ViewVC Help

View File | Revision Log | Show Annotations | Download File | View Changeset | Root Listing

root/proj/trunk/Search/HttpHandler.cpp

Comparing trunk/Search/HttpHandler.cpp (file contents):
Revision 25 by douglas, 2002-12-22T23:32:58-08:00 vs.
Revision 220 by douglas, 2003-07-23T21:37:53-07:00

#	Line 1 | Line 1
1		/* ============================================================================
2		* Douglas Thrift's Search Engine License
3		*
4	<	* Copyright (C) 2002, Douglas Thrift. All Rights Reserved.
4	>	* Copyright (C) 2002-2003, Douglas Thrift. All Rights Reserved.
5		* Redistribution and use in source and binary forms, with or without
6		* modification, are permitted provided that the following conditions are met:
7		*
#	Line 46 | Line 46
46		//
47		// Douglas Thrift
48		//
49	<	// HttpHandler.cpp
49	>	// $Id: HttpHandler.cpp,v 1.26 2003/07/24 04:37:53 douglas Exp $
50
51		#include "HttpHandler.h"
52
53	+	// Lovely C Sockets!
54	+	#ifndef _WIN32
55	+	// BSD Sockets
56	+	#include <unistd.h>
57	+	#include <sys/types.h>
58	+	#include <sys/socket.h>
59	+	#include <netinet/in.h>
60	+	#include <netdb.h>
61	+
62	+	inline int closesocket(SOCKET s) { return close(s); }
63	+	#endif
64	+
65		HttpHandler::HttpHandler()
66		{
67		buffer = new char[BUFSIZ + 1];
#	Line 62 | Line 74 | HttpHandler::HttpHandler()
74		}
75		#endif // _WIN32
76
77	+	binary = false;
78		length = 0;
79		chunked = false;
80	+	#ifdef _OpenSSL_
81	+	tls = false;
82	+	#endif
83		}
84
85		HttpHandler::~HttpHandler()
#	Line 106 | Line 122 | bool HttpHandler::handle(URL &url, const
122		return answer;
123		}
124
125	+	#ifdef _OpenSSL_
126	+	if (url.getTls())
127	+	{
128	+	tls = true;
129	+
130	+	if (!starttls()) return answer;
131	+	}
132	+	#endif
133	+
134		if (head)
135		{
136		putline("HEAD " + url.getPath() + " HTTP/1.1");
#	Line 116 | Line 141 | bool HttpHandler::handle(URL &url, const
141		}
142
143		putline("Accept: text/html; text/plain");
144	+	#ifndef _OpenSSL_
145		putline("User-Agent: " + agent(true) + ' ' + platform());
146
147		if (url.getPort() == 80)
148	+	#else
149	+	putline("User-Agent: " + agent(true) + ' ' + platform() + ' '
150	+	+ openssl(true));
151	+
152	+	if (url.getPort() == 80 && tls || url.getPort() == 443 && tls)
153	+	#endif
154		{
155		putline("Host: " + url.getAddress());
156		}
157		else
158		{
159	<	char* port = new char[1024];
128	<	sprintf(port, "%u", url.getPort());
159	>	ostringstream port;
160
161	<	putline("Host: " + url.getAddress() + ':' + port);
161	>	port << url.getPort();
162
163	<	delete [] port;
163	>	putline("Host: " + url.getAddress() + ':' + port.str());
164		}
165
166		if (referer != "")
#	Line 155 | Line 186 | bool HttpHandler::handle(URL &url, const
186		unsigned dot = line.find('.');
187		unsigned space = line.find(' ');
188
189	<	unsigned major = strtoul(line.substr(5, dot - 5).c_str(), 0, 10);
190	<	unsigned minor = strtoul(line.substr(dot + 1, space - dot - 1).c_str(),
191	<	0, 10);
189	>	unsigned major;
190	>	unsigned minor;
191	>
192	>	istringstream number(line.substr(5, dot - 5) + " " + line.substr(dot
193	>	+ 1, space - dot - 1));
194	>
195	>	number >> major;
196	>	number >> minor;
197
198		if (major > 1)
199		{
#	Line 167 | Line 203 | bool HttpHandler::handle(URL &url, const
203		return answer;
204		}
205
206	<	response = code(strtoul(line.substr(space + 1).c_str(), 0, 10));
206	>	number.clear();
