#include "basehtmlparser.h"

#ifndef WIN32
#define strnicmp strncasecmp
#endif

CBaseHtmlParser::CBaseHtmlParser()
{
}

CBaseHtmlParser::~CBaseHtmlParser()
{
}

void CBaseHtmlParser::Relativity2AbsoluteURL(string& URL)
{
 int pos=-1;
 pos=URL.rfind("#");
 if(pos!=string::npos)
 {
  URL=URL.substr(0,pos);
 }

 do
 {
  pos=URL.find("&");
  if(pos!=string::npos)
   URL=URL.substr(0,pos)+"&"+URL.substr(pos+5);
 }while(pos>=0);

 do
 {
  pos=URL.find(">");
  if(pos!=string::npos)
   URL=URL.substr(0,pos)+">"+URL.substr(pos+4);
 }while(pos>=0);

 do
 {
  pos=URL.find("<");
  if(pos!=string::npos)
   URL=URL.substr(0,pos)+"<"+URL.substr(pos+4);
 }while(pos>=0);

 if((URL.length()>=1)&&(URL[0]=='/'))
 {
  URL=m_szBaseDomain+URL;
 }else if((URL.length()>=7)&&(strnicmp((char*)URL.c_str(),"http://",7)==0))
 {
  return;
 }else
 {
  URL=m_szBaseURL+URL;
 }
}

bool CBaseHtmlParser::Parse(const string& URL,const string& Content)
{
 SZ_STRING strUrl,strContent;
 strUrl.pbData=URL.c_str();
 strUrl.cbData=URL.length();
 strContent.pbData=Content.c_str();
 strContent.cbData=Content.length();
 
 return Parse(strUrl,strContent);
}

bool CBaseHtmlParser::Parse(const SZ_STRING &strUrl,const SZ_STRING &strContent)
{
 sprintf(m_szBaseURL,"%.*s",strUrl.cbData,strUrl.pbData);
 char* pend=strrchr(m_szBaseURL,'/');
 if(pend!=NULL)
 {
  pend++;
  *pend='\0';
 }
 sprintf(m_szBaseDomain,"%.*s",strUrl.cbData,strUrl.pbData);
 pend=strchr(m_szBaseDomain+strlen("http://")+1,'/');
 if(pend!=NULL)
 {
  *pend='\0';
 }

 size_t i;
 size_t nContent=strContent.cbData;
 const char* pContent=(const char*)(strContent.pbData);
 for(i=0;i<nContent;)
 {
  if(pContent[i]=='<')//TagStart
  {
   if(nContent>4)
   if((i<nContent-4)&&(pContent[i+1]=='!')&&(pContent[i+2]=='-')&&(pContent[i+3]=='-'))//comment start
   {
    i+=4;
    size_t nCommentStart=i;

    if(nContent>3)
    while(i<nContent-3)
    {
     if((pContent[i]=='-')&&(pContent[i+1]=='-')&&(pContent[i+2]=='>'))//comment end
     {
      SZ_STRING strComment;
      strComment.pbData=pContent+nCommentStart;
      strComment.cbData=i-nCommentStart;
      OnComment(strComment);
      i+=3;
      break;
     }
     i++;
    }
    continue;
   }
   //tag here
   size_t nTagNameStart=i+1;
   while((nTagNameStart<nContent)&&(pContent[nTagNameStart]==' '))
    nTagNameStart++;
   size_t nTagNameEnd=nTagNameStart+1;
   while((nTagNameEnd<nContent)&&(pContent[nTagNameEnd]!=' ')&&(pContent[nTagNameEnd]!='>'))
    nTagNameEnd++;

