X-Git-Url: https://git.saurik.com/wxWidgets.git/blobdiff_plain/c989314692827562d3d2147ae880886b5eaa530e..1ae82ba9f23ee688bc4db2cd449eca5bf97cfdf0:/src/html/htmltag.cpp diff --git a/src/html/htmltag.cpp b/src/html/htmltag.cpp index efc57a674d..6686cb9cba 100644 --- a/src/html/htmltag.cpp +++ b/src/html/htmltag.cpp @@ -1,31 +1,27 @@ ///////////////////////////////////////////////////////////////////////////// -// Name: htmltag.cpp +// Name: src/html/htmltag.cpp // Purpose: wxHtmlTag class (represents single tag) // 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 -#ifdef __BORDLANDC__ -#pragma hdrstop -#endif +#include "wx/html/htmltag.h" #ifndef WXPRECOMP -#include "wx/wx.h" + #include "wx/colour.h" #endif -#include "wx/html/htmltag.h" +#include "wx/html/htmlpars.h" #include // for vsscanf #include @@ -56,57 +52,121 @@ IMPLEMENT_CLASS(wxHtmlTagsCache,wxObject) #define CACHE_INCREMENT 64 +bool wxIsCDATAElement(const wxChar *tag) +{ + return (wxStrcmp(tag, _T("SCRIPT")) == 0) || + (wxStrcmp(tag, _T("STYLE")) == 0); +} + wxHtmlTagsCache::wxHtmlTagsCache(const wxString& source) { const wxChar *src = source.c_str(); - int i, tg, pos, stpos; - int lng = source.Length(); - wxChar dummy[256]; + int lng = source.length(); + wxChar tagBuffer[256]; m_Cache = NULL; m_CacheSize = 0; m_CachePos = 0; - pos = 0; - while (pos < lng) + int pos = 0; + while (pos < lng) { if (src[pos] == wxT('<')) // tag found: - { + { if (m_CacheSize % CACHE_INCREMENT == 0) m_Cache = (wxHtmlCacheItem*) realloc(m_Cache, (m_CacheSize + CACHE_INCREMENT) * sizeof(wxHtmlCacheItem)); - tg = m_CacheSize++; - m_Cache[tg].Key = stpos = pos++; - dummy[0] = 0; i = 0; - while (pos < lng && - src[pos] != wxT('>') && - src[pos] != wxT(' ') && src[pos] != wxT('\r') && - src[pos] != wxT('\n') && src[pos] != wxT('\t')) - { - dummy[i] = src[pos++]; - if ((dummy[i] >= wxT('a')) && (dummy[i] <= wxT('z'))) dummy[i] -= (wxT('a') - wxT('A')); - i++; + 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]); } - dummy[i] = 0; + tagBuffer[i] = _T('\0'); + m_Cache[tg].Name = new wxChar[i+1]; - memcpy(m_Cache[tg].Name, dummy, (i+1)*sizeof(wxChar)); + 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, dummy+1) == 0)) - { + 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 - { + 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; + } + } } } @@ -114,22 +174,34 @@ wxHtmlTagsCache::wxHtmlTagsCache(const wxString& source) } // ok, we're done, now we'll free .Name members of cache - we don't need it anymore: - for (i = 0; i < m_CacheSize; i++) + for (int i = 0; i < m_CacheSize; i++) { delete[] m_Cache[i].Name; m_Cache[i].Name = NULL; } } - - void wxHtmlTagsCache::QueryTag(int at, int* end1, int* end2) { if (m_Cache == NULL) return; - if (m_Cache[m_CachePos].Key != at) + if (m_Cache[m_CachePos].Key != at) { int delta = (at < m_Cache[m_CachePos].Key) ? -1 : 1; - do {m_CachePos += delta;} while (m_Cache[m_CachePos].Key != at); + do + { + if ( m_CachePos < 0 || m_CachePos == m_CacheSize ) + { + // 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; + return; + } + + m_CachePos += delta; + } + while (m_Cache[m_CachePos].Key != at); } *end1 = m_Cache[m_CachePos].End1; *end2 = m_Cache[m_CachePos].End2; @@ -144,178 +216,297 @@ void wxHtmlTagsCache::QueryTag(int at, int* end1, int* end2) IMPLEMENT_CLASS(wxHtmlTag,wxObject) -wxHtmlTag::wxHtmlTag(const wxString& source, int pos, int end_pos, wxHtmlTagsCache* cache) : wxObject() +wxHtmlTag::wxHtmlTag(wxHtmlTag *parent, + const wxString& source, int pos, int end_pos, + wxHtmlTagsCache *cache, + wxHtmlEntitiesParser *entParser) : wxObject() { + /* Setup DOM relations */ + + m_Next = NULL; + m_FirstChild = m_LastChild = NULL; + m_Parent = parent; + if (parent) + { + m_Prev = m_Parent->m_LastChild; + if (m_Prev == NULL) + m_Parent->m_FirstChild = this; + else + m_Prev->m_Next = this; + m_Parent->m_LastChild = this; + } + else + m_Prev = NULL; + + /* Find parameters and their values: */ + int i; - char c; + wxChar c; // fill-in name, params and begin pos: - m_Name = m_Params = wxEmptyString; i = pos+1; - if (source[i] == wxT('/')) { m_Ending = TRUE; i++; } - else m_Ending = FALSE; // find tag's name and convert it to uppercase: - while ((i < end_pos) && - ((c = source[i++]) != wxT(' ') && c != wxT('\r') && - c != wxT('\n') && c != wxT('\t') && - c != wxT('>'))) - { - if ((c >= wxT('a')) && (c <= wxT('z'))) c -= (wxT('a') - wxT('A')); - m_Name += c; + while ((i < end_pos) && + ((c = source[i++]) != wxT(' ') && c != wxT('\r') && + c != wxT('\n') && c != wxT('\t') && + c != wxT('>'))) + { + if ((c >= wxT('a')) && (c <= wxT('z'))) + c -= (wxT('a') - wxT('A')); + m_Name << c; } // 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 '=': if (source[i-1] != wxT('>')) - while ((i < end_pos) && ((c = source[i++]) != wxT('>'))) - { - if ((c >= wxT('a')) && (c <= wxT('z'))) - c -= (wxT('a') - wxT('A')); - if (c == wxT('\r') || c == wxT('\n') || c == wxT('\t')) - c = wxT(' '); // make future parsing a bit simpler - m_Params += c; - if (c == wxT('"')) - { - // remove spaces around the '=' character: - if (m_Params.Length() > 1 && - m_Params[m_Params.Length()-2] == wxT(' ')) + { + #define IS_WHITE(c) (c == wxT(' ') || c == wxT('\r') || \ + c == wxT('\n') || c == wxT('\t')) + wxString pname, pvalue; + wxChar quote; + enum + { + ST_BEFORE_NAME = 1, + ST_NAME, + ST_BEFORE_EQ, + ST_BEFORE_VALUE, + ST_VALUE + } state; + + quote = 0; + state = ST_BEFORE_NAME; + while (i < end_pos) + { + c = source[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); + } + else if (state == ST_VALUE && quote == 0) { - m_Params.RemoveLast(); - while (m_Params.Length() > 0 && m_Params.Last() == wxT(' ')) - m_Params.RemoveLast(); - m_Params += wxT('"'); + m_ParamNames.Add(pname); + if (entParser) + m_ParamValues.Add(entParser->Parse(pvalue)); + else + m_ParamValues.Add(pvalue); } - while ((i < end_pos) && (source[i++] == wxT(' '))) {} - if (i < end_pos) i--; - - // ...and copy the value to m_Params: - while ((i < end_pos) && ((c = source[i++]) != wxT('"'))) - m_Params += c; - m_Params += c; + break; } - else if (c == wxT('\'')) - { - while ((i < end_pos) && ((c = source[i++]) != wxT('\''))) - m_Params += c; - m_Params += c; + switch (state) + { + case ST_BEFORE_NAME: + if (!IS_WHITE(c)) + { + pname = c; + state = ST_NAME; + } + break; + case ST_NAME: + if (IS_WHITE(c)) + state = ST_BEFORE_EQ; + else