nanohttp.c 42.4 KB
Newer Older
1 2 3 4 5 6 7 8 9 10
/*
 * nanohttp.c: minimalist HTTP GET implementation to fetch external subsets.
 *             focuses on size, streamability, reentrancy and portability
 *
 * This is clearly not a general purpose HTTP implementation
 * If you look for one, check:
 *         http://www.w3.org/Library/
 *
 * See Copyright for the status of this software.
 *
11
 * daniel@veillard.com
12 13 14 15 16
 */
 
/* TODO add compression support, Send the Accept- , and decompress on the
        fly with ZLIB if found at compile-time */

17
#define NEED_SOCKETS
18
#define IN_LIBXML
19
#include "libxml.h"
20 21 22 23 24 25 26 27 28 29

#ifdef LIBXML_HTTP_ENABLED
#include <string.h>

#ifdef HAVE_STDLIB_H
#include <stdlib.h>
#endif
#ifdef HAVE_UNISTD_H
#include <unistd.h>
#endif
30 31 32
#ifdef HAVE_SYS_TYPES_H
#include <sys/types.h>
#endif
33 34 35 36 37 38 39 40 41 42 43 44
#ifdef HAVE_SYS_SOCKET_H
#include <sys/socket.h>
#endif
#ifdef HAVE_NETINET_IN_H
#include <netinet/in.h>
#endif
#ifdef HAVE_ARPA_INET_H
#include <arpa/inet.h>
#endif
#ifdef HAVE_NETDB_H
#include <netdb.h>
#endif
45
#ifdef HAVE_RESOLV_H
46 47 48
#ifdef HAVE_ARPA_NAMESER_H
#include <arpa/nameser.h>
#endif
49 50
#include <resolv.h>
#endif
51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75
#ifdef HAVE_FCNTL_H
#include <fcntl.h> 
#endif
#ifdef HAVE_ERRNO_H
#include <errno.h>
#endif
#ifdef HAVE_SYS_TIME_H
#include <sys/time.h>
#endif
#ifdef HAVE_SYS_SELECT_H
#include <sys/select.h>
#endif
#ifdef HAVE_STRINGS_H
#include <strings.h>
#endif
#ifdef SUPPORT_IP6
#include <resolv.h>
#endif

#ifdef VMS
#include <stropts>
#define SOCKLEN_T unsigned int
#define SOCKET int
#endif

76 77 78 79 80 81 82 83 84 85

#ifdef __MINGW32__
#define _WINSOCKAPI_
#include <wsockcompat.h>
#include <winsock2.h>
#undef SOCKLEN_T
#define SOCKLEN_T unsigned int
#endif


86
#include <libxml/globals.h>
87
#include <libxml/xmlerror.h>
88 89 90
#include <libxml/xmlmemory.h>
#include <libxml/parser.h> /* for xmlStr(n)casecmp() */
#include <libxml/nanohttp.h>
91
#include <libxml/globals.h>
92
#include <libxml/uri.h>
93 94 95 96 97

/**
 * A couple portability macros
 */
#ifndef _WINSOCKAPI_
98
#ifndef __BEOS__
99
#define closesocket(s) close(s)
100
#endif
101 102 103
#define SOCKET int
#endif

104 105 106 107 108 109
#ifdef __BEOS__
#ifndef PF_INET
#define PF_INET AF_INET
#endif
#endif

110 111 112 113 114 115
#ifndef SOCKLEN_T
#define SOCKLEN_T unsigned int
#endif
#ifndef SOCKET
#define SOCKET int
#endif
116

117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147
#ifdef STANDALONE
#define DEBUG_HTTP
#define xmlStrncasecmp(a, b, n) strncasecmp((char *)a, (char *)b, n)
#define xmlStrcasecmpi(a, b) strcasecmp((char *)a, (char *)b)
#endif