207	>	number.str(line.substr(space + 1, 3));
208	>	number >> response;
209
210		if (response < ok) do line = getline(); while (line != "");
211		}
#	Line 192 | Line 230 | bool HttpHandler::handle(URL &url, const
230		}
231		else if (field == "Content-Length")
232		{
233	<	length = strtoul(value.c_str(), 0, 10);
233	>	istringstream number(value);
234	>
235	>	number >> length;
236		}
237		else if (field == "Location")
238		{
#	Line 241 | Line 281 | bool HttpHandler::handle(URL &url, const
281		return answer;
282		}
283
284	<	HttpHandler& HttpHandler::getline(string& line, char endline)
284	>	void HttpHandler::clear()
285		{
286	<	unsigned end = page.find(endline);
287	<	unsigned newline = page.find('\n');
248	<
249	<	if (newline < end || end == string::npos)
286	>	#ifdef _OpenSSL_
287	>	if (tls)
288		{
289	<	end = newline;
289	>	SSL_shutdown(ssl);
290	>	SSL_free(ssl);
291	>	SSL_CTX_free(ctx);
292		}
293	+	#endif
294
254	–	line = page.substr(0, end);
255	–	page.erase(0, (end == string::npos ? end : end + 1));
256	–
257	–	return *this;
258	–	}
259	–
260	–	void HttpHandler::clear()
261	–	{
295		closesocket(http);
296
297		type = "";
298		length = 0;
299		location = "";
300	<	page = "";
300	>	page.clear();
301	>	page.str("");
302		chunked = false;
303	+	#ifdef _OpenSSL_
304	+	tls = false;
305	+	#endif
306		}
307
308		void HttpHandler::populate()
#	Line 279 | Line 316 | void HttpHandler::populate()
316		memset(buffer, 0, BUFSIZ + 1);
317
318		unsigned bytes = left > BUFSIZ ? BUFSIZ : left;
319	<	unsigned received;
319	>	long received;
320
321	<	if ((received = recv(http, buffer, bytes, 0)) == SOCKET_ERROR)
285	<	{
286	<	error(program + ": Recv");
287	<	exit(1);
288	<	}
289	<	else if (received != bytes)
321	>	while (true)
322		{
323	<	left -= received;
324	<	page += buffer;
293	<
294	<	memset(buffer, 0, BUFSIZ + 1);
295	<
296	<	bytes -= received;
297	<	if (recv(http, buffer, bytes, 0) == SOCKET_ERROR)
323	>	#ifndef _OpenSSL_
324	>	if ((received = recv(http, buffer, bytes, 0)) == SOCKET_ERROR)
325		{
326		error(program + ": Recv");
327		exit(1);
328		}
329	+	#else
330	+	if ((received = !tls ? recv(http, buffer, bytes, 0) :
331	+	SSL_read(ssl, buffer, bytes)) <= 0)
332	+	{
333	+	!tls ? error(program + ": Recv") : error(program +
334	+	": SSL Read", int(received));
335	+	}
336	+	#endif
337	+	else if (received != bytes)
338	+	{
339	+	left -= received;
340	+	page << buffer;
341	+
342	+	memset(buffer, 0, BUFSIZ + 1);
343	+
344	+	bytes -= received;
345	+	}
346	+	else
347	+	{
348	+	break;
349	+	}
350		}
351
352	<	page += buffer;
352	>	page << buffer;
353		left -= bytes;
354		}
355		}
#	Line 311 | Line 359 | void HttpHandler::populate()
359
360		do
361		{
362	<	chunk = strtoul(getline().c_str(), 0, 16);
362	>	istringstream number(getline());
363	>
364	>	number.setf(ios_base::hex, ios_base::basefield);
365	>	number >> chunk;
366
367		unsigned left = chunk;
368
#	Line 320 | Line 371 | void HttpHandler::populate()
371		memset(buffer, 0, BUFSIZ + 1);
372
373		unsigned bytes = left > BUFSIZ ? BUFSIZ : left;
374	<	unsigned received;
374	>	long received;
375
376	<	if ((received = recv(http, buffer, bytes, 0)) == SOCKET_ERROR)
326	<	{
327	<	error(program + ": Recv");
328	<	exit(1);
329	<	}
330	<	else if (received != bytes)
376	>	while (true)
377		{
378	<	left -= received;
379	<	page += buffer;
380	<
335	<	memset(buffer, 0, BUFSIZ + 1);
336	<
337	<	bytes -= received;