if (c == wxT('=')) + state = ST_BEFORE_VALUE; + else + pname << c; + break; + case ST_BEFORE_EQ: + if (c == wxT('=')) + state = ST_BEFORE_VALUE; + else if (!IS_WHITE(c)) + { + m_ParamNames.Add(pname); + m_ParamValues.Add(wxEmptyString); + pname = c; + state = ST_NAME; + } + break; + case ST_BEFORE_VALUE: + if (!IS_WHITE(c)) + { + if (c == wxT('"') || c == wxT('\'')) + quote = c, pvalue = wxEmptyString; + else + quote = 0, pvalue = c; + state = ST_VALUE; + } + break; + case ST_VALUE: + if ((quote != 0 && c == quote) || + (quote == 0 && IS_WHITE(c))) + { + m_ParamNames.Add(pname); + if (quote == 0) + { + // VS: backward compatibility, no real reason, + // but wxHTML code relies on this... :( + pvalue.MakeUpper(); + } + if (entParser) + m_ParamValues.Add(entParser->Parse(pvalue)); + else + m_ParamValues.Add(pvalue); + state = ST_BEFORE_NAME; + } + else + pvalue << c; + break; } } - m_Begin = i; - cache->QueryTag(pos, &m_End1, &m_End2); - if (m_End1 > end_pos) m_End1 = end_pos; - if (m_End2 > end_pos) m_End2 = end_pos; + #undef IS_WHITE + } + m_Begin = i; + + cache->QueryTag(pos, &m_End1, &m_End2); + if (m_End1 > end_pos) m_End1 = end_pos; + if (m_End2 > end_pos) m_End2 = end_pos; } - +wxHtmlTag::~wxHtmlTag() +{ + wxHtmlTag *t1, *t2; + t1 = m_FirstChild; + while (t1) + { + t2 = t1->GetNextSibling(); + delete t1; + t1 = t2; + } +} bool wxHtmlTag::HasParam(const wxString& par) const { - const wxChar *st = m_Params, *p = par; - const wxChar *st2, *p2; - const wxChar invalid = wxT(1); + return (m_ParamNames.Index(par, false) != wxNOT_FOUND); +} - if (*st == 0) return FALSE; - if (*p == 0) return FALSE; - for (st2 = st, p2 = p; ; st2++) +wxString wxHtmlTag::GetParam(const wxString& par, bool with_commas) const +{ + int index = m_ParamNames.Index(par, false); + if (index == wxNOT_FOUND) + return wxEmptyString; + if (with_commas) { - if (*p2 == 0) return TRUE; - if (*st2 == 0) return FALSE; - if (*p2 != *st2) p2 = &invalid; - if (*p2 == *st2) p2++; - if (*st2 == wxT(' ')) p2 = p; - else if (*st2 == wxT('=')) - { - p2 = p; - while (*st2 != wxT(' ')) - { - if (*st2 == wxT('"')) - { - st2++; - while (*st2 != wxT('"')) st2++; - } - st2++; - if (*st2 == 0) return FALSE; - } - } + // VS: backward compatibility, seems to be never used by wxHTML... + wxString s; + s << wxT('"') << m_ParamValues[index] << wxT('"'); + return s; } + else + return m_ParamValues[index]; } +int wxHtmlTag::ScanParam(const wxString& par, + const wxChar *format, + void *param) const +{ + wxString parval = GetParam(par); + return wxSscanf(parval, format, param); +} +bool wxHtmlTag::GetParamAsColour(const wxString& par, wxColour *clr) const +{ + wxASSERT(clr); + wxString str = GetParam(par); -wxString wxHtmlTag::GetParam(const wxString& par, bool with_commas) const + if (clr->Set(str)) + return true; + + if (!str.empty()) + { + // Handle colours defined in HTML 4.0: + #define HTML_COLOUR(name,r,g,b) \ + if (str.IsSameAs(wxT(name), false)) \ + { *clr = wxColour(r,g,b); return true; } + HTML_COLOUR("black", 0x00,0x00,0x00) + HTML_COLOUR("silver", 0xC0,0xC0,0xC0) + HTML_COLOUR("gray", 0x80,0x80,0x80) + HTML_COLOUR("white", 0xFF,0xFF,0xFF) + HTML_COLOUR("maroon", 