#define XML_NANO_HTTP_MAX_REDIR	10

#define XML_NANO_HTTP_CHUNK	4096

#define XML_NANO_HTTP_CLOSED	0
#define XML_NANO_HTTP_WRITE	1
#define XML_NANO_HTTP_READ	2
#define XML_NANO_HTTP_NONE	4

typedef struct xmlNanoHTTPCtxt {
    char *protocol;	/* the protocol name */
    char *hostname;	/* the host name */
    int port;		/* the port */
    char *path;		/* the path within the URL */
    SOCKET fd;		/* the file descriptor for the socket */
    int state;		/* WRITE / READ / CLOSED */
    char *out;		/* buffer sent (zero terminated) */
    char *outptr;	/* index within the buffer sent */
    char *in;		/* the receiving buffer */
    char *content;	/* the start of the content */
    char *inptr;	/* the next byte to read from network */
    char *inrptr;	/* the next byte to give back to the client */
    int inlen;		/* len of the input buffer */
    int last;		/* return code for last operation */
    int returnValue;	/* the protocol return value */
148
    int ContentLength;  /* specified content length from HTTP header */
149 150 151
    char *contentType;	/* the MIME type for the input */
    char *location;	/* the new URL in case of redirect */
    char *authHeader;	/* contents of {WWW,Proxy}-Authenticate header */
152
    char *encoding;	/* encoding extracted from the contentType */
153
    char *mimeType;	/* Mime-Type extracted from the contentType */
154 155 156 157 158 159 160
} xmlNanoHTTPCtxt, *xmlNanoHTTPCtxtPtr;

static int initialized = 0;
static char *proxy = NULL;	 /* the proxy name if any */
static int proxyPort;	/* the proxy port if any */
static unsigned int timeout = 60;/* the select() timeout in seconds */

161 162 163
int xmlNanoHTTPFetchContent( void * ctx, char ** ptr, int * len );
int xmlNanoHTTPContentLength( void * ctx );

164 165 166 167 168 169 170 171 172 173 174 175
/**
 * xmlHTTPErrMemory:
 * @extra:  extra informations
 *
 * Handle an out of memory condition
 */
static void
xmlHTTPErrMemory(const char *extra)
{
    __xmlSimpleError(XML_FROM_HTTP, XML_ERR_NO_MEMORY, NULL, NULL, extra);
}

176 177 178
/**
 * A portability function
 */
179
static int socket_errno(void) {
180 181 182 183 184 185 186
#ifdef _WINSOCKAPI_
    return(WSAGetLastError());
#else
    return(errno);
#endif
}

187
#ifdef SUPPORT_IP6
188 189
static
int have_ipv6(void) {
190 191 192 193 194 195 196 197 198 199 200
    int s;

    s = socket (AF_INET6, SOCK_STREAM, 0);
    if (s != -1) {
	close (s);
	return (1);
    }
    return (0);
}
#endif

201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243
/**
 * xmlNanoHTTPInit:
 *
 * Initialize the HTTP protocol layer.
 * Currently it just checks for proxy informations
 */

void
xmlNanoHTTPInit(void) {
    const char *env;
#ifdef _WINSOCKAPI_
    WSADATA wsaData;    
#endif

    if (initialized)
	return;

#ifdef _WINSOCKAPI_
    if (WSAStartup(MAKEWORD(1, 1), &wsaData) != 0)
	return;
#endif

    if (proxy == NULL) {
	proxyPort = 80;
	env = getenv("no_proxy");
	if (env != NULL)
	    goto done;
	env = getenv("http_proxy");
	if (env != NULL) {
	    xmlNanoHTTPScanProxy(env);
	    goto done;
	}
	env = getenv("HTTP_PROXY");
	if (env != NULL) {
	    xmlNanoHTTPScanProxy(env);
	    goto done;
	}
    }
done:
    initialized = 1;
}

/**
244
 * xmlNanoHTTPCleanup:
245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273
 *
 * Cleanup the HTTP protocol layer.
 */

void
xmlNanoHTTPCleanup(void) {
    if (proxy != NULL)
	xmlFree(proxy);
#ifdef _WINSOCKAPI_
    if (initialized)
	WSACleanup();
#endif
    initialized = 0;
    return;
}

