Amebis/wxWidgets
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fixed wxHTML parsing to run in O(n) even in UTF8 build

Browse Source 
git-svn-id: https://svn.wxwidgets.org/svn/wx/wxWidgets/trunk@48390 c3d73ce0-8a6f-49c7-b76d-6d57e0e08775
This commit is contained in:
Václav Slavík
2007-08-26 10:04:11 +00:00
parent 4f33a4dd4c
commit b1a3a96499
5 changed files with 191 additions and 120 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
include/wx/html/htmlpars.h
View File

@@ -73,7 +73,8 @@ public:
// Parses the m_Source from begin_pos to end_pos-1. // Parses the m_Source from begin_pos to end_pos-1.
// (in noparams version it parses whole m_Source) // (in noparams version it parses whole m_Source)
void DoParsing(int begin_pos, int end_pos); void DoParsing(const wxString::const_iterator& begin_pos,
const wxString::const_iterator& end_pos);
void DoParsing(); void DoParsing();
// Returns pointer to the tag at parser's current position // Returns pointer to the tag at parser's current position
@@ -104,7 +105,7 @@ public:
// Restores state before last call to PushTagHandler // Restores state before last call to PushTagHandler
void PopTagHandler(); void PopTagHandler();
wxString* GetSource() {return &m_Source;} const wxString* GetSource() {return m_Source;}
void SetSource(const wxString& src); void SetSource(const wxString& src);
// Sets HTML source and remembers current parser's state so that it can // Sets HTML source and remembers current parser's state so that it can
@@ -140,7 +141,8 @@ protected:
void CreateDOMTree(); void CreateDOMTree();
void DestroyDOMTree(); void DestroyDOMTree();
void CreateDOMSubTree(wxHtmlTag *cur, void CreateDOMSubTree(wxHtmlTag *cur,
int begin_pos, int end_pos, const wxString::const_iterator& begin_pos,
const wxString::const_iterator& end_pos,
wxHtmlTagsCache *cache); wxHtmlTagsCache *cache);
// Adds text to the output. // Adds text to the output.
@@ -163,7 +165,7 @@ protected:
wxHtmlTextPieces *m_TextPieces; wxHtmlTextPieces *m_TextPieces;
size_t m_CurTextPiece; size_t m_CurTextPiece;
wxString m_Source; const wxString *m_Source;
wxHtmlParserState *m_SavedStates; wxHtmlParserState *m_SavedStates;
@@ -232,7 +234,7 @@ protected:
// parses input between beginning and ending tag. // parses input between beginning and ending tag.
// m_Parser must be set. // m_Parser must be set.
void ParseInner(const wxHtmlTag& tag) void ParseInner(const wxHtmlTag& tag)
{ m_Parser->DoParsing(tag.GetBeginPos(), tag.GetEndPos1()); } { m_Parser->DoParsing(tag.GetBeginIter(), tag.GetEndIter1()); }
// Parses given source as if it was tag's inner code (see // Parses given source as if it was tag's inner code (see
// wxHtmlParser::GetInnerSource). Unlike ParseInner(), this method lets // wxHtmlParser::GetInnerSource). Unlike ParseInner(), this method lets

45
include/wx/html/htmltag.h
View File

@@ -31,7 +31,7 @@ class WXDLLIMPEXP_HTML wxHtmlTagsCache
{ {
private: private:
wxHtmlTagsCacheData *m_Cache; wxHtmlTagsCacheData *m_Cache;
size_t m_CachePos; int m_CachePos;
wxHtmlTagsCacheData& Cache() { return *m_Cache; } wxHtmlTagsCacheData& Cache() { return *m_Cache; }
@@ -41,7 +41,11 @@ public:
virtual ~wxHtmlTagsCache(); virtual ~wxHtmlTagsCache();
// Finds parameters for tag starting at at and fills the variables // Finds parameters for tag starting at at and fills the variables
void QueryTag(int at, int* end1, int* end2); void QueryTag(const wxString::const_iterator& at,
const wxString::const_iterator& inputEnd,
wxString::const_iterator *end1,
wxString::const_iterator *end2,
bool *hasEnding);
DECLARE_NO_COPY_CLASS(wxHtmlTagsCache) DECLARE_NO_COPY_CLASS(wxHtmlTagsCache)
}; };
@@ -60,7 +64,9 @@ protected:
// The tag begins (with '<' character) at position pos in source // The tag begins (with '<' character) at position pos in source
// end_pos is position where parsing ends (usually end of document) // end_pos is position where parsing ends (usually end of document)
wxHtmlTag(wxHtmlTag *parent, 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, wxHtmlTagsCache *cache,
wxHtmlEntitiesParser *entParser); wxHtmlEntitiesParser *entParser);
friend class wxHtmlParser; friend class wxHtmlParser;
@@ -108,24 +114,37 @@ public:
// Returns string containing all params. // Returns string containing all params.
