Populate the hzDocHtml object with HTML source code in the supplied chain. Two scenarios are permitted - Full or Partial as follows:- 1) Full: If the HTML source has the <html> as its first tag it will be considered as a full page and tested as such. It will be expected to have the standard sub-tags of <head> and <body> and thier corresponding anti-tags. If either of these are missing or in error (malformed or containing unxpected or malformed tags) the HTML source code is deemed to be syntactically in error and the load fails. 2) Partial: If the opening tag of the HTML source code is not the <html> tag it is viable only if it would be viable as a HTML fragment that could be seemlessly inserted into the <body> part of a whole HTML page. This is to say that all it's tags must be legal sub-tags of <body> and not of <head> and nor must the <body> or <head> tag or anti-tag be present. In either case, tags are loaded into a tree of nodes (tags). The nodes/tags may be searched for and examined. Note: Unlike XML where tags are named so that content in the tree can be searched directly, the nodes in HTML are not named named and so cannot be definitely referenced (they only have type). Some other process must apply application specific criteria to read meaning into the data.
| Return Type | Function name | Arguments |
|---|---|---|
| hzEcode | hzDocHtml::Load | (hzChain&,) |
Declared in file: hzDocument.h
Defined in file : hzDocHtml.cpp
Function Logic:
Function body:
hzEcode hzDocHtml::Load (hzChain& Z)
{
// Populate the hzDocHtml object with HTML source code in the supplied chain.
//
// Two scenarios are permitted - Full or Partial as follows:-
// 1) Full: If the HTML source has the <html> as its first tag it will be considered as a full page and tested as such.
// It will be expected to have the standard sub-tags of <head> and <body> and thier corresponding anti-tags.
// If either of these are missing or in error (malformed or containing unxpected or malformed tags) the HTML
// source code is deemed to be syntactically in error and the load fails.
//
// 2) Partial: If the opening tag of the HTML source code is not the <html> tag it is viable only if it would be viable as
// a HTML fragment that could be seemlessly inserted into the <body> part of a whole HTML page. This is to say
// that all it's tags must be legal sub-tags of <body> and not of <head> and nor must the <body> or <head> tag
// or anti-tag be present.
//
// In either case, tags are loaded into a tree of nodes (tags). The nodes/tags may be searched for and examined.
//
// Arguments: 1) Z The chain containing the HTML document
//
// Returns: E_FORMAT If the HTML was rejected by the the HTML pre-processor _htmlPreproc() OR if any tags could not be processed by _proctag()
// E_OK If the HTML was loaded successfully
//
// Note: Unlike XML where tags are named so that content in the tree can be searched directly, the nodes in HTML are not named
// named and so cannot be definitely referenced (they only have type). Some other process must apply application specific criteria
// to read meaning into the data.
_hzfunc("hzDocHtml::Load") ;
hzChain nc ; // Node content
hzChain T ; // For token building
hzChain W ; // For token building
chIter zi ; // Chain iterator
chIter tw_start ; // Start of tagword marker
chIter tmp ; // Start of tagword marker
chIter limit ; // End of tag marker - Protection against malformed tags (NLA style)
hzHtmElem* pCN = 0; // Current HTML node
hzHtmElem* pNN ; // New HTML node
hzHtmElem* pX ; // HTML node for diagnostics
hzHtmElem* pCurForm = 0; // HTML node for diagnostics
hzAttrset ai ; // Attribute iterator
hzHtmForm* pForm = 0; // Form found in page
hzPair P ; // Name value pair (for forms and fields)
hzString strval ; // To test if current tag is being closed
hzString tagword ; // From MakeTag - just the tagname.
