C++ XmlDoc::loadFromOldTitleRec方法代码示例

// . returns false if blocked, true otherwise
// . sets g_errno on error
// . make a web page displaying the titleRec of "docId" given via cgi
// . call g_httpServer.sendDynamicPage() to send it
bool sendPageTitledb ( TcpSocket *s , HttpRequest *r ) {
	// get the docId from the cgi vars
	long long docId = r->getLongLong ("d", 0LL );
	// set up a msg22 to get the next titleRec
	State4 *st ;
	try { st = new (State4); }
	catch ( ... ) {
		g_errno = ENOMEM;
		log("PageTitledb: new(%i): %s", 
		    (int)sizeof(State4),mstrerror(g_errno));
		return g_httpServer.sendErrorReply(s,500,mstrerror(g_errno));}
	mnew ( st , sizeof(State4) , "PageTitledb");
	// save the socket
	st->m_socket = s;
	// copy it
	st->m_r.copy ( r );
	// remember if http request is internal/local or not
	st->m_isRootAdmin = g_conf.isCollAdmin ( s , r );
	st->m_isLocal = r->isLocal();
	st->m_docId   = docId;
	// password, too
	st->m_pwd = r->getString ( "pwd" );
	// get the collection
	long  collLen = 0;
	char *coll    = st->m_r.getString("c",&collLen);
	if ( ! coll || ! coll[0] ) {
		//coll    = g_conf.m_defaultColl;
		coll = g_conf.getDefaultColl( r->getHost(), r->getHostLen() );
		collLen = gbstrlen(coll);
	}
	st->m_coll    = coll;
	st->m_collLen = collLen;

	// just print page if no docid provided
	if ( ! docId ) return gotTitleRec ( st );

	// get the handy XmlDoc
	XmlDoc *xd = &st->m_xd;
	// use 0 for niceness
	xd->set3 ( docId , coll , 0 );
	// callback
	xd->setCallback ( st , gotTitleRec );
	// . and tell it to load from old title rec
	// . this sets all the member vars from it and also sets
	//   m_titleRecBuf to contain the actual compressed title rec
	if ( ! xd->loadFromOldTitleRec ( ) ) return false;
	// we got it without blocking. cached?
	return gotTitleRec ( st );
}

// returns false if blocked, true otherwise
bool processLoop ( void *state ) {
	// get it
	State60 *st = (State60 *)state;
	// get the tcp socket from the state
	TcpSocket *s = st->m_socket;
	// get it
	XmlDoc *xd = &st->m_xd;

	if ( ! xd->m_loaded ) {
		// setting just the docid. niceness is 0.
		xd->set3 ( st->m_docId , st->m_coll , 0 );
		// callback
		xd->setCallback ( state , processLoop );
		// . and tell it to load from the old title rec
		// . if it returns false it blocked and will call our callback
		//   processLoop() when it completes
		if ( ! xd->loadFromOldTitleRec ( ) ) return false;
	}

	if ( g_errno ) return sendErrorReply ( st , g_errno );


	// get the utf8 content
	char **utf8 = xd->getUtf8Content();
	//int32_t   len  = xd->size_utf8Content - 1;
	// wait if blocked???
	if ( utf8 == (void *)-1 ) return false;
	// strange
	if ( xd->size_utf8Content<=0) return sendErrorReply(st,EBADENGINEER );
	// alloc error?
	if ( ! utf8 ) return sendErrorReply ( st , g_errno );

	// get this host
	Host *h = g_hostdb.getHost ( g_hostdb.m_hostId );
	if ( ! h ) return sendErrorReply ( st , EBADENGINEER );
	
	// make it into an editable page now for the turk guy
	sendTurkPageReply ( st );
}

//.........这里部分代码省略.........
 	//no more setting the default root quality to 30, instead if we do not
 	// know it setting it to -1
 	st->m_rootQuality=-1;

	// header
	//xbuf->safePrintf("<meta http-equiv=\"Content-Type\" "
	//		 "content=\"text/html; charset=utf-8\">\n");

	XmlDoc *xd = &st->m_xd;

	long isXml = r->getLong("xml",0);

	// if got docid, use that
	if ( st->m_docId != -1 ) {
		if ( ! xd->set3 ( st->m_docId,
				  st->m_coll,
				  0 ) ) // niceness
			// return error reply if g_errno is set
			return sendErrorReply ( st , g_errno );
		// make this our callback in case something blocks
		xd->setCallback ( st , gotXmlDoc );
		xd->m_pbuf = &st->m_wbuf;
		// reset this flag
		st->m_donePrinting = false;
		// . set xd from the old title rec if recycle is true
		// . can also use XmlDoc::m_loadFromOldTitleRec flag
		//if ( st->m_recycle ) xd->m_recycleContent = true;
		xd->m_recycleContent = true;
		// force this on
		//xd->m_useSiteLinkBuf = true;
		//xd->m_usePageLinkBuf = true;
		if ( isXml ) xd->m_printInXml = true;
		// now tell it to fetch the old title rec
		if ( ! xd->loadFromOldTitleRec () )
			// return false if this blocks
			return false;
		return gotXmlDoc ( st );
	}			