   SZ_STRING strTagName;
   strTagName.pbData = pContent+nTagNameStart;
   strTagName.cbData = nTagNameEnd-nTagNameStart;

   size_t nTagEnd=nTagNameEnd;
   while((nTagEnd<nContent)&&(pContent[nTagEnd]!='>'))
    nTagEnd++;
   nTagEnd++;

   const char* pTag=pContent+i;
   size_t nTag=nTagEnd-i;

   i=nTagEnd;

   vector< pair<SZ_STRING,SZ_STRING> > Attribs;

   if((strTagName.cbData ==6)&&(strnicmp((char*)strTagName.pbData,"script",6)==0))//<script></script>
   {
    size_t nScriptStart=i;
    if(nContent>8)
    while(i<nContent-8)
    {
     if(strnicmp((char*)pContent+i,"</script",8)==0)
     {
      OnStartTag(strTagName,Attribs);
      SZ_STRING strComment;
      strComment.pbData = pContent+nScriptStart;
      strComment.cbData = i-nScriptStart;
      OnComment(strComment);
      OnEndTag(strTagName);
      i+=8;
      break;
     }
     i++;
    }
    while((i<nContent)&&(pContent[i]!='>'))
     i++;
    i++;
    continue;
   }

   if(strTagName.pbData[0]=='/')
   {
    strTagName.pbData +=1;
    strTagName.cbData -=1;
    OnEndTag(strTagName);
    continue;
   }

   size_t m=strTagName.cbData+1;
   do
   {
    while((m<nTag)&&(((char)pTag[m]==' ')||((char)pTag[m]=='\r')||((char)pTag[m]=='\n')||((char)pTag[m]=='\t')))
     m++;

    size_t nAttribStart=m;

    while((m<nTag)&&((char)pTag[m]!='='))
     m++;

    size_t nAttribStop=m-1;
    while((nAttribStop>0)&&((pTag[nAttribStop]==' ')||(pTag[nAttribStop]=='\r')||(pTag[nAttribStop]=='\n')||(pTag[nAttribStop]=='\t')))
      nAttribStop--;
    if(nAttribStop>0)
     nAttribStop++;

    SZ_STRING strAttribName;
    strAttribName.pbData =pTag+nAttribStart;
    strAttribName.cbData =nAttribStop-nAttribStart;

    m++;
    bool bStartWithDQ=false;
    bool bStartWithQ=false;
    bool bStartWithSpace=false;
    if((m<nTag)&&((char)pTag[m]=='\"'))
    {
     bStartWithDQ=true;
     m++;
    }else if((m<nTag)&&((char)pTag[m]=='\''))
    {
     bStartWithQ=true;
     m++;
    }else
     bStartWithSpace=true;

    if(m>=nTag)
     continue;

    size_t nValueStart=m;

    while((m<nTag)
     &&((char)pTag[m]!='>')
     &&((char)pTag[m]!='\r')
     &&((char)pTag[m]!='\n')
     &&(!bStartWithDQ||((char)pTag[m]!='\"'))
     &&(!bStartWithQ||((char)pTag[m]!='\''))
     &&(!bStartWithSpace||((char)pTag[m]!=' ')))
     m++;

    SZ_STRING strAttribValue;
    strAttribValue.pbData =pTag+nValueStart;
    strAttribValue.cbData =m-nValueStart;

    if(strAttribName.cbData !=0)
     Attribs.push_back(pair<SZ_STRING,SZ_STRING>(strAttribName,strAttribValue));

    while((m<nTag)&&((char)pTag[m]!=' '))
     m++;

   }while(m<nTag);

   OnStartTag(strTagName,Attribs);
  }else
  {
   while((i<nContent)&&((pContent[i]==' ')||(pContent[i]=='\r')||(pContent[i]=='\n')||(pContent[i]=='\t')))
    i++;
   size_t nTextBegin=i;
   while((i<nContent)&&(pContent[i]!='<'))
    i++;
   size_t nTextEnd=i;

   while((nTextEnd>=nTextBegin)&&((pContent[nTextEnd-1]==' ')||(pContent[nTextEnd-1]=='\r')||(pContent[nTextEnd-1]=='\n')||(pContent[nTextEnd-1]=='\t')))
    nTextEnd--;
   if(nTextEnd<=nTextBegin)
    continue;
   SZ_STRING strData;
   strData.pbData =pContent+nTextBegin;
   strData.cbData =nTextEnd-nTextBegin;
   OnData(strData);
  }
 }
 return true;
}