/////////////////////////////////////////////////////////////////////////////
-// Name: htmlpars.cpp
+// Name: src/html/htmlpars.cpp
// Purpose: wxHtmlParser class (generic parser)
// 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 && wxUSE_STREAMS
-#ifdef __BORDLANDC__
-#pragma hdrstop
-#endif
-
-#ifndef WXPRECOMP
+#ifndef WX_PRECOMP
+ #include "wx/dynarray.h"
#include "wx/log.h"
#include "wx/intl.h"
+ #include "wx/app.h"
+ #include "wx/wxcrtvararg.h"
#endif
#include "wx/tokenzr.h"
#include "wx/fontmap.h"
#include "wx/html/htmldefs.h"
#include "wx/html/htmlpars.h"
-#include "wx/dynarray.h"
#include "wx/arrimpl.cpp"
+#ifdef __WXWINCE__
+ #include "wx/msw/wince/missing.h" // for bsearch()
+#endif
+
+// DLL options compatibility check:
+WX_CHECK_BUILD_OPTIONS("wxHTML")
+
+const wxChar *wxTRACE_HTML_DEBUG = _T("htmldebug");
+
//-----------------------------------------------------------------------------
// wxHtmlParser helpers
//-----------------------------------------------------------------------------
};
WX_DECLARE_OBJARRAY(wxHtmlTextPiece, wxHtmlTextPieces);
-WX_DEFINE_OBJARRAY(wxHtmlTextPieces);
+WX_DEFINE_OBJARRAY(wxHtmlTextPieces)
class wxHtmlParserState
{
wxHtmlParser::~wxHtmlParser()
{
- while (RestoreState())
- DestroyDOMTree();
+ while (RestoreState()) {}
+ DestroyDOMTree();
+
+ if (m_HandlersStack)
+ {
+ wxList& tmp = *m_HandlersStack;
+ wxList::iterator it, en;
+ for( it = tmp.begin(), en = tmp.end(); it != en; ++it )
+ delete (wxHashTable*)*it;
+ tmp.clear();
+ }
delete m_HandlersStack;
m_HandlersHash.Clear();
- m_HandlersList.DeleteContents(TRUE);
- m_HandlersList.Clear();
+ WX_CLEAR_LIST(wxList, m_HandlersList);
delete m_entitiesParser;
}
wxObject* wxHtmlParser::Parse(const wxString& source)
{
- wxObject *result;
-
InitParser(source);
DoParsing();
- result = GetProduct();
+ wxObject *result = GetProduct();
DoneParser();
return result;
}
void wxHtmlParser::InitParser(const wxString& source)
{
SetSource(source);
+ m_stopParsing = false;
}
void wxHtmlParser::DoneParser()
{
wxHtmlTagsCache cache(m_Source);
m_TextPieces = new wxHtmlTextPieces;
- CreateDOMSubTree(NULL, 0, m_Source.Length(), &cache);
+ CreateDOMSubTree(NULL, 0, m_Source.length(), &cache);
m_CurTextPiece = 0;
}
+extern bool wxIsCDATAElement(const wxChar *tag);
+
void wxHtmlParser::CreateDOMSubTree(wxHtmlTag *cur,
int begin_pos, int end_pos,
wxHtmlTagsCache *cache)
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);
wxHtmlTextPiece(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('-'))
+ wxString::const_iterator iter = m_Source.begin() + i;
+ if ( SkipCommentTag(iter, m_Source.end()) )
{
- // 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;
- }
+ textBeginning =
+ i = iter - m_Source.begin() + 1; // skip closing '>' too
}
-
+
// add another tag to the tree:
else if (i < end_pos-1 && m_Source.GetChar(i+1) != wxT('/'))
- {
+ {
wxHtmlTag *chd;
- if (cur)
- chd = new wxHtmlTag(cur, m_Source,
+ if (cur)
+ chd = new wxHtmlTag(cur, m_Source,
i, end_pos, cache, m_entitiesParser);
- else
+ else
{
chd = new wxHtmlTag(NULL, m_Source,
i, end_pos, cache, m_entitiesParser);
- if (!m_Tags)
+ if (!m_Tags)
{
- // if this is the first tag to be created make the root
+ // 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
+ // 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(),
+ chd->GetBeginPos(), chd->GetEndPos1(),