wxString GetAllParams() const; wxString GetAllParams() const;
// return true if this there is matching ending tag // return true if there is matching ending tag
inline bool HasEnding() const {return m_End1 >= 0;} inline bool HasEnding() const {return m_hasEnding;}
// returns beginning position of _internal_ block of text // returns beginning position of _internal_ block of text
// See explanation (returned value is marked with *): // See explanation (returned value is marked with *):
// bla bla bla <MYTAG>* bla bla intenal text</MYTAG> bla bla // bla bla bla <MYTAG>* bla bla intenal text</MYTAG> bla bla
inline int GetBeginPos() const {return m_Begin;} wxString::const_iterator GetBeginIter() const
{ return m_Begin; }
// returns ending position of _internal_ block of text. // returns ending position of _internal_ block of text.
// bla bla bla <MYTAG> bla bla intenal text*</MYTAG> bla bla // bla bla bla <MYTAG> bla bla intenal text*</MYTAG> bla bla
inline int GetEndPos1() const {return m_End1;} wxString::const_iterator GetEndIter1() const
{ wxASSERT(m_hasEnding); return m_End1; }
// returns end position 2 : // returns end position 2 :
// bla bla bla <MYTAG> bla bla internal text</MYTAG>* bla bla // bla bla bla <MYTAG> bla bla internal text</MYTAG>* bla bla
inline int GetEndPos2() const {return m_End2;} wxString::const_iterator GetEndIter2() const
{ wxASSERT(m_hasEnding); return m_End2; }
#if WXWIN_COMPATIBILITY_2_8
wxDEPRECATED( inline int GetBeginPos() const );
wxDEPRECATED( inline int GetEndPos1() const );
wxDEPRECATED( inline int GetEndPos2() const );
#endif // WXWIN_COMPATIBILITY_2_8
private: private:
wxString m_Name; wxString m_Name;
int m_Begin, m_End1, m_End2; bool m_hasEnding;
wxString::const_iterator m_Begin, m_End1, m_End2;
wxArrayString m_ParamNames, m_ParamValues; wxArrayString m_ParamNames, m_ParamValues;
#if WXWIN_COMPATIBILITY_2_8
wxString::const_iterator m_sourceStart;
#endif
// DOM tree relations: // DOM tree relations:
wxHtmlTag *m_Next; wxHtmlTag *m_Next;
@@ -137,10 +156,16 @@ private:
}; };
#if WXWIN_COMPATIBILITY_2_8
inline int wxHtmlTag::GetBeginPos() const { return m_Begin - m_sourceStart; }
inline int wxHtmlTag::GetEndPos1() const { return m_End1 - m_sourceStart; }
inline int wxHtmlTag::GetEndPos2() const { return m_End2 - m_sourceStart; }
#endif // WXWIN_COMPATIBILITY_2_8
#endif
#endif // wxUSE_HTML
#endif // _WX_HTMLTAG_H_ #endif // _WX_HTMLTAG_H_

112
src/html/htmlpars.cpp
View File

@@ -48,8 +48,10 @@ class wxHtmlTextPiece
{ {
public: public:
wxHtmlTextPiece() {} wxHtmlTextPiece() {}
wxHtmlTextPiece(int pos, int lng) : m_pos(pos), m_lng(lng) {} wxHtmlTextPiece(const wxString::const_iterator& start,
int m_pos, m_lng; const wxString::const_iterator& end)
: m_start(start), m_end(end) {}
wxString::const_iterator m_start, m_end;
}; };
// NB: this is an empty class and not typedef because of forward declaration // NB: this is an empty class and not typedef because of forward declaration
@@ -64,7 +66,7 @@ public:
wxHtmlTag *m_tags; wxHtmlTag *m_tags;