	// set this up
	SpiderRequest sreq;
	sreq.reset();
	if ( st->m_u ) strcpy(sreq.m_url,st->m_u);
	long firstIp = hash32n(st->m_u);
	if ( firstIp == -1 || firstIp == 0 ) firstIp = 1;
	// parentdocid of 0
	sreq.setKey( firstIp, 0LL, false );
	sreq.m_isPageParser = 1;
	sreq.m_hopCount = st->m_hopCount;
	sreq.m_hopCountValid = 1;
	sreq.m_fakeFirstIp   = 1;
	sreq.m_firstIp = firstIp;
	Url nu;
	nu.set(sreq.m_url);
	sreq.m_domHash32 = nu.getDomainHash32();
	sreq.m_siteHash32 = nu.getHostHash32();

	// . get provided content if any
	// . will be NULL if none provided
	// . "content" may contain a MIME
	long  contentLen = 0;
	char *content = r->getString ( "content" , &contentLen , NULL );
	// is the "content" url-encoded? default is true.
	bool contentIsEncoded = true;
	// mark doesn't like to url-encode his content
	if ( ! content ) {

// returns false if blocked, true otherwise
bool processLoop ( void *state ) {
	// get it
	State2 *st = (State2 *)state;
	// get the tcp socket from the state
	TcpSocket *s = st->m_socket;
	// get it
	XmlDoc *xd = &st->m_xd;

	if ( ! xd->m_loaded ) {
		// setting just the docid. niceness is 0.
		//xd->set3 ( st->m_docId , st->m_coll , 0 );
		// callback
		xd->setCallback ( state , processLoop );
		// . and tell it to load from the old title rec
		// . this sets xd->m_oldTitleRec/m_oldTitleRecSize
		// . this sets xd->ptr_* and all other member vars from
		//   the old title rec if found in titledb.
		if ( ! xd->loadFromOldTitleRec ( ) ) return false;
	}

	if ( g_errno ) return sendErrorReply ( st , g_errno );
	// now force it to load old title rec
	//char **tr = xd->getTitleRec();
	SafeBuf *tr = xd->getTitleRecBuf();
	// blocked? return false if so. it will call processLoop() when it rets
	if ( tr == (void *)-1 ) return false;
	// we did not block. check for error? this will free "st" too.
	if ( ! tr ) return sendErrorReply ( st , g_errno );
	// if title rec was empty, that is a problem
	if ( xd->m_titleRecBuf.length() == 0 ) 
		return sendErrorReply ( st , ENOTFOUND);

	// set callback
	char *na = xd->getIsNoArchive();
	// wait if blocked
	if ( na == (void *)-1 ) return false;
	// error?
	if ( ! na ) return sendErrorReply ( st , g_errno );
	// forbidden? allow turkeys through though...
	if ( ! st->m_isAdmin && *na )
		return sendErrorReply ( st , ENOCACHE );

	SafeBuf *sb = &st->m_sb;


	// &page=4 will print rainbow sections
	if ( ! st->m_printed && st->m_r.getLong("page",0) ) {
		// do not repeat this call
		st->m_printed = true;
		// this will call us again since we called
		// xd->setCallback() above to us
		if ( ! xd->printDocForProCog ( sb , &st->m_r ) )
			return false;
	}

	char *contentType = "text/html";
	char format = st->m_format;
	if ( format == FORMAT_XML ) contentType = "text/xml";
	if ( format == FORMAT_JSON ) contentType = "application/json";

	// if we printed a special page (like rainbow sections) then return now
	if ( st->m_printed ) {
		bool status = g_httpServer.sendDynamicPage (s,
							    //buf,bufLen,
							    sb->getBufStart(),
							    sb->getLength(),
							    -1,false,
							    //"text/html",
							    contentType,
							    -1, NULL, "utf8" );
		// nuke state2
		mdelete ( st , sizeof(State2) , "PageGet1" );
		delete (st);
		return status;
	}

	/*
	  // this was calling XmlDoc and setting sections, etc. to
	  // get the SpiderReply junk... no no no
	// is it banned or filtered? this ignores the TagRec in the titleRec
	// and uses msg8a to get it fresh instead
	char *vi = xd->getIsFiltered();//Visible( );
	// wait if blocked
	if ( vi == (void *)-1 ) return false;
	// error?
	if ( ! vi ) return sendErrorReply ( st , g_errno );
	// banned?
	if ( ! st->m_isAdmin && ! *vi ) return sendErrorReply (st,EDOCBANNED);
	*/

	// get the utf8 content
	char **utf8 = xd->getUtf8Content();
	//long   len  = xd->size_utf8Content - 1;
	// wait if blocked???
	if ( utf8 == (void *)-1 ) return false;
	// strange
	if ( xd->size_utf8Content<=0) {
		log("pageget: utf8 content <= 0");
		return sendErrorReply(st,EBADENGINEER );
//.........这里部分代码省略.........