/**
 * xmlNanoHTTPScanURL:
 * @ctxt:  an HTTP context
 * @URL:  The URL used to initialize the context
 *
 * (Re)Initialize an HTTP context by parsing the URL and finding
 * the protocol host port and path it indicates.
 */

static void
xmlNanoHTTPScanURL(xmlNanoHTTPCtxtPtr ctxt, const char *URL) {
    const char *cur = URL;
    char buf[4096];
274
    int indx = 0;
275
    const int indxMax = 4096 - 1;
276 277 278 279 280 281 282 283 284 285 286 287 288 289 290
    int port = 0;

    if (ctxt->protocol != NULL) { 
        xmlFree(ctxt->protocol);
	ctxt->protocol = NULL;
    }
    if (ctxt->hostname != NULL) { 
        xmlFree(ctxt->hostname);
	ctxt->hostname = NULL;
    }
    if (ctxt->path != NULL) { 
        xmlFree(ctxt->path);
	ctxt->path = NULL;
    }
    if (URL == NULL) return;
291
    buf[indx] = 0;
292
    while ((*cur != 0) && (indx < indxMax)) {
293
        if ((cur[0] == ':') && (cur[1] == '/') && (cur[2] == '/')) {
294
	    buf[indx] = 0;
295
	    ctxt->protocol = xmlMemStrdup(buf);
296
	    indx = 0;
297 298 299
            cur += 3;
	    break;
	}
300
	buf[indx++] = *cur++;
301 302 303
    }
    if (*cur == 0) return;

304
    buf[indx] = 0;
305
    while (indx < indxMax) {
306 307
	if ((strchr (cur, '[') && !strchr (cur, ']')) ||
		(!strchr (cur, '[') && strchr (cur, ']'))) {
308 309
	    __xmlIOErr(XML_FROM_HTTP, XML_HTTP_URL_SYNTAX, 
	               "Syntax Error\n");
310 311 312 313 314
	    return;
	}

	if (cur[0] == '[') {
	    cur++;
315
	    while ((cur[0] != ']') && (indx < indxMax))
316 317 318
		buf[indx++] = *cur++;
    
	    if (!strchr (buf, ':')) {
319
		__xmlIOErr(XML_FROM_HTTP, XML_HTTP_USE_IP,
320 321 322 323
			"Use [IPv6]/IPv4 format\n");
		return;
	    }

324
	    buf[indx] = 0;
325
	    ctxt->hostname = xmlMemStrdup (buf);
326
	    indx = 0;
327
	    cur += 1;
328
	    if (cur[0] == ':') {
329
		cur++;
330 331 332 333 334 335 336 337 338
		while (*cur >= '0' && *cur <= '9') {
		    port *= 10;
		    port += *cur - '0';
		    cur++;
		}

		if (port != 0) ctxt->port = port;
		while ((cur[0] != '/') && (*cur != 0))
		    cur++;
339 340
	    }
	    break;
341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363
        }
	else {
	    if (cur[0] == ':') {
		buf[indx] = 0;
		ctxt->hostname = xmlMemStrdup (buf);
		indx = 0;
		cur += 1;
		while ((*cur >= '0') && (*cur <= '9')) {
		    port *= 10;
		    port += *cur - '0';
		    cur++;
		}
		if (port != 0) ctxt->port = port;
		while ((cur[0] != '/') && (*cur != 0)) 
		    cur++;
		break;
	    }
	    if ((*cur == '/') || (*cur == 0)) {
		buf[indx] = 0;
		ctxt->hostname = xmlMemStrdup (buf);
		indx = 0;
		break;
	    }
364
	}
365
	buf[indx++] = *cur++;
366 367 368 369
    }
    if (*cur == 0) 
        ctxt->path = xmlMemStrdup("/");
    else {
370 371
        indx = 0;
        buf[indx] = 0;
372
	while ((*cur != 0) && (indx < indxMax))
373 374
	    buf[indx++] = *cur++;
	buf[indx] = 0;
375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392
	ctxt->path = xmlMemStrdup(buf);
    }	
}

/**
 * xmlNanoHTTPScanProxy:
 * @URL:  The proxy URL used to initialize the proxy context
 *
 * (Re)Initialize the HTTP Proxy context by parsing the URL and finding
 * the protocol host port it indicates.
 * Should be like http://myproxy/ or http://myproxy:3128/
 * A NULL URL cleans up proxy informations.
 */

void
xmlNanoHTTPScanProxy(const char *URL) {
    const char *cur = URL;
    char buf[4096];
393
    int indx = 0;
394
    const int indxMax = 4096 - 1;
395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412
    int port = 0;

    if (proxy != NULL) { 
        xmlFree(proxy);
	proxy = NULL;
    }
    if (proxyPort != 0) { 
	proxyPort = 0;
    }
#ifdef DEBUG_HTTP
    if (URL == NULL)
	xmlGenericError(xmlGenericErrorContext,
		"Removing HTTP proxy info\n");
    else
	xmlGenericError(xmlGenericErrorContext,
		"Using HTTP proxy %s\n", URL);
#endif
    if (URL == NULL) return;
413
    buf[indx] = 0;
414
    while ((*cur != 0) && (indx < indxMax)) {
415
        if ((cur[0] == ':') && (cur[1] == '/') && (cur[2] == '/')) {
416 417
	    buf[indx] = 0;
	    indx = 0;
418 419 420
            cur += 3;
	    break;
	}
421
	buf[indx++] = *cur++;
422 423 424
    }
    if (*cur == 0) return;