wxHtmlTextPieces *m_textPieces; wxHtmlTextPieces *m_textPieces;
int m_curTextPiece; int m_curTextPiece;
wxString m_source; const wxString *m_source;
wxHtmlParserState *m_nextState; wxHtmlParserState *m_nextState;
}; };
@@ -78,6 +80,7 @@ wxHtmlParser::wxHtmlParser()
: wxObject(), m_HandlersHash(wxKEY_STRING), : wxObject(), m_HandlersHash(wxKEY_STRING),
m_FS(NULL), m_HandlersStack(NULL) m_FS(NULL), m_HandlersStack(NULL)
{ {
m_Source = NULL;
m_entitiesParser = new wxHtmlEntitiesParser; m_entitiesParser = new wxHtmlEntitiesParser;
m_Tags = NULL; m_Tags = NULL;
m_CurTag = NULL; m_CurTag = NULL;
@@ -103,6 +106,7 @@ wxHtmlParser::~wxHtmlParser()
m_HandlersHash.Clear(); m_HandlersHash.Clear();
WX_CLEAR_LIST(wxList, m_HandlersList); WX_CLEAR_LIST(wxList, m_HandlersList);
delete m_entitiesParser; delete m_entitiesParser;
delete m_Source;
} }
wxObject* wxHtmlParser::Parse(const wxString& source) wxObject* wxHtmlParser::Parse(const wxString& source)
@@ -128,7 +132,10 @@ void wxHtmlParser::DoneParser()
void wxHtmlParser::SetSource(const wxString& src) void wxHtmlParser::SetSource(const wxString& src)
{ {
DestroyDOMTree(); DestroyDOMTree();
m_Source = src; // NB: this is allocated on heap because wxHtmlTag keeps a pointer to
// this string if WXWIN_COMPATIBILITY_2_8
delete m_Source;
m_Source = new wxString(src);
CreateDOMTree(); CreateDOMTree();
m_CurTag = NULL; m_CurTag = NULL;
m_CurTextPiece = 0; m_CurTextPiece = 0;
@@ -136,54 +143,53 @@ void wxHtmlParser::SetSource(const wxString& src)
void wxHtmlParser::CreateDOMTree() void wxHtmlParser::CreateDOMTree()
{ {
wxHtmlTagsCache cache(m_Source); wxHtmlTagsCache cache(*m_Source);
m_TextPieces = new wxHtmlTextPieces; m_TextPieces = new wxHtmlTextPieces;
CreateDOMSubTree(NULL, 0, m_Source.length(), &cache); CreateDOMSubTree(NULL, m_Source->begin(), m_Source->end(), &cache);
m_CurTextPiece = 0; m_CurTextPiece = 0;
} }
extern bool wxIsCDATAElement(const wxChar *tag); extern bool wxIsCDATAElement(const wxString& tag);
void wxHtmlParser::CreateDOMSubTree(wxHtmlTag *cur, void wxHtmlParser::CreateDOMSubTree(wxHtmlTag *cur,
int begin_pos, int end_pos, const wxString::const_iterator& begin_pos,
const wxString::const_iterator& end_pos,
wxHtmlTagsCache *cache) wxHtmlTagsCache *cache)
{ {
if (end_pos <= begin_pos) return; if (end_pos <= begin_pos)
return;
wxChar c; wxChar c;
int i = begin_pos; wxString::const_iterator i = begin_pos;
int textBeginning = begin_pos; wxString::const_iterator textBeginning = begin_pos;
// If the tag contains CDATA text, we include the text between beginning // 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 // 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 // end of this function where text piece is added, bypassing any child
// tags parsing (CDATA element can't have child elements by definition): // tags parsing (CDATA element can't have child elements by definition):
if (cur != NULL && wxIsCDATAElement(cur->GetName().c_str())) if (cur != NULL && wxIsCDATAElement(cur->GetName()))
{ {
i = end_pos; i = end_pos;
} }