hzString wholetag ; // From MakeTag - the entire opening sequence if applicable
hzString anam ; // Attribute name
hzString aval ; // Attribute value
hzHtagform tf ; // Tag form
uint32_t nX ; // For nesting levels/general iteration
uint32_t nColon ; // Does the tagname contain a colon (3rd party tag)
uint32_t nLine ; // Line number for errors
uint32_t quote ; // Are we in a quoted string
bool bAnti ; // Tag is an anti-tag
int32_t cDelim ; // Delimiting char (single/double quote)
hzEcode rc = E_OK ; // return code
Clear() ;
// m_Error.Clear() ;
// Pre-process the HTML
rc = _htmPreproc(Z) ;
if (rc != E_OK)
return rc ;
m_Content = Z ;
// Make sure the HTML tags are loading into the lookup table
if (!s_htagNam.Count())
InitHtml() ;
// Init the iterator
zi = Z ;
zi.Skipwhite() ;
// Bypass the doctype if present
if (zi.Equiv("<!DOCTYPE"))
{
quote = 0;
for (zi += 9; !zi.eof() ; zi++)
{
if (quote)
{
if (*zi == CHAR_DQUOTE)
quote = 0;
continue ;
}
if (*zi == CHAR_MORE)
{ zi++ ; break ; }
if (*zi == CHAR_DQUOTE)
quote = 1;
}
zi.Skipwhite() ;
}
// Look for the opening <html>
for (; !zi.eof() ; zi++)
{
if (zi.Equiv("<html"))
{
m_pRoot = _proctag(0,zi, HTAG_HTML) ;
if (!m_pRoot)
{ threadLog("Could not establist root node (the <html> tag)\n") ; return E_FORMAT ; }
break ;
}
}
if (!m_pRoot)
{
threadLog("No valid contents found before expected <html> tag - assuming a partial page\n") ;
zi = Z ;
zi.Skipwhite() ;
pCN = new hzHtmElem() ;
pCN->Init(this, 0,tagword, HTAG_NULL, m_vecTags.Count(), zi.Line()) ;
m_vecTags.Add(pCN) ;
}
else
{
// A <html> tag has been found so this is a full page. Look for <head> next
for (; !zi.eof() ;)
{
if (zi.Equiv("<head"))
{
m_pHead = _proctag(m_pRoot, zi, HTAG_HEAD) ;
if (!m_pHead)
{ threadLog("Could not process <head> tag\n") ; return E_FORMAT ; }
break ;
}
zi++ ;
}
if (!m_pHead)
{ threadLog("Expected a <head> tag\n") ; return E_FORMAT ; }
pCN = m_pHead ;
// Now get the subtags of <head>
for (; rc == E_OK && pCN && !zi.eof() ;)
{
// Handle tag content
if (*zi != CHAR_LESS)
{
// Ignore certain constructs
if (zi == "//")
{
for (zi += 2; !zi.eof() && *zi != CHAR_NL ; zi++) ;
continue ;
}
// If not part of a construct, just agregate the char to the current tag's content, striping leading whitespace
if (*zi <&eq; CHAR_SPACE && pCN->m_tmpContent.Size() == 0)
{ zi++ ; continue ; }
pCN->m_tmpContent.AddByte(*zi) ;
zi++ ;
continue ;
}
// Ignore deleted text within comment (<!-- and -->) tags. Note these cannot be nested
nLine = zi.Line() ;
if (zi == "<!--[if")
{
for (zi += 7; !zi.eof() ; zi++)
{
if (zi == "<![endif]>") { zi += 10;break ; }
if (zi == "<![endif]-->") { zi += 12;break ; }
}
continue ;
}
if (zi == "<![if")
{
for (zi += 5; !zi.eof() ; zi++)
{
if (zi == "<![endif]>") { zi += 10;break ; }
if (zi == "<![endif]-->") { zi += 12;break ; }
}
continue ;
}
if (zi == "<!--")
{
for (zi += 4; !zi.eof() ; zi++)
{
if (zi == "-->")
{ zi += 3; break ; }
}
continue ;
}