cache);
i = chd->GetEndPos2();
}
else
i = chd->GetBeginPos();
+
textBeginning = i;
}
// ... or skip ending tag:
- else
+ else
{
while (i < end_pos && m_Source.GetChar(i) != wxT('>')) i++;
textBeginning = i+1;
m_TextPieces = NULL;
}
-void wxHtmlParser::DoParsing()
+void wxHtmlParser::DoParsing()
{
m_CurTag = m_Tags;
m_CurTextPiece = 0;
- DoParsing(0, m_Source.Length());
+ DoParsing(0, m_Source.length());
}
void wxHtmlParser::DoParsing(int begin_pos, int end_pos)
{
if (end_pos <= begin_pos) return;
-
+
wxHtmlTextPieces& pieces = *m_TextPieces;
size_t piecesCnt = pieces.GetCount();
-
+
while (begin_pos < end_pos)
{
while (m_CurTag && m_CurTag->GetBeginPos() < begin_pos)
m_CurTag = m_CurTag->GetNextTag();
- while (m_CurTextPiece < piecesCnt &&
+ while (m_CurTextPiece < piecesCnt &&
pieces[m_CurTextPiece].m_pos < begin_pos)
m_CurTextPiece++;
- if (m_CurTextPiece < piecesCnt &&
- (!m_CurTag ||
+ if (m_CurTextPiece < piecesCnt &&
+ (!m_CurTag ||
pieces[m_CurTextPiece].m_pos < m_CurTag->GetBeginPos()))
{
// Add text:
- AddText(m_Source.Mid(pieces[m_CurTextPiece].m_pos,
- pieces[m_CurTextPiece].m_lng));
- begin_pos = pieces[m_CurTextPiece].m_pos +
+ AddText(GetEntitiesParser()->Parse(
+ m_Source.Mid(pieces[m_CurTextPiece].m_pos,
+ pieces[m_CurTextPiece].m_lng)));
+ begin_pos = pieces[m_CurTextPiece].m_pos +
pieces[m_CurTextPiece].m_lng;
m_CurTextPiece++;
}
else if (m_CurTag)
{
- // Add tag:
- if (m_CurTag)
- {
- if (m_CurTag->HasEnding())
- begin_pos = m_CurTag->GetEndPos2();
- else
- begin_pos = m_CurTag->GetBeginPos();
- }
+ if (m_CurTag->HasEnding())
+ begin_pos = m_CurTag->GetEndPos2();
+ else
+ begin_pos = m_CurTag->GetBeginPos();
wxHtmlTag *t = m_CurTag;
m_CurTag = m_CurTag->GetNextTag();
AddTag(*t);
+ if (m_stopParsing)
+ return;
}
else break;
}
void wxHtmlParser::AddTag(const wxHtmlTag& tag)
{
wxHtmlTagHandler *h;
- bool inner = FALSE;
+ bool inner = false;
h = (wxHtmlTagHandler*) m_HandlersHash.Get(tag.GetName());
if (h)
+ {
inner = h->HandleTag(tag);
+ if (m_stopParsing)
+ return;
+ }
if (!inner)
{
if (tag.HasEnding())
wxStringTokenizer tokenizer(s, wxT(", "));
while (tokenizer.HasMoreTokens())
- m_HandlersHash.Put(tokenizer.NextToken(), handler);
+ m_HandlersHash.Put(tokenizer.GetNextToken(), handler);
if (m_HandlersList.IndexOf(handler) == wxNOT_FOUND)
m_HandlersList.Append(handler);
handler->SetParser(this);
}
-void wxHtmlParser::PushTagHandler(wxHtmlTagHandler *handler, wxString tags)
+void wxHtmlParser::PushTagHandler(wxHtmlTagHandler *handler, const wxString& tags)
{
wxStringTokenizer tokenizer(tags, wxT(", "));
wxString key;
if (m_HandlersStack == NULL)
{
m_HandlersStack = new wxList;
- m_HandlersStack->DeleteContents(TRUE);
}
- m_HandlersStack->Insert(new wxHashTable(m_HandlersHash));
+ m_HandlersStack->Insert((wxObject*)new wxHashTable(m_HandlersHash));
while (tokenizer.HasMoreTokens())
{
- key = tokenizer.NextToken();
+ key = tokenizer.GetNextToken();
m_HandlersHash.Delete(key);
m_HandlersHash.Put(key, handler);
}
void wxHtmlParser::PopTagHandler()
{
- wxNode *first;
-
- if (m_HandlersStack == NULL ||
- (first = m_HandlersStack->GetFirst()) == NULL)
+ wxList::compatibility_iterator first;
+
+ if ( !m_HandlersStack ||
+#if wxUSE_STL
+ !(first = m_HandlersStack->GetFirst())
+#else // !wxUSE_STL
+ ((first = m_HandlersStack->GetFirst()) == NULL)