while (i < end_pos) while (i < end_pos)
{ {
c = m_Source.GetChar(i); c = *i;
if (c == wxT('<')) if (c == wxT('<'))
{ {
// add text to m_TextPieces: // add text to m_TextPieces:
if (i - textBeginning > 0) if (i > textBeginning)
m_TextPieces->push_back( m_TextPieces->push_back(wxHtmlTextPiece(textBeginning, i));
wxHtmlTextPiece(textBeginning, i - textBeginning));
// if it is a comment, skip it: // if it is a comment, skip it:
wxString::const_iterator iter = m_Source.begin() + i; if ( SkipCommentTag(i, m_Source->end()) )
if ( SkipCommentTag(iter, m_Source.end()) )
{ {
textBeginning = textBeginning = i = i + 1; // skip closing '>' too
i = iter - m_Source.begin() + 1; // skip closing '>' too
} }
// add another tag to the tree: // add another tag to the tree:
else if (i < end_pos-1 && m_Source.GetChar(i+1) != wxT('/')) else if (i < end_pos-1 && *(i+1) != wxT('/'))
{ {
wxHtmlTag *chd; wxHtmlTag *chd;
if (cur) if (cur)
@@ -211,12 +217,12 @@ void wxHtmlParser::CreateDOMSubTree(wxHtmlTag *cur,
if (chd->HasEnding()) if (chd->HasEnding())
{ {
CreateDOMSubTree(chd, CreateDOMSubTree(chd,
chd->GetBeginPos(), chd->GetEndPos1(), chd->GetBeginIter(), chd->GetEndIter1(),
cache); cache);
i = chd->GetEndPos2(); i = chd->GetEndIter2();
} }
else else
i = chd->GetBeginPos(); i = chd->GetBeginIter();
textBeginning = i; textBeginning = i;
} }
@@ -224,17 +230,16 @@ void wxHtmlParser::CreateDOMSubTree(wxHtmlTag *cur,
// ... or skip ending tag: // ... or skip ending tag:
else else
{ {
while (i < end_pos && m_Source.GetChar(i) != wxT('>')) i++; while (i < end_pos && *i != wxT('>')) ++i;
textBeginning = i+1; textBeginning = i+1;
} }
} }
else i++; else ++i;
} }
// add remaining text to m_TextPieces: // add remaining text to m_TextPieces:
if (end_pos - textBeginning > 0) if (end_pos > textBeginning)
m_TextPieces->push_back( m_TextPieces->push_back(wxHtmlTextPiece(textBeginning, end_pos));
wxHtmlTextPiece(textBeginning, end_pos - textBeginning));
} }
void wxHtmlParser::DestroyDOMTree() void wxHtmlParser::DestroyDOMTree()
@@ -257,42 +262,45 @@ void wxHtmlParser::DoParsing()
{ {
m_CurTag = m_Tags; m_CurTag = m_Tags;
m_CurTextPiece = 0; m_CurTextPiece = 0;
DoParsing(0, m_Source.length()); DoParsing(m_Source->begin(), m_Source->end());
} }
void wxHtmlParser::DoParsing(int begin_pos, int end_pos) void wxHtmlParser::DoParsing(const wxString::const_iterator& begin_pos_,
const wxString::const_iterator& end_pos)
{ {
if (end_pos <= begin_pos) return; wxString::const_iterator begin_pos(begin_pos_);
if (end_pos <= begin_pos)
return;
wxHtmlTextPieces& pieces = *m_TextPieces; wxHtmlTextPieces& pieces = *m_TextPieces;
size_t piecesCnt = pieces.size(); size_t piecesCnt = pieces.size();
while (begin_pos < end_pos) while (begin_pos < end_pos)
{ {
while (m_CurTag && m_CurTag->GetBeginPos() < begin_pos) while (m_CurTag && m_CurTag->GetBeginIter() < begin_pos)
m_CurTag = m_CurTag->GetNextTag(); m_CurTag = m_CurTag->GetNextTag();
while (m_CurTextPiece < piecesCnt && while (m_CurTextPiece < piecesCnt &&
pieces[m_CurTextPiece].m_pos < begin_pos) pieces[m_CurTextPiece].m_start < begin_pos)