// Handle <![CDATA[...]]> block by converting the innards to straight data (apparently CDATA now legal in HTML)
if (zi == "<![CDATA[")
{
for (zi += 9; !zi.eof() ; zi++)
{
if (zi == "]]>")
{ zi += 3; break ; }
pCN->m_tmpContent.AddByte(*zi) ;
}
continue ;
}
// Eliminate <noscript> tags from header (we don't use them)
if (zi == "<noscript")
{
for (zi += 9; !zi.eof() ; zi++)
{
if (zi == "</noscript>")
{ zi += 11;break ; }
}
if (zi.eof())
{ threadLog("Unclosed <noscript> block\n") ; rc = E_FORMAT ; break ; }
continue ;
}
// At this point we have the '<' start of tag char. Establish whole and tagword of possible HTML tag
wholetag.Clear() ;
tagword.Clear() ;
limit = zi ;
limit++ ;
W.AddByte(CHAR_LESS) ;
bAnti = false ;
if (*limit == CHAR_FWSLASH)
{ W.AddByte(CHAR_FWSLASH) ; bAnti = true ; limit++ ; }
nColon = 0;
for (tw_start = limit ; !limit.eof() ; limit++)
{
if (*limit == CHAR_COLON || IsAlphanum(*limit))
{
if (*limit == CHAR_COLON)
nColon++ ;
T.AddByte(*limit) ;
W.AddByte(*limit) ;
continue ;
}
break ;
}
tagword = T ;
T.Clear() ;
for (; !limit.eof() ;)
{
W.AddByte(*limit) ;
if (*limit == CHAR_DQUOTE || *limit == CHAR_SQUOTE)
{
cDelim = *limit ;
for (limit++ ; !limit.eof() ; limit++)
{
if (*limit == CHAR_BKSLASH)
{
limit++ ;
if (*limit == cDelim)
continue ;
}
if (*limit == cDelim)
break ;
}
}
if (*limit == CHAR_MORE)
break ;
limit++ ;
}
wholetag = W ;
W.Clear() ;
if (*limit != CHAR_MORE)
{
threadLog("Malformed tag (%s)\n", *wholetag) ;
zi = limit ;
continue ;
}
limit++ ;
// tagword.ToLower() ;
if (nColon)
{
if (!s_htagNam.Exists(tagword))
{
tf.klas = HTCLASS_3RD ;
tf.rule = HTRULE_OPTION ;
tf.name = tagword ;
s_htagTyp.Insert(tf.type, tf) ;
s_htagNam.Insert(tf.name, tf) ;
threadLog("Inserted 3rd party HTML tag %s\n", *tagword) ;
}
}
if (!s_htagNam.Exists(tagword))
{
if (bAnti)
threadLog("Line %d case 1 Unknown lookup anti-tag </%s> (%s)\n", zi.Line(), *tagword, *wholetag) ;
else
threadLog("Line %d Case 1 Unknown lookup tag <%s> (%d bytes)\n", zi.Line(), *tagword, wholetag.Length()) ;
pCN->m_tmpContent << wholetag ;
zi = limit ;
continue ;
}
tf = s_htagNam[tagword] ;
if (tf.type == HTAG_NULL)
{
if (bAnti)
threadLog("Line %d case 2 Unknown lookup anti-tag </%s> (%s)\n", zi.Line(), *tagword, *wholetag) ;
else
threadLog("Line %d Case 2 Unknown lookup tag <%s> (%d bytes)\n", zi.Line(), *tagword, wholetag.Length()) ;
pCN->m_tmpContent << wholetag ;
zi = limit ;
continue ;
}
// Obtain tag name
if (bAnti == false)
{
if (zi.Equiv("<title>"))
{
pCN = _proctag(m_pHead, zi, HTAG_TITLE) ;
if (!pCN)
{ rc = E_FORMAT ; threadLog("Line %d Could not process <meta> tags\n", zi.Line()) ; }
}
else if (zi.Equiv("<meta"))
{
pCN = _proctag(m_pHead, zi, HTAG_META) ;
if (!pCN)
{ rc = E_FORMAT ; threadLog("Line %d Could not process <meta> tags\n", zi.Line()) ; }
}
else if (zi.Equiv("<style"))
{
pCN = _proctag(m_pHead, zi, HTAG_STYLE) ;
if (!pCN)
{ rc = E_FORMAT ; threadLog("Line %d Could not process <style> tags\n", zi.Line()) ; }