+#endif // wxUSE_STL/!wxUSE_STL
+ )
{
wxLogWarning(_("Warning: attempt to remove HTML tag handler from empty stack."));
return;
}
m_HandlersHash = *((wxHashTable*) first->GetData());
- m_HandlersStack->DeleteNode(first);
+ delete (wxHashTable*) first->GetData();
+ m_HandlersStack->Erase(first);
}
void wxHtmlParser::SetSourceAndSaveState(const wxString& src)
m_TextPieces = NULL;
m_CurTextPiece = 0;
m_Source = wxEmptyString;
-
+
SetSource(src);
}
bool wxHtmlParser::RestoreState()
{
- if (!m_SavedStates) return FALSE;
-
+ if (!m_SavedStates) return false;
+
+ DestroyDOMTree();
+
wxHtmlParserState *s = m_SavedStates;
m_SavedStates = s->m_nextState;
-
+
m_CurTag = s->m_curTag;
m_Tags = s->m_tags;
m_TextPieces = s->m_textPieces;
m_CurTextPiece = s->m_curTextPiece;
m_Source = s->m_source;
-
+
delete s;
- return TRUE;
+ return true;
+}
+
+wxString wxHtmlParser::GetInnerSource(const wxHtmlTag& tag)
+{
+ return GetSource()->Mid(tag.GetBeginPos(),
+ tag.GetEndPos1() - tag.GetBeginPos());
}
//-----------------------------------------------------------------------------
IMPLEMENT_ABSTRACT_CLASS(wxHtmlTagHandler,wxObject)
+void wxHtmlTagHandler::ParseInnerSource(const wxString& source)
+{
+ // It is safe to temporarily change the source being parsed,
+ // provided we restore the state back after parsing
+ m_Parser->SetSourceAndSaveState(source);
+ m_Parser->DoParsing();
+ m_Parser->RestoreState();
+}
+
//-----------------------------------------------------------------------------
// wxHtmlEntitiesParser
void wxHtmlEntitiesParser::SetEncoding(wxFontEncoding encoding)
{
#if wxUSE_WCHAR_T && !wxUSE_UNICODE
- if (encoding == m_encoding) return;
+ if (encoding == m_encoding)
+ return;
+
delete m_conv;
- m_conv = NULL;
+
m_encoding = encoding;
- if (m_encoding != wxFONTENCODING_SYSTEM)
+ if (m_encoding == wxFONTENCODING_SYSTEM)
+ m_conv = NULL;
+ else
m_conv = new wxCSConv(wxFontMapper::GetEncodingName(m_encoding));
+#else
+ (void) encoding;
#endif
}
const wxChar *in_str = input.c_str();
wxString output;
+ output.reserve(input.length());
+
for (c = in_str, last = in_str; *c != wxT('\0'); c++)
{
if (*c == wxT('&'))
{
if (c - last > 0)
output.append(last, c - last);
- if (++c == wxT('\0')) break;
+ if ( *++c == wxT('\0') )
+ break;
+
wxString entity;
const wxChar *ent_s = c;
+ wxChar entity_char;
+
for (; (*c >= wxT('a') && *c <= wxT('z')) ||
(*c >= wxT('A') && *c <= wxT('Z')) ||
(*c >= wxT('0') && *c <= wxT('9')) ||
entity.append(ent_s, c - ent_s);
if (*c != wxT(';')) c--;
last = c+1;
- output << GetEntityChar(entity);
+ entity_char = GetEntityChar(entity);
+ if (entity_char)
+ output << entity_char;
+ else
+ {
+ output.append(ent_s-1, c-ent_s+2);
+ wxLogTrace(wxTRACE_HTML_DEBUG,
+ wxT("Unrecognized HTML entity: '%s'"),
+ entity.c_str());
+ }
}
}
if (*last != wxT('\0'))
unsigned code;
};
-static int LINKAGEMODE compar_entity(const void *key, const void *item)
+extern "C" int LINKAGEMODE wxHtmlEntityCompare(const void *key, const void *item)
{
return wxStrcmp((wxChar*)key, ((wxHtmlEntityInfo*)item)->name);
}
+#if !wxUSE_UNICODE
wxChar wxHtmlEntitiesParser::GetCharForCode(unsigned code)
{
-#if wxUSE_UNICODE
- return (wxChar)code;
-#elif wxUSE_WCHAR_T
+#if wxUSE_WCHAR_T
char buf[2];
wchar_t wbuf[2];
wbuf[0] = (wchar_t)code;
return (code < 256) ? (wxChar)code : '?';
#endif
}
+#endif
wxChar wxHtmlEntitiesParser::GetEntityChar(const wxString& entity)