m_CurTextPiece++; m_CurTextPiece++;
if (m_CurTextPiece < piecesCnt && if (m_CurTextPiece < piecesCnt &&
(!m_CurTag || (!m_CurTag ||
pieces[m_CurTextPiece].m_pos < m_CurTag->GetBeginPos())) pieces[m_CurTextPiece].m_start < m_CurTag->GetBeginIter()))
{ {
// Add text: // Add text:
AddText(GetEntitiesParser()->Parse( AddText(GetEntitiesParser()->Parse(
m_Source.Mid(pieces[m_CurTextPiece].m_pos, wxString(pieces[m_CurTextPiece].m_start,
pieces[m_CurTextPiece].m_lng))); pieces[m_CurTextPiece].m_end)));
begin_pos = pieces[m_CurTextPiece].m_pos + begin_pos = pieces[m_CurTextPiece].m_end;
pieces[m_CurTextPiece].m_lng;
m_CurTextPiece++; m_CurTextPiece++;
} }
else if (m_CurTag) else if (m_CurTag)
{ {
if (m_CurTag->HasEnding()) if (m_CurTag->HasEnding())
begin_pos = m_CurTag->GetEndPos2(); begin_pos = m_CurTag->GetEndIter2();
else else
begin_pos = m_CurTag->GetBeginPos(); begin_pos = m_CurTag->GetBeginIter();
wxHtmlTag *t = m_CurTag; wxHtmlTag *t = m_CurTag;
m_CurTag = m_CurTag->GetNextTag(); m_CurTag = m_CurTag->GetNextTag();
AddTag(*t); AddTag(*t);
@@ -318,7 +326,7 @@ void wxHtmlParser::AddTag(const wxHtmlTag& tag)
if (!inner) if (!inner)
{ {
if (tag.HasEnding()) if (tag.HasEnding())
DoParsing(tag.GetBeginPos(), tag.GetEndPos1()); DoParsing(tag.GetBeginIter(), tag.GetEndIter1());
} }
} }
@@ -393,7 +401,7 @@ void wxHtmlParser::SetSourceAndSaveState(const wxString& src)
m_Tags = NULL; m_Tags = NULL;
m_TextPieces = NULL; m_TextPieces = NULL;
m_CurTextPiece = 0; m_CurTextPiece = 0;
m_Source = wxEmptyString; m_Source = NULL;
SetSource(src); SetSource(src);
} }
@@ -419,8 +427,7 @@ bool wxHtmlParser::RestoreState()
wxString wxHtmlParser::GetInnerSource(const wxHtmlTag& tag) wxString wxHtmlParser::GetInnerSource(const wxHtmlTag& tag)
{ {
return GetSource()->Mid(tag.GetBeginPos(), return wxString(tag.GetBeginIter(), tag.GetEndIter1());
tag.GetEndPos1() - tag.GetBeginPos());
} }
//----------------------------------------------------------------------------- //-----------------------------------------------------------------------------
@@ -501,11 +508,16 @@ wxString wxHtmlEntitiesParser::Parse(const wxString& input) const
const wxString::const_iterator ent_s = c; const wxString::const_iterator ent_s = c;
wxChar entity_char; wxChar entity_char;
for (; c != end && for ( ; c != end; ++c )
((*c >= wxT('a') && *c <= wxT('z')) || {
(*c >= wxT('A') && *c <= wxT('Z')) || wxChar ch = *c;
(*c >= wxT('0') && *c <= wxT('9')) || if ( !((ch >= wxT('a') && ch <= wxT('z')) ||
*c == wxT('_') || *c == wxT('#')); ++c) {} (ch >= wxT('A') && ch <= wxT('Z')) ||
(ch >= wxT('0') && ch <= wxT('9')) ||
ch == wxT('_') || ch == wxT('#')) )
break;
}
entity.append(ent_s, c); entity.append(ent_s, c);
if (c == end || *c != wxT(';')) --c; if (c == end || *c != wxT(';')) --c;
last = c+1; last = c+1;

138
src/html/htmltag.cpp
View File

@@ -36,14 +36,21 @@ struct wxHtmlCacheItem
{ {
// this is "pos" value passed to wxHtmlTag's constructor. // this is "pos" value passed to wxHtmlTag's constructor.