}
else if (zi.Equiv("<script"))
{
pCN = _proctag(m_pHead, zi, HTAG_SCRIPT) ;
if (!pCN)
{ rc = E_FORMAT ; threadLog("Line %d Could not process <script> tags\n", zi.Line()) ; }
}
else if (zi.Equiv("<link"))
{
pCN = _proctag(m_pHead, zi, HTAG_LINK) ;
if (!pCN)
{ rc = E_FORMAT ; threadLog("Line %d Could not process <link> tags\n", zi.Line()) ; }
}
else if (zi.Equiv("<base"))
{
pCN = _proctag(m_pHead, zi, HTAG_BASE) ;
if (!pCN)
{ rc = E_FORMAT ; threadLog("Line %d Could not process <link> tags\n", zi.Line()) ; }
// Set m_Base
if (pCN->m_tmpContent.Size())
m_Base = pCN->m_tmpContent ;
else
{
// set the m_Base to the first param
ai = pCN ;
if (ai.Value())
m_Base = ai.Value() ;
// pAttr = pCN->GetFirstAttr() ;
// if (pAttr)
// m_Base = pAttr->value ;
}
}
else
{ rc = E_FORMAT ; threadLog("Line %d Could not process <%s> tag within <head>\n", zi.Line(), *tagword) ; }
continue ;
}
// Handle antitag
if (bAnti)
{
if (zi.Equiv("</head>"))
{ zi += 7; break ; }
// Inactive (text rendering only) anti-tags
if (tf.klas == HTCLASS_TXT)
{ pCN->m_tmpContent << wholetag ; zi = limit ; continue ; }
// { zi = limit ; continue ; }
zi = limit ;
if (pCN->Type() == tf.type || tf.rule == HTRULE_SINGLE)
pCN = pCN->Parent() ;
else
{
threadLog("case 1 Tag mis-match. Current highest tag is <%s id=%d, level=%d> but on line %d we have an anti-tag for %s\n",
*Tagtype2Txt(pCN->Type()), pCN->GetUid(), pCN->Level(), zi.Line(), *Tagtype2Txt(tf.type)) ;
if (tf.rule == HTRULE_SINGLE)
{
// pCN = pX ;
pCN = pCN->Parent() ;
threadLog("Case 2 Corrected by allowing last tag as anti-tag\n") ;
}
if (pCN->Type() == HTAG_TBL_CEL && tf.type == HTAG_TR)
{
for (pX = pCN ; pX ; pX = pX->Parent())
{
if (pX->Type() == tf.type)
{
pCN = pX ;
threadLog("Corrected by decending to level %d\n", pCN->Level()) ;
break ;
}
}
}
}
continue ;
}
// If none of the above just advance
zi++ ;
}
// Advance to the <body> tag
for (; !zi.eof() ;)
{
if (zi.Equiv("<body"))
{
m_pBody = _proctag(m_pRoot, zi, HTAG_BODY) ;
if (!m_pBody)
{ threadLog("Expected an actual body\n") ; return E_FORMAT ; }
break ;
}
zi++ ;
}
if (!m_pBody)
{ threadLog("Expected a <body> tag\n") ; return E_FORMAT ; }
pCN = m_pBody ;
}
//
// Process document body. Here everything is either a tag, an anti-tag or it is tag-content. Both tags and antitags begin with a '<' so the
// raw HTML is iterated and whenever the < is found, it is tested for a known tag/antitag. In the general case of "<tag>content</tag>", the
// process is to call _procTag() to parse the tag, garner the attributes and to create a new element (which the current element is then set
// to). Bytes after the tag are agregated to the current element's content until the antitag occurs (at which point the current element is
// then set back to the parent tag).
//
// The exceptions to the general case:-
//
// 1) Paragraph tags can be left open (antitag omited). These tags are closed by the parent antitag or by another paragraph tag.
