realloc(m_Cache, (m_CacheSize + CACHE_INCREMENT) * sizeof(wx28HtmlCacheItem));
int tg = m_CacheSize++;
int stpos = pos++;
m_Cache[tg].Key = stpos;
int i;
for ( i = 0;
pos < lng && i < (int)WXSIZEOF(tagBuffer) - 1 &&
src[pos] != wxT('>') && !wxIsspace(src[pos]);
i++, pos++ )
{
tagBuffer[i] = (wxChar)wxToupper(src[pos]);
}
tagBuffer[i] = wxT('\0');
m_Cache[tg].Name = new wxChar[i+1];
memcpy(m_Cache[tg].Name, tagBuffer, (i+1)*sizeof(wxChar));
while (pos < lng && src[pos] != wxT('>')) pos++;
if (src[stpos+1] == wxT('/')) // ending tag:
{
m_Cache[tg].End1 = m_Cache[tg].End2 = -2;
// find matching begin tag:
for (i = tg; i >= 0; i--)
if ((m_Cache[i].End1 == -1) && (wxStrcmp(m_Cache[i].Name, tagBuffer+1) == 0))
{
m_Cache[i].End1 = stpos;
m_Cache[i].End2 = pos + 1;
break;
}
}
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:
for (int i = 0; i < m_CacheSize; i++)
{
delete[] m_Cache[i].Name;
m_Cache[i].Name = NULL;
}
}