// it is position of '<' character of the tag // 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: // end positions for the tag:
// end1 is '<' of ending tag, // end1 is '<' of ending tag,
// end2 is '>' or both are // end2 is '>' or both are
// -1 if there is no ending tag for this one... wxString::const_iterator End1, End2;
// or -2 if this is ending tag </...>
int End1, End2;
// name of this tag // name of this tag
wxChar *Name; wxChar *Name;
@@ -60,95 +67,108 @@ bool wxIsCDATAElement(const wxChar *tag)
(wxStrcmp(tag, _T("STYLE")) == 0); (wxStrcmp(tag, _T("STYLE")) == 0);
} }
bool wxIsCDATAElement(const wxString& tag)
{
return (wxStrcmp(tag.wx_str(), wxSTRING_TEXT("SCRIPT")) == 0) ||
(wxStrcmp(tag.wx_str(), wxSTRING_TEXT("STYLE")) == 0);
}
wxHtmlTagsCache::wxHtmlTagsCache(const wxString& source) wxHtmlTagsCache::wxHtmlTagsCache(const wxString& source)
{ {
m_Cache = new wxHtmlTagsCacheData; m_Cache = new wxHtmlTagsCacheData;
m_CachePos = 0; m_CachePos = 0;
const wxChar *src = source.c_str();
int lng = source.length();
wxChar tagBuffer[256]; 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 // don't cache comment tags
wxString::const_iterator iter = source.begin() + pos; if ( wxHtmlParser::SkipCommentTag(pos, source.end()) )
if ( wxHtmlParser::SkipCommentTag(iter, source.end()) )
{
pos = iter - source.begin();
continue; continue;
}
size_t tg = Cache().size(); size_t tg = Cache().size();
Cache().push_back(wxHtmlCacheItem()); Cache().push_back(wxHtmlCacheItem());
int stpos = pos++; wxString::const_iterator stpos = pos++;
Cache()[tg].Key = stpos; Cache()[tg].Key = stpos;
int i; int i;
for ( i = 0; for ( i = 0;
pos < lng && i < (int)WXSIZEOF(tagBuffer) - 1 && pos < end && i < (int)WXSIZEOF(tagBuffer) - 1 &&
src[pos] != wxT('>') && !wxIsspace(src[pos]); *pos != wxT('>') && !wxIsspace(*pos);
i++, pos++ ) ++i, ++pos )
{ {
tagBuffer[i] = (wxChar)wxToupper(src[pos]); tagBuffer[i] = (wxChar)wxToupper(*pos);
} }
tagBuffer[i] = _T('\0'); tagBuffer[i] = _T('\0');
Cache()[tg].Name = new wxChar[i+1]; Cache()[tg].Name = new wxChar[i+1];
memcpy(Cache()[tg].Name, tagBuffer, (i+1)*sizeof(wxChar)); 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: // find matching begin tag:
for (i = tg; i >= 0; i--) 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].End1 = stpos;
Cache()[i].End2 = pos + 1; Cache()[i].End2 = pos + 1;
break; break;
} }
}
} }
else else
{ {
Cache()[tg].End1 = Cache()[tg].End2 = -1; Cache()[tg].type = wxHtmlCacheItem::Type_NoMatchingEndingTag;
if (wxIsCDATAElement(tagBuffer)) if (wxIsCDATAElement(tagBuffer))
{ {
// store the orig pos in case we are missing the closing // store the orig pos in case we are missing the closing
// tag (see below) // tag (see below)
wxInt32 old_pos = pos; const wxString::const_iterator old_pos = pos;
bool foundCloseTag = false; bool foundCloseTag = false;
// find next matching tag // find next matching tag
int tag_len = wxStrlen(tagBuffer); int tag_len = wxStrlen(tagBuffer);
while (pos < lng) while (pos < end)
{ {
// find the ending tag // find the ending tag
while (pos + 1 < lng && while (pos + 1 < end &&
(src[pos] != '<' || src[pos+1] != '/')) (*pos != '<' || *(pos+1) != '/'))
++pos; ++pos;
if (src[pos] == '<') if (*pos == '<')
++pos; ++pos;
// see if it matches // see if it matches