338	<	if (recv(http, buffer, bytes, 0) == SOCKET_ERROR)
378	>	#ifndef _OpenSSL_
379	>	if ((received = recv(http, buffer, bytes, 0)) ==
380	>	SOCKET_ERROR)
381		{
382		error(program + ": Recv");
383		exit(1);
384		}
385	+	#else
386	+	if ((received = !tls ? recv(http, buffer, bytes, 0) :
387	+	SSL_read(ssl, buffer, bytes)) <= 0)
388	+	{
389	+	!tls ? error(program + ": Recv") : error(program +
390	+	": SSL Read", int(received));
391	+	exit(1);
392	+	}
393	+	#endif
394	+	else if (received != bytes)
395	+	{
396	+	left -= received;
397	+	page << buffer;
398	+
399	+	memset(buffer, 0, BUFSIZ + 1);
400	+
401	+	bytes -= received;
402	+	}
403	+	else
404	+	{
405	+	break;
406	+	}
407		}
408
409	<	page += buffer;
409	>	page << buffer;
410		left -= bytes;
411		}
412
#	Line 352 | Line 416 | void HttpHandler::populate()
416		while (chunk > 0);
417		}
418
419	<	for (unsigned index = 0; index < page.length(); index++)
419	>	if (!binary)
420		{
421	<	if (page[index] == '\r' && (index + 1 < page.length()) ? page[index +
422	<	1] == '\n' : false)
423	<	{
360	<	page.erase(index, 1);
361	<	}
362	<	else if (page[index] == '\r')
421	>	string page = this->page.str();
422	>
423	>	for (unsigned index = 0; index < page.length(); index++)
424		{
425	<	page[index] = '\n';
425	>	if (page[index] == '\r' && (index + 1 < page.length()) ? page[index +
426	>	1] == '\n' : false)
427	>	{
428	>	page.erase(index, 1);
429	>	}
430	>	else if (page[index] == '\r')
431	>	{
432	>	page[index] = '\n';
433	>	}
434		}
435	+
436	+	this->page.str(page);
437		}
438		}
439
440		void HttpHandler::putline(const string line)
441		{
442		sprintf(buffer, "%s\r\n", line.c_str());
443	+
444	+	#ifndef _OpenSSL_
445		if (send(http, buffer, strlen(buffer), 0) == SOCKET_ERROR)
446		{
447		error(program + ": Send");
448		exit(1);
449		}
450	+	#else
451	+	if (!tls)
452	+	{
453	+	if (send(http, buffer, strlen(buffer), 0) == SOCKET_ERROR)
454	+	{
455	+	error(program + ": Send");
456	+	exit(1);
457	+	}
458	+	}
459	+	else
460	+	{
461	+	int number;
462	+
463	+	if ((number = SSL_write(ssl, buffer, strlen(buffer))) <= 0)
464	+	{
465	+	error(program + ": SSL Write", number);
466	+	exit(1);
467	+	}
468	+	}
469	+	#endif
470		}
471
472		string HttpHandler::getline()
#	Line 383 | Line 476 | string HttpHandler::getline()
476
477		do
478		{
479	+	#ifndef _OpenSSL_
480		if (recv(http, &byte, 1, 0) == SOCKET_ERROR)
481		{
482		error(program + ": Recv");
483		}
484	+	#else
485	+	if (!tls)
486	+	{
487	+	if (recv(http, &byte, 1, 0) == SOCKET_ERROR)
488	+	{
489	+	error(program + ": Recv");
490	+	}
491	+	}
492	+	else
493	+	{
494	+	int number;
495	+
496	+	if ((number = SSL_read(ssl, &byte, 1)) <= 0)
497	+	{
498	+	error(program + ": SSL Read", number);
499	+	}
500	+	}
501	+	#endif
502
503		if (byte != '\r' && byte != '\n')
504		{
#	Line 406 | Line 518 | void HttpHandler::error(const string& pr
518		switch (WSAGetLastError())
519		{
520		case WSAEACCES:
521	<	error = "Permission denied.";
521	>	error = "Permission denied";
522		break;
523		case WSAEADDRINUSE:
524	<	error = "Address already in use.";
524	>	error = "Address already in use";
525		break;
526		case WSAEADDRNOTAVAIL:
527	<	error = "Cannot assign requested address.";
527	>	error = "Cannot assign requested address";
528		break;
529		case WSAEAFNOSUPPORT:
530	<	error = "Address family not supported by protocol family.";
530	>	error = "Address family not supported by protocol family";
