#endif
#include "wx/html/htmlpars.h"
+#include "wx/html/styleparams.h"
+
#include "wx/vector.h"
#include <stdio.h> // for vsscanf
{
// this is "pos" value passed to wxHtmlTag's constructor.
// it is position of '<' character of the tag
- int Key;
+ wxString::const_iterator Key;
+
+ // Tag type
+ enum Type
+ {
+ Type_Normal, // normal tag with a matching ending tag
+ Type_NoMatchingEndingTag, // there's no ending tag for this tag
+ Type_EndingTag // this is ending tag </..>
+ };
+ Type type;
// end positions for the tag:
// end1 is '<' of ending tag,
// end2 is '>' or both are
- // -1 if there is no ending tag for this one...
- // or -2 if this is ending tag </...>
- int End1, End2;
+ wxString::const_iterator End1, End2;
// name of this tag
wxChar *Name;
bool wxIsCDATAElement(const wxChar *tag)
{
- return (wxStrcmp(tag, _T("SCRIPT")) == 0) ||
- (wxStrcmp(tag, _T("STYLE")) == 0);
+ return (wxStrcmp(tag, wxT("SCRIPT")) == 0) ||
+ (wxStrcmp(tag, wxT("STYLE")) == 0);
+}
+
+bool wxIsCDATAElement(const wxString& tag)
+{
+ return (wxStrcmp(tag.wx_str(), wxS("SCRIPT")) == 0) ||
+ (wxStrcmp(tag.wx_str(), wxS("STYLE")) == 0);
}
wxHtmlTagsCache::wxHtmlTagsCache(const wxString& source)
m_Cache = new wxHtmlTagsCacheData;
m_CachePos = 0;
- const wxChar *src = source.c_str();
- int lng = source.length();
wxChar tagBuffer[256];
- for ( int pos = 0; pos < lng; pos++ )
+ const wxString::const_iterator end = source.end();
+ for ( wxString::const_iterator pos = source.begin(); pos < end; ++pos )
{
- if (src[pos] == wxT('<')) // tag found:
+ if (*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();
+ if ( wxHtmlParser::SkipCommentTag(pos, source.end()) )
continue;
- }
size_t tg = Cache().size();
Cache().push_back(wxHtmlCacheItem());
- int stpos = pos++;
+ wxString::const_iterator stpos = pos++;
Cache()[tg].Key = stpos;
int i;
for ( i = 0;
- pos < lng && i < (int)WXSIZEOF(tagBuffer) - 1 &&
- src[pos] != wxT('>') && !wxIsspace(src[pos]);
- i++, pos++ )
+ pos < end && i < (int)WXSIZEOF(tagBuffer) - 1 &&
+ *pos != wxT('>') && !wxIsspace(*pos);
+ ++i, ++pos )
{
- tagBuffer[i] = (wxChar)wxToupper(src[pos]);
+ tagBuffer[i] = (wxChar)wxToupper(*pos);
}
- tagBuffer[i] = _T('\0');
+ tagBuffer[i] = wxT('\0');
Cache()[tg].Name = new wxChar[i+1];
memcpy(Cache()[tg].Name, tagBuffer, (i+1)*sizeof(wxChar));
- while (pos < lng && src[pos] != wxT('>')) pos++;
+ while (pos < end && *pos != wxT('>'))
+ ++pos;
- if (src[stpos+1] == wxT('/')) // ending tag:
+ if ((stpos+1) < end && *(stpos+1) == wxT('/')) // ending tag:
{
- Cache()[tg].End1 = Cache()[tg].End2 = -2;
+ Cache()[tg].type = wxHtmlCacheItem::Type_EndingTag;
// find matching begin tag:
for (i = tg; i >= 0; i--)
- if ((Cache()[i].End1 == -1) && (wxStrcmp(Cache()[i].Name, tagBuffer+1) == 0))
+ {
+ 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].End1 = Cache()[tg].End2 = -1;
+ Cache()[tg].type = wxHtmlCacheItem::Type_NoMatchingEndingTag;
if (wxIsCDATAElement(tagBuffer))
{
// store the orig pos in case we are missing the closing
// tag (see below)
- wxInt32 old_pos = pos;
+ const wxString::const_iterator old_pos = pos;
bool foundCloseTag = false;
// find next matching tag
int tag_len = wxStrlen(tagBuffer);
- while (pos < lng)
+ while (pos < end)
{
// find the ending tag
- while (pos + 1 < lng &&
- (src[pos] != '<' || src[pos+1] != '/'))
+ while (pos + 1 < end &&