{
while (substitutions[substitutions_cnt].code != 0)
substitutions_cnt++;
- wxHtmlEntityInfo *info;
+ wxHtmlEntityInfo *info = NULL;
+#ifdef __WXWINCE__
+ // bsearch crashes under WinCE for some reason
+ size_t i;
+ for (i = 0; i < substitutions_cnt; i++)
+ {
+ if (entity == substitutions[i].name)
+ {
+ info = & substitutions[i];
+ break;
+ }
+ }
+#else
info = (wxHtmlEntityInfo*) bsearch(entity.c_str(), substitutions,
substitutions_cnt,
sizeof(wxHtmlEntityInfo),
- compar_entity);
+ wxHtmlEntityCompare);
+#endif
if (info)
code = info->code;
}
if (code == 0)
- return wxT('?');
+ return 0;
else
return GetCharForCode(code);
}
-#endif
+wxFSFile *wxHtmlParser::OpenURL(wxHtmlURLType WXUNUSED(type),
+ const wxString& url) const
+{
+ return m_FS ? m_FS->OpenFile(url) : NULL;
+
+}
+
+
+//-----------------------------------------------------------------------------
+// wxHtmlParser::ExtractCharsetInformation
+//-----------------------------------------------------------------------------
+
+class wxMetaTagParser : public wxHtmlParser
+{
+public:
+ wxMetaTagParser() { }
+
+ wxObject* GetProduct() { return NULL; }
+
+protected:
+ virtual void AddText(const wxChar* WXUNUSED(txt)) {}
+
+ DECLARE_NO_COPY_CLASS(wxMetaTagParser)
+};
+
+class wxMetaTagHandler : public wxHtmlTagHandler
+{
+public:
+ wxMetaTagHandler(wxString *retval) : wxHtmlTagHandler(), m_retval(retval) {}
+ wxString GetSupportedTags() { return wxT("META,BODY"); }
+ bool HandleTag(const wxHtmlTag& tag);
+
+private:
+ wxString *m_retval;
+
+ DECLARE_NO_COPY_CLASS(wxMetaTagHandler)
+};
+
+bool wxMetaTagHandler::HandleTag(const wxHtmlTag& tag)
+{
+ if (tag.GetName() == _T("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")))
+ {
+ wxString content = tag.GetParam(_T("CONTENT")).Lower();
+ if (content.Left(19) == _T("text/html; charset="))
+ {
+ *m_retval = content.Mid(19);
+ m_Parser->StopParsing();
+ }
+ }
+ return false;
+}
+
+
+/*static*/
+wxString wxHtmlParser::ExtractCharsetInformation(const wxString& markup)
+{
+ wxString charset;
+ wxMetaTagParser *parser = new wxMetaTagParser();
+ if(parser)
+ {
+ parser->AddTagHandler(new wxMetaTagHandler(&charset));
+ parser->Parse(markup);
+ delete parser;
+ }
+ return charset;
+}
+
+/* static */
+bool
+wxHtmlParser::SkipCommentTag(wxString::const_iterator& start,
+ wxString::const_iterator end)
+{
+ wxASSERT_MSG( *start == '<', _T("should be called on the tag start") );
+
+ wxString::const_iterator p = start;
+
+ // comments begin with "<!--" in HTML 4.0
+ if ( end - p < 3 || *++p != '!' || *++p != '-' || *++p != '-' )
+ {
+ // not a comment at all
+ 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
+ start = p;
+
+ // comments end with "--[ \t\r\n]*>", i.e. white space is allowed between
+ // comment delimiter and the closing tag character (section 3.2.4 of
+ // http://www.w3.org/TR/html401/)
+ int dashes = 0;
+ while ( ++p < end )
+ {
+ const wxChar c = *p;
+
+ if ( (c == wxT(' ') || c == wxT('\n') ||
+ c == wxT('\r') || c == wxT('\t')) && dashes >= 2 )
+ {
+ // ignore white space before potential tag end
+ continue;
+ }
+
+ if ( c == wxT('>') && dashes >= 2 )
+ {
+ // found end of comment
+ start = p;
+ break;
+ }
+
+ if ( c == wxT('-') )
+ dashes++;
+ else
+ dashes = 0;
+ }
+
+ return true;
+}
+
+#endif // wxUSE_HTML
projects / wxWidgets.git / blobdiff