//
// 2) Print control tags which are completely ignored. These can never become the current tag so any content they have is aggregated to
// their parent tag.
//
// 3) Links which do become current, but will have thier content aggregated to the parent tag.
//
for (; pCN && !zi.eof() ;)
{
// Handle tag content
if (*zi != CHAR_LESS)
{
if (pCN->Type() != HTAG_ANCHOR)
{
if (*zi <&eq; CHAR_SPACE && pCN->m_tmpContent.Size() == 0)
{ zi++ ; continue ; }
pCN->m_tmpContent.AddByte(*zi) ;
}
else
{
if (pCN->Parent())
pCN->Parent()->m_tmpContent.AddByte(*zi) ;
}
zi++ ;
continue ;
}
// Ignore deleted text within <strike></strike> tags
nLine = zi.Line() ;
if (zi == "<strike>")
{
for (zi += 8; !zi.eof() ; zi++)
{
if (zi == "</strike>")
{ zi += 9; break ; }
}
if (zi.eof())
{ threadLog("Unclosed comment block\n") ; rc = E_FORMAT ; break ; }
continue ;
}
if (zi == "<fb:like>")
{
for (zi += 9; !zi.eof() ; zi++)
{
if (zi == "</fb:like>")
{ zi += 10;break ; }
}
if (zi.eof())
{ threadLog("Facebook special\n") ; rc = E_FORMAT ; break ; }
continue ;
}
if (zi == "<g:plusone>")
{
for (zi += 11;!zi.eof() ; zi++)
{
if (zi == "</g:plusone>")
{ zi += 12;break ; }
}
if (zi.eof())
{ threadLog("Google special\n") ; rc = E_FORMAT ; break ; }
continue ;
}
// Ignore deleted text within comment (<!-- and -->) tags
if (zi == "<!--[if")
{
for (zi += 7; !zi.eof() ; zi++)
{
if (zi == "<![endif]>") { zi += 10;break ; }
if (zi == "<![endif]-->") { zi += 12;break ; }
}
if (zi.eof())
{ threadLog("Unterminated <!--[if cond]..> tag starting line %d\n", nLine) ; rc = E_FORMAT ; break ; }
continue ;
}
if (zi == "<![if")
{
for (zi += 5; !zi.eof() ; zi++)
{
if (zi == "<![endif]>") { zi += 10;break ; }
if (zi == "<![endif]-->") { zi += 12;break ; }
}
if (zi.eof())
{ threadLog("Unterminated <![if cond]..> tag starting line %d\n", nLine) ; rc = E_FORMAT ; break ; }
continue ;
}
if (zi == "<!--")
{
for (zi += 4; !zi.eof() ; zi++)
{
if (zi == "-->")
{ zi += 3; break ; }
}
if (zi.eof())
{ threadLog("Unterminated <!--> tag starting line %d\n", nLine) ; rc = E_FORMAT ; break ; }
continue ;
}
/*
** ** At this point we have the '<' start of tag char. Establish whole and tagword of possible HTML tag
** */
wholetag.Clear() ;
tagword.Clear() ;
limit = zi ;
limit++ ;
W.AddByte(CHAR_LESS) ;
bAnti = false ;
if (*limit == CHAR_FWSLASH)
{ W.AddByte(CHAR_FWSLASH) ; bAnti = true ; limit++ ; }
nColon = 0;
for (tw_start = limit ; !limit.eof() ; limit++)
{
if (*limit == CHAR_COLON || IsAlphanum(*limit))
{
if (*limit == CHAR_COLON)
nColon++ ;
T.AddByte(*limit) ;
W.AddByte(*limit) ;
continue ;
}
break ;
}
tagword = T ;
T.Clear() ;
for (; !limit.eof() ;)
{
W.AddByte(*limit) ;