+ (*pos != '<' || *(pos+1) != '/'))
++pos;
- if (src[pos] == '<')
+ if (*pos == '<')
++pos;
// see if it matches
int match_pos = 0;
- while (pos < lng && match_pos < tag_len && src[pos] != '>' && src[pos] != '<') {
+ 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(src[pos]) == tagBuffer[match_pos]) {
+ if ((wxChar)wxToupper(c) == tagBuffer[match_pos])
+ {
++match_pos;
}
- else if (src[pos] == wxT(' ') || src[pos] == wxT('\n') ||
- src[pos] == wxT('\r') || src[pos] == wxT('\t')) {
+ else if (c == wxT(' ') || c == wxT('\n') ||
+ c == wxT('\r') || c == wxT('\t'))
+ {
// need to skip over these
}
- else {
+ else
+ {
match_pos = 0;
}
++pos;
for ( wxHtmlTagsCacheData::iterator i = Cache().begin();
i != Cache().end(); ++i )
{
- delete[] i->Name;
- i->Name = NULL;
+ wxDELETEA(i->Name);
}
}
delete m_Cache;
}
-void wxHtmlTagsCache::QueryTag(int at, int* end1, int* end2)
+void wxHtmlTagsCache::QueryTag(const wxString::const_iterator& at,
+ const wxString::const_iterator& inputEnd,
+ wxString::const_iterator *end1,
+ wxString::const_iterator *end2,
+ bool *hasEnding)
{
if (Cache().empty())
return;
int delta = (at < Cache()[m_CachePos].Key) ? -1 : 1;
do
{
- if ( m_CachePos < 0 || m_CachePos == Cache().size() )
+ m_CachePos += delta;
+
+ if ( m_CachePos < 0 || m_CachePos >= (int)Cache().size() )
{
+ if ( m_CachePos < 0 )
+ m_CachePos = 0;
+ else
+ m_CachePos = Cache().size() - 1;
// 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;
+ *end2 = inputEnd;
+ *hasEnding = true;
return;
}
-
- m_CachePos += delta;
}
while (Cache()[m_CachePos].Key != at);
}
- *end1 = Cache()[m_CachePos].End1;
- *end2 = Cache()[m_CachePos].End2;
+
+ switch ( Cache()[m_CachePos].type )
+ {
+ case wxHtmlCacheItem::Type_Normal:
+ *end1 = Cache()[m_CachePos].End1;
+ *end2 = Cache()[m_CachePos].End2;
+ *hasEnding = true;
+ break;
+
+ case wxHtmlCacheItem::Type_EndingTag:
+ wxFAIL_MSG("QueryTag called for ending tag - can't be");
+ // but if it does happen, fall through, better than crashing
+
+ case wxHtmlCacheItem::Type_NoMatchingEndingTag:
+ // If input HTML is invalid and there's no closing tag for this
+ // one, pretend that it runs all the way to the end of input
+ *end1 = inputEnd;
+ *end2 = inputEnd;
+ *hasEnding = false;
+ break;
+ }
}
//-----------------------------------------------------------------------------
wxHtmlTag::wxHtmlTag(wxHtmlTag *parent,
- const wxString& source, int pos, int end_pos,
+ const wxString *source,
+ const wxString::const_iterator& pos,
+ const wxString::const_iterator& end_pos,
wxHtmlTagsCache *cache,
wxHtmlEntitiesParser *entParser)
{
/* Find parameters and their values: */
- int i;
- wxChar c;
+ wxChar c wxDUMMY_INITIALIZE(0);
// fill-in name, params and begin pos:
- i = pos+1;
+ wxString::const_iterator i(pos+1);
// find tag's name and convert it to uppercase:
while ((i < end_pos) &&
- ((c = source[i++]) != wxT(' ') && c != wxT('\r') &&
+ ((c = *(i++)) != wxT(' ') && c != wxT('\r') &&
c != wxT('\n') && c != wxT('\t') &&
- c != wxT('>')))
+ c != wxT('>') && c != wxT('/')))
{
if ((c >= wxT('a')) && (c <= wxT('z')))
c -= (wxT('a') - wxT('A'));
// if the tag has parameters, read them and "normalize" them,
// i.e. convert to uppercase, replace whitespaces by spaces and
// remove whitespaces around '=':
- if (source[i-1] != wxT('>'))
+ if (*(i-1) != wxT('>'))
{