425
    buf[indx] = 0;
426
    while (indx < indxMax) {
427 428
	if ((strchr (cur, '[') && !strchr (cur, ']')) ||
		(!strchr (cur, '[') && strchr (cur, ']'))) {
429
	    __xmlIOErr(XML_FROM_HTTP, XML_HTTP_URL_SYNTAX, "Syntax Error\n");
430 431 432 433 434
	    return;
	}

	if (cur[0] == '[') {
	    cur++;
435
	    while ((cur[0] != ']') && (indx < indxMax))
436 437 438
		buf[indx++] = *cur++;

	    if (!strchr (buf, ':')) {
439
		__xmlIOErr(XML_FROM_HTTP, XML_HTTP_USE_IP,
440 441 442 443
			"Use [IPv6]/IPv4 format\n");
		return;
	    }

444
	    buf[indx] = 0;
445
	    proxy = xmlMemStrdup (buf);
446
	    indx = 0;
447
	    cur += 1;
448
	    if (cur[0] == ':') {
449
	        cur++;
450 451 452 453 454 455 456 457 458 459
		while (*cur >= '0' && *cur <= '9') {
		    port *= 10;
		    port += *cur - '0';
		    cur++;
		}

		if (port != 0) proxyPort = port;
		while ((cur[0] != '/') && (*cur != 0))
		    cur ++;
	    }
460 461
	    break;
	}
462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483
	else {
	    if (cur[0] == ':') {
		buf[indx] = 0;
		proxy = xmlMemStrdup (buf);
		indx = 0;
		cur += 1;
		while ((*cur >= '0') && (*cur <= '9')) {
		    port *= 10;
		    port += *cur - '0';
		    cur++;
		}
		if (port != 0) proxyPort = port;
		while ((cur[0] != '/') && (*cur != 0)) 
		    cur++;
		break;
	    }
	    if ((*cur == '/') || (*cur == 0)) {
		buf[indx] = 0;
		proxy = xmlMemStrdup (buf);
		indx = 0;
		break;
	    }
484
	}
485
	buf[indx++] = *cur++;
486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502
    }
}

/**
 * xmlNanoHTTPNewCtxt:
 * @URL:  The URL used to initialize the context
 *
 * Allocate and initialize a new HTTP context.
 *
 * Returns an HTTP context or NULL in case of error.
 */

static xmlNanoHTTPCtxtPtr
xmlNanoHTTPNewCtxt(const char *URL) {
    xmlNanoHTTPCtxtPtr ret;

    ret = (xmlNanoHTTPCtxtPtr) xmlMalloc(sizeof(xmlNanoHTTPCtxt));
503 504 505 506
    if (ret == NULL) {
        xmlHTTPErrMemory("allocating context");
        return(NULL);
    }
507 508 509 510 511

    memset(ret, 0, sizeof(xmlNanoHTTPCtxt));
    ret->port = 80;
    ret->returnValue = 0;
    ret->fd = -1;
512
    ret->ContentLength = -1;
513

514
    xmlNanoHTTPScanURL(ret, URL);
515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534

    return(ret);
}

/**
 * xmlNanoHTTPFreeCtxt:
 * @ctxt:  an HTTP context
 *
 * Frees the context after closing the connection.
 */