0x80,0x00,0x00) + HTML_COLOUR("red", 0xFF,0x00,0x00) + HTML_COLOUR("purple", 0x80,0x00,0x80) + HTML_COLOUR("fuchsia", 0xFF,0x00,0xFF) + HTML_COLOUR("green", 0x00,0x80,0x00) + HTML_COLOUR("lime", 0x00,0xFF,0x00) + HTML_COLOUR("olive", 0x80,0x80,0x00) + HTML_COLOUR("yellow", 0xFF,0xFF,0x00) + HTML_COLOUR("navy", 0x00,0x00,0x80) + HTML_COLOUR("blue", 0x00,0x00,0xFF) + HTML_COLOUR("teal", 0x00,0x80,0x80) + HTML_COLOUR("aqua", 0x00,0xFF,0xFF) + #undef HTML_COLOUR + } + + return false; +} + +bool wxHtmlTag::GetParamAsInt(const wxString& par, int *clr) const { - const wxChar *st = m_Params, *p = par; - const wxChar *st2, *p2; - const wxChar invalid = wxT(1); - bool comma; - wxChar comma_char; - - if (*st == 0) return wxEmptyString; - if (*p == 0) return wxEmptyString; - for (st2 = st, p2 = p; ; st2++) + if (!HasParam(par)) return false; + long i; + bool succ = GetParam(par).ToLong(&i); + *clr = (int)i; + return succ; +} + +wxString wxHtmlTag::GetAllParams() const +{ + // VS: this function is for backward compatibility only, + // never used by wxHTML + wxString s; + size_t cnt = m_ParamNames.GetCount(); + for (size_t i = 0; i < cnt; i++) { - if (*p2 == 0) // found - { - wxString fnd = wxEmptyString; - st2++; // '=' character - comma = FALSE; - comma_char = wxT('\0'); - if (!with_commas && (*(st2) == wxT('"'))) - { - st2++; - comma = TRUE; - comma_char = wxT('"'); - } - else if (!with_commas && (*(st2) == wxT('\''))) - { - st2++; - comma = TRUE; - comma_char = wxT('\''); - } - - while (*st2 != 0) - { - if (comma && *st2 == comma_char) comma = FALSE; - else if ((*st2 == wxT(' ')) && (!comma)) break; - fnd += (*(st2++)); - } - if (!with_commas && (*(st2-1) == comma_char)) fnd.RemoveLast(); - return fnd; - } - if (*st2 == 0) return wxEmptyString; - if (*p2 != *st2) p2 = &invalid; - if (*p2 == *st2) p2++; - if (*st2 == wxT(' ')) p2 = p; - else if (*st2 == wxT('=')) - { - p2 = p; - while (*st2 != wxT(' ')) - { - if (*st2 == wxT('"')) - { - st2++; - while (*st2 != wxT('"')) st2++; - } - else if (*st2 == wxT('\'')) - { - st2++; - while (*st2 != wxT('\'')) st2++; - } - st2++; - } - } + s << m_ParamNames[i]; + s << wxT('='); + if (m_ParamValues[i].Find(wxT('"')) != wxNOT_FOUND) + s << wxT('\'') << m_ParamValues[i] << wxT('\''); + else + s << wxT('"') << m_ParamValues[i] << wxT('"'); } + return s; } +wxHtmlTag *wxHtmlTag::GetFirstSibling() const +{ + if (m_Parent) + return m_Parent->m_FirstChild; + else + { + wxHtmlTag *cur = (wxHtmlTag*)this; + while (cur->m_Prev) + cur = cur->m_Prev; + return cur; + } +} +wxHtmlTag *wxHtmlTag::GetLastSibling() const +{ + if (m_Parent) + return m_Parent->m_LastChild; + else + { + wxHtmlTag *cur = (wxHtmlTag*)this; + while (cur->m_Next) + cur = cur->m_Next; + return cur; + } +} -int wxHtmlTag::ScanParam(const wxString& par, wxChar *format, void *param) const +wxHtmlTag *wxHtmlTag::GetNextTag() const { - wxString parval = GetParam(par); - return wxSscanf((const wxChar*)parval, format, param); + if (m_FirstChild) return m_FirstChild; + if (m_Next) return m_Next; + wxHtmlTag *cur = m_Parent; + if (!cur) return NULL; + while (cur->m_Parent && !cur->m_Next) + cur = cur->m_Parent; + return cur->m_Next; } #endif