#define IS_WHITE(c) (c == wxT(' ') || c == wxT('\r') || \
c == wxT('\n') || c == wxT('\t'))
state = ST_BEFORE_NAME;
while (i < end_pos)
{
- c = source[i++];
+ c = *(i++);
if (c == wxT('>') && !(state == ST_VALUE && quote != 0))
{
if (state == ST_BEFORE_EQ || state == ST_NAME)
{
m_ParamNames.Add(pname);
- m_ParamValues.Add(wxEmptyString);
+ m_ParamValues.Add(wxGetEmptyString());
}
else if (state == ST_VALUE && quote == 0)
{
else if (!IS_WHITE(c))
{
m_ParamNames.Add(pname);
- m_ParamValues.Add(wxEmptyString);
+ m_ParamValues.Add(wxGetEmptyString());
pname = c;
state = ST_NAME;
}
if (!IS_WHITE(c))
{
if (c == wxT('"') || c == wxT('\''))
- quote = c, pvalue = wxEmptyString;
+ quote = c, pvalue = wxGetEmptyString();
else
quote = 0, pvalue = c;
state = ST_VALUE;
#undef IS_WHITE
}
m_Begin = i;
-
- cache->QueryTag(pos, &m_End1, &m_End2);
+ cache->QueryTag(pos, source->end(), &m_End1, &m_End2, &m_hasEnding);
if (m_End1 > end_pos) m_End1 = end_pos;
if (m_End2 > end_pos) m_End2 = end_pos;
+
+#if WXWIN_COMPATIBILITY_2_8
+ m_sourceStart = source->begin();
+#endif
+
+ // Try to parse any style parameters that can be handled simply by
+ // converting them to the equivalent HTML 3 attributes: this is a far cry
+ // from perfect but better than nothing.
+ static const struct EquivAttr
+ {
+ const char *style;
+ const char *attr;
+ } equivAttrs[] =
+ {
+ { "text-align", "ALIGN" },
+ { "width", "WIDTH" },
+ { "vertical-align", "VALIGN" },
+ { "background", "BGCOLOR" },
+ };
+
+ wxHtmlStyleParams styleParams(*this);
+ for ( unsigned n = 0; n < WXSIZEOF(equivAttrs); n++ )
+ {
+ const EquivAttr& ea = equivAttrs[n];
+ if ( styleParams.HasParam(ea.style) && !HasParam(ea.attr) )
+ {
+ m_ParamNames.Add(ea.attr);
+ m_ParamValues.Add(styleParams.GetParam(ea.style));
+ }
+ }
}
wxHtmlTag::~wxHtmlTag()
return (m_ParamNames.Index(par, false) != wxNOT_FOUND);
}
-wxString wxHtmlTag::GetParam(const wxString& par, bool with_commas) const
+wxString wxHtmlTag::GetParam(const wxString& par, bool with_quotes) const
{
int index = m_ParamNames.Index(par, false);
if (index == wxNOT_FOUND)
- return wxEmptyString;
- if (with_commas)
+ return wxGetEmptyString();
+ if (with_quotes)
{
// VS: backward compatibility, seems to be never used by wxHTML...
wxString s;
return wxSscanf(parval, format, param);
}
-bool wxHtmlTag::GetParamAsColour(const wxString& par, wxColour *clr) const
+/* static */
+bool wxHtmlTag::ParseAsColour(const wxString& str, wxColour *clr)
{
- wxCHECK_MSG( clr, false, _T("invalid colour argument") );
-
- wxString str = GetParam(par);
+ wxCHECK_MSG( clr, false, wxT("invalid colour argument") );
// handle colours defined in HTML 4.0 first:
- if (str.length() > 1 && str[0] != _T('#'))
+ if (str.length() > 1 && str[0] != wxT('#'))
{
#define HTML_COLOUR(name, r, g, b) \
- if (str.IsSameAs(wxT(name), false)) \
+ if (str.IsSameAs(wxS(name), false)) \
{ clr->Set(r, g, b); return true; }
HTML_COLOUR("black", 0x00,0x00,0x00)
HTML_COLOUR("silver", 0xC0,0xC0,0xC0)
return false;
}
+bool wxHtmlTag::GetParamAsColour(const wxString& par, wxColour *clr) const
+{
+ const wxString str = GetParam(par);
+ return !str.empty() && ParseAsColour(str, clr);
+}
+
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);
+ if ( !GetParam(par).ToLong(&i) )
+ return false;
+
*clr = (int)i;
- return succ;
+ return true;
}
wxString wxHtmlTag::GetAllParams() const