static void
xmlNanoHTTPFreeCtxt(xmlNanoHTTPCtxtPtr ctxt) {
    if (ctxt == NULL) return;
    if (ctxt->hostname != NULL) xmlFree(ctxt->hostname);
    if (ctxt->protocol != NULL) xmlFree(ctxt->protocol);
    if (ctxt->path != NULL) xmlFree(ctxt->path);
    if (ctxt->out != NULL) xmlFree(ctxt->out);
    if (ctxt->in != NULL) xmlFree(ctxt->in);
    if (ctxt->contentType != NULL) xmlFree(ctxt->contentType);
535
    if (ctxt->encoding != NULL) xmlFree(ctxt->encoding);
536
    if (ctxt->mimeType != NULL) xmlFree(ctxt->mimeType);
537 538 539 540 541 542 543 544 545 546 547 548 549
    if (ctxt->location != NULL) xmlFree(ctxt->location);
    if (ctxt->authHeader != NULL) xmlFree(ctxt->authHeader);
    ctxt->state = XML_NANO_HTTP_NONE;
    if (ctxt->fd >= 0) closesocket(ctxt->fd);
    ctxt->fd = -1;
    xmlFree(ctxt);
}

/**
 * xmlNanoHTTPSend:
 * @ctxt:  an HTTP context
 *
 * Send the input needed to initiate the processing on the server side
550
 * Returns number of bytes sent or -1 on error.
551 552
 */

553 554 555 556 557 558 559 560 561
static int
xmlNanoHTTPSend(xmlNanoHTTPCtxtPtr ctxt, const char * xmt_ptr, int outlen) {

    int 	total_sent = 0;

    if ( (ctxt->state & XML_NANO_HTTP_WRITE) && (xmt_ptr != NULL ) ) {
        while (total_sent < outlen) {
            int nsent = send(ctxt->fd, xmt_ptr + total_sent,
                                      outlen - total_sent, 0);
562 563
            if (nsent>0)
                total_sent += nsent;
564
	    else if ( ( nsent == -1 ) && 
565
#if defined(EAGAIN) && EAGAIN != EWOULDBLOCK
566
	    	      ( socket_errno( ) != EAGAIN ) &&
567
#endif
568
		        ( socket_errno( ) != EWOULDBLOCK ) ) {
569
		__xmlIOErr(XML_FROM_HTTP, 0, "send failed\n");
570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590
		if ( total_sent == 0 )
		    total_sent = -1;
		break;
	    }
	    else {
	        /*
		**  No data sent
		**  Since non-blocking sockets are used, wait for 
		**  socket to be writable or default timeout prior
		**  to retrying.
		*/

		struct timeval	tv;
		fd_set		wfd;

		tv.tv_sec = timeout;
		tv.tv_usec = 0;
		FD_ZERO( &wfd );
		FD_SET( ctxt->fd, &wfd );
		(void)select( ctxt->fd + 1, NULL, &wfd, NULL, &tv );
	    }
591
	}
592
    }
593 594

    return total_sent;
595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614
}

/**
 * xmlNanoHTTPRecv:
 * @ctxt:  an HTTP context
 *
 * Read information coming from the HTTP connection.
 * This is a blocking call (but it blocks in select(), not read()).
 *
 * Returns the number of byte read or -1 in case of error.
 */

static int
xmlNanoHTTPRecv(xmlNanoHTTPCtxtPtr ctxt) {
    fd_set rfd;
    struct timeval tv;


    while (ctxt->state & XML_NANO_HTTP_READ) {
	if (ctxt->in == NULL) {
615
	    ctxt->in = (char *) xmlMallocAtomic(65000 * sizeof(char));
616
	    if (ctxt->in == NULL) {
617
		xmlHTTPErrMemory("allocating input");
618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636
	        ctxt->last = -1;
		return(-1);
	    }
	    ctxt->inlen = 65000;
	    ctxt->inptr = ctxt->content = ctxt->inrptr = ctxt->in;
	}
	if (ctxt->inrptr > ctxt->in + XML_NANO_HTTP_CHUNK) {
	    int delta = ctxt->inrptr - ctxt->in;
	    int len = ctxt->inptr - ctxt->inrptr;
	    
	    memmove(ctxt->in, ctxt->inrptr, len);
	    ctxt->inrptr -= delta;
	    ctxt->content -= delta;
	    ctxt->inptr -= delta;
	}
        if ((ctxt->in + ctxt->inlen) < (ctxt->inptr + XML_NANO_HTTP_CHUNK)) {
	    int d_inptr = ctxt->inptr - ctxt->in;
	    int d_content = ctxt->content - ctxt->in;
	    int d_inrptr = ctxt->inrptr - ctxt->in;
637
	    char *	tmp_ptr = ctxt->in;
638 639