if (*limit == CHAR_DQUOTE || *limit == CHAR_SQUOTE)
{
cDelim = *limit ;
for (limit++ ; !limit.eof() ; limit++)
{
if (*limit == CHAR_BKSLASH)
{
limit++ ;
if (*limit == cDelim)
continue ;
}
if (*limit == cDelim)
break ;
}
}
if (*limit == CHAR_MORE)
break ;
limit++ ;
}
wholetag = W ;
W.Clear() ;
if (*limit != CHAR_MORE)
{
threadLog("Malformed tag (%s)\n", *wholetag) ;
zi = limit ;
continue ;
}
tagword.ToLower() ;
if (nColon)
{
if (!s_htagNam.Exists(tagword))
{
tf.klas=HTCLASS_3RD ;
tf.rule=HTRULE_OPTION ;
tf.name = tagword ;
s_htagTyp.Insert(tf.type, tf) ;
s_htagNam.Insert(tf.name, tf) ;
threadLog("Inserted 3rd party HTML tag %s\n", *tagword) ;
}
}
// if (bAnti)
// threadLog("Case 2 line %d Doing antitag %s\n", zi.Line(), *tagword) ;
// else
// threadLog("Case 2 line %d Doing tag %s\n", zi.Line(), *tagword) ;
tf = s_htagNam[tagword] ;
if (tf.type == HTAG_NULL)
{
// Unrecognized tags are just made part of the content of the currently applicable tag
if (bAnti)
threadLog("Line %d Unknown lookup anti-tag </%s> (%s)\n", zi.Line(), *tagword, *wholetag) ;
else
threadLog("Line %d Case 3 Unknown lookup tag <%s> (%d bytes)\n", zi.Line(), *tagword, wholetag.Length()) ;
pCN->m_tmpContent << wholetag ;
zi = limit ;
continue ;
}
if (bAnti == false)
{
// Ignore graphic tags
if (tf.klas == HTCLASS_IMG)
{ zi = limit ; continue ; }
// Ignore self-closed 'system' tags
if (tf.klas == HTCLASS_SYS)
{
if (tf.type == HTAG_EMBED)
pCN->m_tmpContent << "<embed/>" ;
if (tf.type == HTAG_NOEMBED)
pCN->m_tmpContent << "<noembed/>" ;
for (; !zi.eof() ; zi++)
{
if (*zi == CHAR_MORE)
{ zi++ ; break ; }
}
threadLog("Line %d Bypassed system tag <%s> (%s)\n", zi.Line(), *tagword, *wholetag) ;
zi = limit ;
continue ;
}
// Handle HTCLASS_TXT 'in-content' tags. We just copy these through, complete with tag, antitag and content, to the content of the
// current tag. However these tags should still be placed in the m_mapTags and m_vecTags member.
if (tf.klas == HTCLASS_TXT) // || tf.type == HTAG_ANCHOR)
{
pCN->m_tmpContent << wholetag ; zi = limit ;
continue ;
}
// If we are suppressing anchors, we only want the content of a <a href=...>...</a> sequence.
// if (m_bOpflags & HDOC_SUPPRESS_LINKS && tf.klas == HTCLASS_LNK && tf.type == HTAG_ANCHOR)
// if (bFlags & HDOC_ONLOAD_LINKS && tf.klas == HTCLASS_LNK && tf.type == HTAG_ANCHOR)
// { zi = limit ; continue ; }
// Eliminate scripts (may revisit)
if (zi.Equiv("<script"))
{
// plog->Out("%s. ignoring a script tag ...\n", __FUNCTION__) ;
for (tmp = zi ; !tmp.eof() ; tmp++)
{
if (tmp.Equiv("</script>"))
{ tmp += 9; zi = tmp ; break ; }
}
if (zi.eof())
{ threadLog("Unclosed script tag\n") ; rc = E_FORMAT ; break ; }
continue ;