int match_pos = 0; 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 // cast to wxChar needed to suppress warning in
// Unicode build // Unicode build
if ((wxChar)wxToupper(src[pos]) == tagBuffer[match_pos]) { if ((wxChar)wxToupper(c) == tagBuffer[match_pos])
{
++match_pos; ++match_pos;
} }
else if (src[pos] == wxT(' ') || src[pos] == wxT('\n') || else if (c == wxT(' ') || c == wxT('\n') ||
src[pos] == wxT('\r') || src[pos] == wxT('\t')) { c == wxT('\r') || c == wxT('\t'))
{
// need to skip over these // need to skip over these
} }
else { else
{
match_pos = 0; match_pos = 0;
} }
++pos; ++pos;
@@ -193,7 +213,11 @@ wxHtmlTagsCache::~wxHtmlTagsCache()
delete m_Cache; 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()) if (Cache().empty())
return; return;
@@ -203,22 +227,28 @@ void wxHtmlTagsCache::QueryTag(int at, int* end1, int* end2)
int delta = (at < Cache()[m_CachePos].Key) ? -1 : 1; int delta = (at < Cache()[m_CachePos].Key) ? -1 : 1;
do 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 // something is very wrong with HTML, give up by returning an
// impossibly large value which is going to be ignored by the // impossibly large value which is going to be ignored by the
// caller // caller
*end1 = *end1 =
*end2 = INT_MAX; *end2 = inputEnd;
*hasEnding = true;
return; return;
} }
m_CachePos += delta;
} }
while (Cache()[m_CachePos].Key != at); while (Cache()[m_CachePos].Key != at);
} }
*end1 = Cache()[m_CachePos].End1; *end1 = Cache()[m_CachePos].End1;
*end2 = Cache()[m_CachePos].End2; *end2 = Cache()[m_CachePos].End2;
*hasEnding = (Cache()[m_CachePos].type == wxHtmlCacheItem::Type_Normal);
} }
@@ -229,7 +259,9 @@ void wxHtmlTagsCache::QueryTag(int at, int* end1, int* end2)
//----------------------------------------------------------------------------- //-----------------------------------------------------------------------------
wxHtmlTag::wxHtmlTag(wxHtmlTag *parent, 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, wxHtmlTagsCache *cache,
wxHtmlEntitiesParser *entParser) wxHtmlEntitiesParser *entParser)
{ {
@@ -252,15 +284,14 @@ wxHtmlTag::wxHtmlTag(wxHtmlTag *parent,
/* Find parameters and their values: */ /* Find parameters and their values: */
int i;
wxChar c; wxChar c;
// fill-in name, params and begin pos: // 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: // find tag's name and convert it to uppercase:
while ((i < end_pos) && 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('\n') && c != wxT('\t') &&
c != wxT('>'))) c != wxT('>')))
{ {
@@ -272,7 +303,7 @@ wxHtmlTag::wxHtmlTag(wxHtmlTag *parent,
// if the tag has parameters, read them and "normalize" them, // if the tag has parameters, read them and "normalize" them,
// i.e. convert to uppercase, replace whitespaces by spaces and // i.e. convert to uppercase, replace whitespaces by spaces and
// remove whitespaces around '=': // remove whitespaces around '=':
if (source[i-1] != wxT('>')) if (*(i-1) != wxT('>'))
{ {
#define IS_WHITE(c) (c == wxT(' ') || c == wxT('\r') || \ #define IS_WHITE(c) (c == wxT(' ') || c == wxT('\r') || \
c == wxT('\n') || c == wxT('\t')) c == wxT('\n') || c == wxT('\t'))
@@ -291,14 +322,14 @@ wxHtmlTag::wxHtmlTag(wxHtmlTag *parent,
state = ST_BEFORE_NAME; state = ST_BEFORE_NAME;
while (i < end_pos) while (i < end_pos)