	    ctxt->inlen *= 2;
640
            ctxt->in = (char *) xmlRealloc(tmp_ptr, ctxt->inlen);
641
	    if (ctxt->in == NULL) {
642
		xmlHTTPErrMemory("allocating input buffer");
643
		xmlFree( tmp_ptr );
644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666
	        ctxt->last = -1;
		return(-1);
	    }
            ctxt->inptr = ctxt->in + d_inptr;
            ctxt->content = ctxt->in + d_content;
            ctxt->inrptr = ctxt->in + d_inrptr;
	}
	ctxt->last = recv(ctxt->fd, ctxt->inptr, XML_NANO_HTTP_CHUNK, 0);
	if (ctxt->last > 0) {
	    ctxt->inptr += ctxt->last;
	    return(ctxt->last);
	}
	if (ctxt->last == 0) {
	    return(0);
	}
	if (ctxt->last == -1) {
	    switch (socket_errno()) {
		case EINPROGRESS:
		case EWOULDBLOCK:
#if defined(EAGAIN) && EAGAIN != EWOULDBLOCK
		case EAGAIN:
#endif
		    break;
667 668 669 670 671

		case ECONNRESET:
		case ESHUTDOWN:
		    return ( 0 );

672
		default:
673
		    __xmlIOErr(XML_FROM_HTTP, 0, "recv failed\n");
674
		    return(-1);
675 676 677 678 679 680 681 682
	    }
	}

	tv.tv_sec = timeout;
	tv.tv_usec = 0;
	FD_ZERO(&rfd);
	FD_SET(ctxt->fd, &rfd);
	
683 684 685 686 687
	if ( (select(ctxt->fd+1, &rfd, NULL, NULL, &tv)<1)
#if defined(EINTR)
		&& (errno != EINTR)
#endif
	)
688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707
		return(0);
    }
    return(0);
}

/**
 * xmlNanoHTTPReadLine:
 * @ctxt:  an HTTP context
 *
 * Read one line in the HTTP server output, usually for extracting
 * the HTTP protocol informations from the answer header.
 *
 * Returns a newly allocated string with a copy of the line, or NULL
 *         which indicate the end of the input.
 */

static char *
xmlNanoHTTPReadLine(xmlNanoHTTPCtxtPtr ctxt) {
    char buf[4096];
    char *bp = buf;
708
    int	rc;
709 710 711
    
    while (bp - buf < 4095) {
	if (ctxt->inrptr == ctxt->inptr) {
712
	    if ( (rc = xmlNanoHTTPRecv(ctxt)) == 0) {
713 714 715 716 717 718
		if (bp == buf)
		    return(NULL);
		else
		    *bp = 0;
		return(xmlMemStrdup(buf));
	    }
719 720 721
	    else if ( rc == -1 ) {
	        return ( NULL );
	    }
722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743
	}
	*bp = *ctxt->inrptr++;
	if (*bp == '\n') {
	    *bp = 0;
	    return(xmlMemStrdup(buf));
	}
	if (*bp != '\r')
	    bp++;
    }
    buf[4095] = 0;
    return(xmlMemStrdup(buf));
}


/**
 * xmlNanoHTTPScanAnswer:
 * @ctxt:  an HTTP context
 * @line:  an HTTP header line
 *
 * Try to extract useful informations from the server answer.
 * We currently parse and process:
 *  - The HTTP revision/ return code
744
 *  - The Content-Type, Mime-Type and charset used
745
 *  - The Location for redirect processing.
746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787
 *
 * Returns -1 in case of failure, the file descriptor number otherwise
 */

static void
xmlNanoHTTPScanAnswer(xmlNanoHTTPCtxtPtr ctxt, const char *line) {
    const char *cur = line;

    if (line == NULL) return;

    if (!strncmp(line, "HTTP/", 5)) {
        int version = 0;
	int ret = 0;