531		break;
532		case WSAEALREADY:
533	<	error = "Operation already in progress.";
533	>	error = "Operation already in progress";
534		break;
535		case WSAECONNABORTED:
536	<	error = "Software caused connection abort.";
536	>	error = "Software caused connection abort";
537		break;
538		case WSAECONNREFUSED:
539	<	error = "Connection refused.";
539	>	error = "Connection refused";
540		break;
541		case WSAECONNRESET:
542	<	error = "Connection reset by peer.";
542	>	error = "Connection reset by peer";
543		break;
544		case WSAEDESTADDRREQ:
545	<	error = "Destination address required.";
545	>	error = "Destination address required";
546		break;
547		case WSAEFAULT:
548	<	error = "Bad address.";
548	>	error = "Bad address";
549		break;
550		case WSAEHOSTDOWN:
551	<	error = "Host is down.";
551	>	error = "Host is down";
552		break;
553		case WSAEHOSTUNREACH:
554	<	error = "No route to host.";
554	>	error = "No route to host";
555		break;
556		case WSAEINPROGRESS:
557	<	error = "Operation now in progress.";
557	>	error = "Operation now in progress";
558		break;
559		case WSAEINTR:
560	<	error = "Interrupted function call.";
560	>	error = "Interrupted function call";
561		break;
562		case WSAEINVAL:
563	<	error = "Invalid argument.";
563	>	error = "Invalid argument";
564		break;
565		case WSAEISCONN:
566	<	error = "Socket is already connected.";
566	>	error = "Socket is already connected";
567		break;
568		case WSAEMFILE:
569	<	error = "Too many open files.";
569	>	error = "Too many open files";
570		break;
571		case WSAEMSGSIZE:
572	<	error = "Message too long.";
572	>	error = "Message too long";
573		break;
574		case WSAENETDOWN:
575	<	error = "Network is down.";
575	>	error = "Network is down";
576		break;
577		case WSAENETRESET:
578	<	error = "Network dropped connection on reset.";
578	>	error = "Network dropped connection on reset";
579		break;
580		case WSAENETUNREACH:
581	<	error = "Network is unreachable.";
581	>	error = "Network is unreachable";
582		break;
583		case WSAENOBUFS:
584	<	error = "No buffer space available.";
584	>	error = "No buffer space available";
585		break;
586		case WSAENOPROTOOPT:
587	<	error = "Bad protocol option.";
587	>	error = "Bad protocol option";
588		break;
589		case WSAENOTCONN:
590	<	error = "Socket is not connected.";
590	>	error = "Socket is not connected";
591		break;
592		case WSAENOTSOCK:
593	<	error = "Socket operation on non-socket.";
593	>	error = "Socket operation on non-socket";
594		break;
595		case WSAEOPNOTSUPP:
596	<	error = "Operation not supported.";
596	>	error = "Operation not supported";
597		break;
598		case WSAEPFNOSUPPORT:
599	<	error = "Protocol family not supported.";
599	>	error = "Protocol family not supported";
600		break;
601		case WSAEPROCLIM:
602	<	error = "Too many processes.";
602	>	error = "Too many processes";
603		break;
604		case WSAEPROTONOSUPPORT:
605	<	error = "Protocol not supported.";
605	>	error = "Protocol not supported";
606		break;
607		case WSAEPROTOTYPE:
608	<	error = "Protocol wrong type for socket.";
608	>	error = "Protocol wrong type for socket";
609		break;
610		case WSAESHUTDOWN:
611	<	error = "Cannot send after socket shutdown.";
611	>	error = "Cannot send after socket shutdown";
612		break;
613		case WSAESOCKTNOSUPPORT:
614	<	error = "Socket type not supported.";
614	>	error = "Socket type not supported";
615		break;
616		case WSAETIMEDOUT:
617	<	error = "Connection timed out.";
617	>	error = "Connection timed out";
618		break;
619		case WSATYPE_NOT_FOUND:
620	<	error = "Class type not found.";
620	>	error = "Class type not found";
621		break;
622		case WSAEWOULDBLOCK:
623	<	error = "Resource temporarily unavailable.";
623	>	error = "Resource temporarily unavailable";
624		break;
625		case WSAHOST_NOT_FOUND:
626	<	error = "Host not found.";
626	>	error = "Host not found";
627		break;
628		case WSA_INVALID_HANDLE:
629	<	error = "Specified event object handle is invalid.";
629	>	error = "Specified event object handle is invalid";
630		break;
631		case WSA_INVALID_PARAMETER:
632	<	error = "One or more parameters are invalid.";
632	>	error = "One or more parameters are invalid";
633		break;
634		//      case WSAINVALIDPROCTABLE:
635	<	//              error = "Invalid procedure table from service provider.";
635	>	//              error = "Invalid procedure table from service provider";
636		//              break;
637		//      case WSAINVALIDPROVIDER:
638	<	//              error = "Invalid service provider version number.";
638	>	//              error = "Invalid service provider version number";
639		//              break;
640		case WSA_IO_INCOMPLETE:
641	<	error = "Overlapped I/O event object not in signaled state.";
641	>	error = "Overlapped I/O event object not in signaled state";
642		break;
643		case WSA_IO_PENDING:
644	<	error = "Overlapped operations will complete later.";
644	>	error = "Overlapped operations will complete later";
645		break;
646		case WSA_NOT_ENOUGH_MEMORY:
647	<	error = "Insufficient memory available.";
647	>	error = "Insufficient memory available";
648		break;
649		case WSANOTINITIALISED:
650	<	error = "Successful WSAStartup not yet performed.";
650	>	error = "Successful WSAStartup not yet performed";
651		break;
652		case WSANO_DATA:
653	<	error = "Valid name, no data record of requested type.";
653	>	error = "Valid name, no data record of requested type";
654		break;
655		case WSANO_RECOVERY:
656	<	error = "This is a non-recoverable error.";
656	>	error = "This is a non-recoverable error";
657		break;
658		//      case WSAPROVIDERFAILEDINIT:
659	<	//              error = "Unable to initialize a service provider.";
659	>	//              error = "Unable to initialize a service provider";
660		//              break;
661		case WSASYSCALLFAILURE:
662	<	error = "System call failure.";
662	>	error = "System call failure";
663		break;
664		case WSASYSNOTREADY:
665	<	error = "Network subsystem is unavailable.";
665	>	error = "Network subsystem is unavailable";
666		break;
667		case WSATRY_AGAIN:
668	<	error = "Non-authoritative host not found.";
668	>	error = "Non-authoritative host not found";
669		break;
670		case WSAVERNOTSUPPORTED:
671	<	error = "WINSOCK.DLL version out of range.";
671	>	error = "WINSOCK.DLL version out of range";
672		break;
673		case WSAEDISCON:
674	<	error = "Graceful shutdown in progress.";
674	>	error = "Graceful shutdown in progress";
675		break;
676		case WSA_OPERATION_ABORTED:
677	<	error = "Overlapped operation aborted.";
677	>	error = "Overlapped operation aborted";
678		break;
679		default:
680	<	error = "Unknown error.";
680	>	error = "Unknown error";
681		break;
682		}
683
#	Line 573 | Line 685 | void HttpHandler::error(const string& pr
685		#else
686		if (host)
687		{
688	<	herror(prefix.c_str());
688	>	string error;
689	>
690	>	switch (h_errno)
