]> git.saurik.com Git - wxWidgets.git/blobdiff - src/html/htmltag.cpp
Add backend descriptions for all backends to the documentation.
[wxWidgets.git] / src / html / htmltag.cpp
index 4365a26554316d4165255e2fb0322b4403650d9c..fdc71fe1710063e3212f899891ceaddb31a95973 100644 (file)
@@ -23,6 +23,8 @@
 #endif
 
 #include "wx/html/htmlpars.h"
 #endif
 
 #include "wx/html/htmlpars.h"
+#include "wx/html/styleparams.h"
+
 #include "wx/vector.h"
 
 #include <stdio.h> // for vsscanf
 #include "wx/vector.h"
 
 #include <stdio.h> // for vsscanf
@@ -63,14 +65,14 @@ class wxHtmlTagsCacheData : public wxVector<wxHtmlCacheItem>
 
 bool wxIsCDATAElement(const wxChar *tag)
 {
 
 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)
 {
 }
 
 bool wxIsCDATAElement(const wxString& tag)
 {
-    return (wxStrcmp(tag.wx_str(), wxSTRING_TEXT("SCRIPT")) == 0) ||
-           (wxStrcmp(tag.wx_str(), wxSTRING_TEXT("STYLE")) == 0);
+    return (wxStrcmp(tag.wx_str(), wxS("SCRIPT")) == 0) ||
+           (wxStrcmp(tag.wx_str(), wxS("STYLE")) == 0);
 }
 
 wxHtmlTagsCache::wxHtmlTagsCache(const wxString& source)
 }
 
 wxHtmlTagsCache::wxHtmlTagsCache(const wxString& source)
@@ -83,118 +85,132 @@ wxHtmlTagsCache::wxHtmlTagsCache(const wxString& source)
     const wxString::const_iterator end = source.end();
     for ( wxString::const_iterator pos = source.begin(); pos < end; ++pos )
     {
     const wxString::const_iterator end = source.end();
     for ( wxString::const_iterator pos = source.begin(); pos < end; ++pos )
     {
-        if (*pos == wxT('<'))   // tag found:
-        {
-            // don't cache comment tags
-            if ( wxHtmlParser::SkipCommentTag(pos, source.end()) )
-                continue;
+        if (*pos != wxT('<'))
+            continue;
 
 
-            size_t tg = Cache().size();
-            Cache().push_back(wxHtmlCacheItem());
+        // possible tag start found:
 
 
-            wxString::const_iterator stpos = pos++;
-            Cache()[tg].Key = stpos;
+        // don't cache comment tags
+        if ( wxHtmlParser::SkipCommentTag(pos, end) )
+            continue;
 
 
-            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] = _T('\0');
+        // 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');
 
 
-            Cache()[tg].Name = new wxChar[i+1];
-            memcpy(Cache()[tg].Name, tagBuffer, (i+1)*sizeof(wxChar));
+        while (pos < end && *pos != wxT('>'))
+            ++pos;
 
 
-            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:
+        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--)
             {
             {
-                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))
                 {
                 {
-                    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;
-                    }
+                    Cache()[i].type = wxHtmlCacheItem::Type_Normal;
+                    Cache()[i].End1 = stpos;
+                    Cache()[i].End2 = pos + 1;
+                    break;
                 }
             }
                 }
             }
-            else
-            {
-                Cache()[tg].type = wxHtmlCacheItem::Type_NoMatchingEndingTag;
+        }
+        else
+        {
+            Cache()[tg].type = wxHtmlCacheItem::Type_NoMatchingEndingTag;
 
 
-                if (wxIsCDATAElement(tagBuffer))
+            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)
                 {
                 {
-                    // 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 )
                     {
                     {
-                        // 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])
                         {
                         {
-                            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;
+                            ++match_pos;
                         }
                         }
-
-                        // found a match
-                        if (match_pos == tag_len)
+                        else if (c == wxT(' ') || c == wxT('\n') ||
+                            c == wxT('\r') || c == wxT('\t'))
                         {
                         {
-                            pos = pos - tag_len - 3;
-                            foundCloseTag = true;
-                            break;
+                            // need to skip over these
                         }
                         }
-                        else // keep looking for the closing tag
+                        else
                         {
                         {
-                            ++pos;
+                            match_pos = 0;
                         }
                         }
+                        ++pos;
+                    }
+
+                    // found a match
+                    if (match_pos == tag_len)
+                    {
+                        pos = pos - tag_len - 3;
+                        foundCloseTag = true;
+                        break;
                     }
                     }
-                    if (!foundCloseTag)
+                    else // keep looking for the closing tag
                     {
                     {
-                        // 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;
                     }
                 }
                     }
                 }
+                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;
+                }
             }
         }
     }
             }
         }
     }
@@ -203,8 +219,7 @@ wxHtmlTagsCache::wxHtmlTagsCache(const wxString& source)
     for ( wxHtmlTagsCacheData::iterator i = Cache().begin();
           i != Cache().end(); ++i )
     {
     for ( wxHtmlTagsCacheData::iterator i = Cache().begin();
           i != Cache().end(); ++i )
     {
-        delete[] i->Name;
-        i->Name = NULL;
+        wxDELETEA(i->Name);
     }
 }
 