}
/*
** ** Process 'data structure' tags into nodes. These are tables (with there rows and columns) but also menus
** ** and ordered and unordered lists.
** */
pNN = 0;
pNN = _proctag(pCN, zi, tf.type) ;
if (!pNN)
{
threadLog("No node allocated for tag <%s>\n", *Tagtype2Txt(tf.type)) ;
return E_FORMAT ;
}
pCN = pNN ;
zi = limit ;
/*
** ** Handle the <input> tag. As this is it's own anti-tag it has no content, only parameters. We need to include the tag
** ** in the tree as it is active, but we need to effect the anti-tag aspect as well (so the level is not raised)
** */
if (tf.type == HTAG_INPUT)
pCN = pCN->Parent() ;
continue ;
}
// Handle anti-tags
if (bAnti)
{
// Inactive (text rendering only) anti-tags
if (tf.klas == HTCLASS_TXT) // || tf.type == HTAG_ANCHOR)
{ pCN->m_tmpContent << wholetag ; zi = limit ; continue ; }
// Ignore self-closed 'system' tags
if (tf.klas == HTCLASS_SYS)
{
if (tf.type == HTAG_EMBED)
pCN->m_tmpContent << "</embed>" ;
if (tf.type == HTAG_NOEMBED)
pCN->m_tmpContent << "</noembed>" ;
for (; !zi.eof() ; zi++)
{
if (*zi == CHAR_MORE)
{ zi++ ; break ; }
}
threadLog("Line %d Bypassed system anti-tag <%s> (%s)\n", zi.Line(), *tagword, *wholetag) ;
zi = limit ;
continue ;
}
zi = limit ;
if (pCN->Type() == tf.type || tf.rule == HTRULE_SINGLE)
pCN = pCN->Parent() ;
else
{
threadLog("case 2 Tag mis-match. Current highest tag is <%s id=%d, level=%d> but on line %d we have an anti-tag for %s\n",
*Tagtype2Txt(pCN->Type()), pCN->GetUid(), pCN->Level(), zi.Line(), *Tagtype2Txt(tf.type)) ;
if (tf.rule == HTRULE_SINGLE)
{
// pCN = pX ;
pCN = pCN->Parent() ;
threadLog("Case 1 Corrected by allowing last tag as anti-tag\n") ;
}
if (pCN->Type() == HTAG_TBL_CEL && tf.type == HTAG_TR)
{
for (pX = pCN ; pX ; pX = pX->Parent())
{
if (pX->Type() == tf.type)
{
pCN = pX ;
threadLog("Corrected by decending to level %d\n", pCN->Level()) ;
break ;
}
}
}
}
continue ;
}
threadLog("HANDLING ABD %s (%s)\n", *tagword, *wholetag) ;
}
if (pCN)
threadLog("End of file encountered whilst inside tag definition\n") ;
// Move thru the tags in thier order of appearence and reduce where appropriate, the tag content held in chains to strings. Place forms in
// the list of forms and place form field tags with thier host forms.
for (nX = 0; nX < m_vecTags.Count() ; nX++)
{
pX = m_vecTags[nX] ;
if (pX->Type() == HTAG_FORM)
{
// Add the form to to m_Forms and set this to the current form
pCurForm = pX ;
pForm = new hzHtmForm() ;
m_Forms.Add(pForm) ;
continue ;
}
if (pCurForm)
{
if (pX->Type() == HTAG_INPUT)
{
// Add this field to the current form (report error if not in a current form)
if (pX->Line() < pCurForm->Anti())
{
P.name = pX->Name() ;
// for (pAttr = pX->GetFirstAttr() ; pAttr ; pAttr = pAttr->next)
// {
// if (pAttr->name == "value")
// { P.value = pAttr->value ; break ; }
// }
for (ai = pX ; ai.Valid() ; ai.Advance())
{
anam = ai.Name() ; aval = ai.Value() ;
if (anam == "value")
{ P.value = aval ; break ; }
}
pForm->fields.Add(P) ;
}
continue ;
}
if (pX->Line() > pCurForm->Anti())
pCurForm = 0;
}
}
threadLog("END OF LOAD page has %d links\n", m_vecLinks.Count()) ;
return rc ;
}