{ {
c = source[i++]; c = *(i++);
if (c == wxT('>') && !(state == ST_VALUE && quote != 0)) if (c == wxT('>') && !(state == ST_VALUE && quote != 0))
{ {
if (state == ST_BEFORE_EQ || state == ST_NAME) if (state == ST_BEFORE_EQ || state == ST_NAME)
{ {
m_ParamNames.Add(pname); m_ParamNames.Add(pname);
m_ParamValues.Add(wxEmptyString); m_ParamValues.Add(wxGetEmptyString());
} }
else if (state == ST_VALUE && quote == 0) else if (state == ST_VALUE && quote == 0)
{ {
@@ -333,7 +364,7 @@ wxHtmlTag::wxHtmlTag(wxHtmlTag *parent,
else if (!IS_WHITE(c)) else if (!IS_WHITE(c))
{ {
m_ParamNames.Add(pname); m_ParamNames.Add(pname);
m_ParamValues.Add(wxEmptyString); m_ParamValues.Add(wxGetEmptyString());
pname = c; pname = c;
state = ST_NAME; state = ST_NAME;
} }
@@ -342,7 +373,7 @@ wxHtmlTag::wxHtmlTag(wxHtmlTag *parent,
if (!IS_WHITE(c)) if (!IS_WHITE(c))
{ {
if (c == wxT('"') || c == wxT('\'')) if (c == wxT('"') || c == wxT('\''))
quote = c, pvalue = wxEmptyString; quote = c, pvalue = wxGetEmptyString();
else else
quote = 0, pvalue = c; quote = 0, pvalue = c;
state = ST_VALUE; state = ST_VALUE;
@@ -374,10 +405,13 @@ wxHtmlTag::wxHtmlTag(wxHtmlTag *parent,
#undef IS_WHITE #undef IS_WHITE
} }
m_Begin = i; m_Begin = i;
cache->QueryTag(pos, source->end(), &m_End1, &m_End2, &m_hasEnding);
cache->QueryTag(pos, &m_End1, &m_End2);
if (m_End1 > end_pos) m_End1 = end_pos; if (m_End1 > end_pos) m_End1 = end_pos;
if (m_End2 > end_pos) m_End2 = end_pos; if (m_End2 > end_pos) m_End2 = end_pos;
#if WXWIN_COMPATIBILITY_2_8
m_sourceStart = source->begin();
#endif
} }
wxHtmlTag::~wxHtmlTag() wxHtmlTag::~wxHtmlTag()
@@ -401,7 +435,7 @@ 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) if (index == wxNOT_FOUND)
return wxEmptyString; return wxGetEmptyString();
if (with_commas) if (with_commas)
{ {
// VS: backward compatibility, seems to be never used by wxHTML... // VS: backward compatibility, seems to be never used by wxHTML...
@@ -438,8 +472,8 @@ bool wxHtmlTag::GetParamAsColour(const wxString& par, wxColour *clr) const
// handle colours defined in HTML 4.0 first: // handle colours defined in HTML 4.0 first:
if (str.length() > 1 && str[0] != _T('#')) if (str.length() > 1 && str[0] != _T('#'))
{ {
#define HTML_COLOUR(name, r, g, b) \ #define HTML_COLOUR(name, r, g, b) \
if (str.IsSameAs(wxT(name), false)) \ if (str.IsSameAs(wxSTRING_TEXT(name), false)) \
{ clr->Set(r, g, b); return true; } { clr->Set(r, g, b); return true; }
HTML_COLOUR("black", 0x00,0x00,0x00) HTML_COLOUR("black", 0x00,0x00,0x00)
HTML_COLOUR("silver", 0xC0,0xC0,0xC0) HTML_COLOUR("silver", 0xC0,0xC0,0xC0)

4
src/html/m_layout.cpp
View File

@@ -289,9 +289,7 @@ TAG_HANDLER_BEGIN(TITLE, "TITLE")
wxHtmlWindowInterface *winIface = m_WParser->GetWindowInterface(); wxHtmlWindowInterface *winIface = m_WParser->GetWindowInterface();
if (winIface) if (winIface)
{ {
wxString title = m_WParser->GetSource()->Mid( wxString title(tag.GetBeginIter(), tag.GetEndIter1());
tag.GetBeginPos(),
tag.GetEndPos1()-tag.GetBeginPos());
#if !wxUSE_UNICODE && wxUSE_WCHAR_T #if !wxUSE_UNICODE && wxUSE_WCHAR_T
const wxFontEncoding enc = m_WParser->GetInputEncoding(); const wxFontEncoding enc = m_WParser->GetInputEncoding();
if ( enc != wxFONTENCODING_DEFAULT ) if ( enc != wxFONTENCODING_DEFAULT )