     }
 }
 
@@ -220,7 +235,12 @@ void wxHtmlTagsCache::QueryTag(const wxString::const_iterator& at,
                                bool *hasEnding)
 {
     if (Cache().empty())
                                bool *hasEnding)
 {
     if (Cache().empty())
+    {
+        *end1 =
+        *end2 = inputEnd;
+        *hasEnding = true;
         return;
         return;
+    }
 
     if (Cache()[m_CachePos].Key != at)
     {
 
     if (Cache()[m_CachePos].Key != at)
     {
@@ -246,9 +266,27 @@ void wxHtmlTagsCache::QueryTag(const wxString::const_iterator& at,
         }
         while (Cache()[m_CachePos].Key != at);
     }
         }
         while (Cache()[m_CachePos].Key != at);
     }
-    *end1 = Cache()[m_CachePos].End1;
-    *end2 = Cache()[m_CachePos].End2;
-    *hasEnding = (Cache()[m_CachePos].type == wxHtmlCacheItem::Type_Normal);
+
+    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;
+    }
 }
 
 
 }
 
 
@@ -284,7 +322,7 @@ wxHtmlTag::wxHtmlTag(wxHtmlTag *parent,
 
     /* Find parameters and their values: */
 
 
     /* Find parameters and their values: */
 
-    wxChar c;
+    wxChar c wxDUMMY_INITIALIZE(0);
 
     // fill-in name, params and begin pos:
     wxString::const_iterator i(pos+1);
 
     // fill-in name, params and begin pos:
     wxString::const_iterator i(pos+1);
@@ -293,7 +331,7 @@ wxHtmlTag::wxHtmlTag(wxHtmlTag *parent,
     while ((i < end_pos) &&
            ((c = *(i++)) != wxT(' ') && c != wxT('\r') &&
              c != wxT('\n') && c != wxT('\t') &&
     while ((i < end_pos) &&
            ((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 ((c >= wxT('a')) && (c <= wxT('z')))
             c -= (wxT('a') - wxT('A'));
@@ -412,6 +450,32 @@ wxHtmlTag::wxHtmlTag(wxHtmlTag *parent,
 #if WXWIN_COMPATIBILITY_2_8
     m_sourceStart = source->begin();
 #endif
 #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()
 }
 
 wxHtmlTag::~wxHtmlTag()
@@ -431,12 +495,12 @@ 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
+wxString wxHtmlTag::GetParam(const wxString& par, bool with_quotes) const
 {
     int index = m_ParamNames.Index(par, false);
     if (index == wxNOT_FOUND)
         return wxGetEmptyString();
 {
     int index = m_ParamNames.Index(par, false);
     if (index == wxNOT_FOUND)
         return wxGetEmptyString();
-    if (with_commas)
+    if (with_quotes)
     {
         // VS: backward compatibility, seems to be never used by wxHTML...
         wxString s;
     {
         // VS: backward compatibility, seems to be never used by wxHTML...
         wxString s;
@@ -463,17 +527,16 @@ int wxHtmlTag::ScanParam(const wxString& par,
     return wxSscanf(parval, format, param);
 }
 
     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:
 
     // 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(wxSTRING_TEXT(name), false))         \
+        #define HTML_COLOUR(name, r, g, b)              \
+            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)
                 { clr->Set(r, g, b); return true; }
         HTML_COLOUR("black",   0x00,0x00,0x00)
         HTML_COLOUR("silver",  0xC0,0xC0,0xC0)
@@ -504,13 +567,23 @@ bool wxHtmlTag::GetParamAsColour(const wxString& par, wxColour *clr) const
     return false;
 }
 
     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
 {
 bool wxHtmlTag::GetParamAsInt(const wxString& par, int *clr) const
 {
-    if (!HasParam(par)) return false;
+    if ( !HasParam(par) )
+        return false;
+
     long i;
     long i;
-    bool succ = GetParam(par).ToLong(&i);
+    if ( !GetParam(par).ToLong(&i) )
+        return false;
+
     *clr = (int)i;
     *clr = (int)i;
-    return succ;
+    return true;
 }
 
 wxString wxHtmlTag::GetAllParams() const
 }
 
 wxString wxHtmlTag::GetAllParams() const