691	>	{
692	>	case HOST_NOT_FOUND:
693	>	error = "Unknown host";
694	>	break;
695	>	case TRY_AGAIN:
696	>	error = "Host name lookup failure";
697	>	break;
698	>	case NO_RECOVERY:
699	>	error = "Unknown server error";
700	>	break;
701	>	case NO_DATA:
702	>	error = "No address associated with name";
703	>	break;
704	>	default:
705	>	error = "Unknown error";
706	>	break;
707	>	}
708	>
709	>	cerr << prefix << ": " << error << "\n";
710		}
711		else
712		{
#	Line 581 | Line 714 | void HttpHandler::error(const string& pr
714		}
715		#endif // _WIN32
716		}
717	+
718	+	#ifdef _OpenSSL_
719	+	void HttpHandler::error(const string& prefix, int number)
720	+	{
721	+	string error;
722	+
723	+	switch (SSL_get_error(ssl, number))
724	+	{
725	+	case SSL_ERROR_NONE:
726	+	error = "The TLS/SSL I/O operation completed";
727	+	break;
728	+	case SSL_ERROR_ZERO_RETURN:
729	+	error = "The TLS/SSL connection has been closed";
730	+	break;
731	+	case SSL_ERROR_WANT_READ:
732	+	case SSL_ERROR_WANT_WRITE:
733	+	case SSL_ERROR_WANT_CONNECT:
734	+	//      case SSL_ERROR_WANT_ACCEPT:
735	+	case SSL_ERROR_WANT_X509_LOOKUP:
736	+	error = "The operation did not complete";
737	+	break;
738	+	case SSL_ERROR_SYSCALL:
739	+	if (int err = ERR_get_error() != 0)
740	+	{
741	+	error = ERR_reason_error_string(err);
742	+	}
743	+	else
744	+	{
745	+	switch (number)
746	+	{
747	+	case 0:
748	+	error = "An EOF was observed that violates the protocol";
749	+	break;
750	+	case -1:
751	+	this->error(prefix);
752	+	return;
753	+	default:
754	+	error = "Unknown error";
755	+	break;
756	+	}
757	+	}
758	+	break;
759	+	case SSL_ERROR_SSL:
760	+	error = ERR_reason_error_string(ERR_get_error());
761	+	break;
762	+	default:
763	+	error = "Unknown error";
764	+	break;
765	+	}
766	+
767	+	cerr << prefix << ": " << error << "\n";
768	+	}
769	+
770	+	bool HttpHandler::starttls()
771	+	{
772	+	SSL_load_error_strings();
773	+	SSL_library_init();
774	+
775	+	#ifndef _urandomdev_
776	+	int pid = getpid();
777	+	int now = time(NULL);
778	+
779	+	unsigned seed = now > pid ? now - pid : pid - now;
780	+
781	+	char* junk = new char[seed % 30 + 2];
782	+	junk[0] = pid;
783	+	junk[seed % 30 + 1] = now;
784	+
785	+	srand(seed);
786	+
787	+	for (int index = 1; index < seed % 30 + 1; index++)
788	+	{
789	+	junk[index] = rand();
790	+	}
791	+
792	+	if (debug)
793	+	{
794	+	cerr << "junk = {\n";
795	+
796	+	for (int index = 1; index < seed % 30 + 2; index++)
797	+	{
798	+	cerr << "   [" << index << "] = " << int(junk[index]) << "\n";
799	+	}
800	+
801	+	cerr << "}\n";
802	+	}
803	+
804	+	RAND_seed(junk, seed % 30 + 2);
805	+
806	+	delete junk;
807	+	#else
808	+	if (debug) cerr << "junk = /dev/urandom\n";
809	+	#endif
810	+
811	+	ctx = SSL_CTX_new(TLSv1_client_method());
812	+
813	+	if (ctx == NULL)
814	+	{
815	+	cerr << program << ": SSL CTX New: "
816	+	<< ERR_reason_error_string(ERR_get_error()) << "\n";
817	+	return false;
818	+	}
819	+
820	+	ssl = SSL_new(ctx);
821	+
822	+	if (SSL_set_fd(ssl, http) == 0)
823	+	{
824	+	cerr << program << ": SSL Set FD: "
825	+	<< ERR_reason_error_string(ERR_get_error()) << "\n";
826	+	return false;
827	+	}
828	+
829	+	int number;
830	+
831	+	if ((number = SSL_connect(ssl)) <= 0)
832	+	{
833	+	error(program + ": SSL Connect", number);
834	+	return false;
835	+	}
836	+
837	+	return true;
838	+	}
839	+	#endif

Diff Legend

–	Removed lines
+	Added lines
<	Changed lines
>	Changed lines