	cur += 5;
	while ((*cur >= '0') && (*cur <= '9')) {
	    version *= 10;
	    version += *cur - '0';
	    cur++;
	}
	if (*cur == '.') {
	    cur++;
	    if ((*cur >= '0') && (*cur <= '9')) {
		version *= 10;
		version += *cur - '0';
		cur++;
	    }
	    while ((*cur >= '0') && (*cur <= '9'))
		cur++;
	} else
	    version *= 10;
	if ((*cur != ' ') && (*cur != '\t')) return;
	while ((*cur == ' ') || (*cur == '\t')) cur++;
	if ((*cur < '0') || (*cur > '9')) return;
	while ((*cur >= '0') && (*cur <= '9')) {
	    ret *= 10;
	    ret += *cur - '0';
	    cur++;
	}
	if ((*cur != 0) && (*cur != ' ') && (*cur != '\t')) return;
	ctxt->returnValue = ret;
    } else if (!xmlStrncasecmp(BAD_CAST line, BAD_CAST"Content-Type:", 13)) {
788
        const xmlChar *charset, *last, *mime;
789 790 791 792 793
        cur += 13;
	while ((*cur == ' ') || (*cur == '\t')) cur++;
	if (ctxt->contentType != NULL)
	    xmlFree(ctxt->contentType);
	ctxt->contentType = xmlMemStrdup(cur);
794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812
	mime = (const xmlChar *) cur;
	last = mime;
	while ((*last != 0) && (*last != ' ') && (*last != '\t') &&
	       (*last != ';') && (*last != ','))
	    last++;
	if (ctxt->mimeType != NULL)
	    xmlFree(ctxt->mimeType);
	ctxt->mimeType = (char *) xmlStrndup(mime, last - mime);
	charset = xmlStrstr(BAD_CAST ctxt->contentType, BAD_CAST "charset=");
	if (charset != NULL) {
	    charset += 8;
	    last = charset;
	    while ((*last != 0) && (*last != ' ') && (*last != '\t') &&
	           (*last != ';') && (*last != ','))
		last++;
	    if (ctxt->encoding != NULL)
	        xmlFree(ctxt->encoding);
	    ctxt->encoding = (char *) xmlStrndup(charset, last - charset);
	}
813
    } else if (!xmlStrncasecmp(BAD_CAST line, BAD_CAST"ContentType:", 12)) {
814
        const xmlChar *charset, *last, *mime;
815 816 817 818
        cur += 12;
	if (ctxt->contentType != NULL) return;
	while ((*cur == ' ') || (*cur == '\t')) cur++;
	ctxt->contentType = xmlMemStrdup(cur);
819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837
	mime = (const xmlChar *) cur;
	last = mime;
	while ((*last != 0) && (*last != ' ') && (*last != '\t') &&
	       (*last != ';') && (*last != ','))
	    last++;
	if (ctxt->mimeType != NULL)
	    xmlFree(ctxt->mimeType);
	ctxt->mimeType = (char *) xmlStrndup(mime, last - mime);
	charset = xmlStrstr(BAD_CAST ctxt->contentType, BAD_CAST "charset=");
	if (charset != NULL) {
	    charset += 8;
	    last = charset;
	    while ((*last != 0) && (*last != ' ') && (*last != '\t') &&
	           (*last != ';') && (*last != ','))
		last++;
	    if (ctxt->encoding != NULL)
	        xmlFree(ctxt->encoding);
	    ctxt->encoding = (char *) xmlStrndup(charset, last - charset);
	}
838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855
    } else if (!xmlStrncasecmp(BAD_CAST line, BAD_CAST"Location:", 9)) {
        cur += 9;
	while ((*cur == ' ') || (*cur == '\t')) cur++;
	if (ctxt->location != NULL)
	    xmlFree(ctxt->location);
	ctxt->location = xmlMemStrdup(cur);
    } else if (!xmlStrncasecmp(BAD_CAST line, BAD_CAST"WWW-Authenticate:", 17)) {
        cur += 17;
	while ((*cur == ' ') || (*cur == '\t')) cur++;
	if (ctxt->authHeader != NULL)
	    xmlFree(ctxt->authHeader);
	ctxt->authHeader = xmlMemStrdup(cur);
    } else if (!xmlStrncasecmp(BAD_CAST line, BAD_CAST"Proxy-Authenticate:", 19)) {
        cur += 19;
	while ((*cur == ' ') || (*cur == '\t')) cur++;
	if (ctxt->authHeader != NULL)
	    xmlFree(ctxt->authHeader);
	ctxt->authHeader = xmlMemStrdup(cur);
856 857 858
    } else if ( !xmlStrncasecmp( BAD_CAST line, BAD_CAST"Content-Length:", 15) ) {
	cur += 15;
	ctxt->ContentLength = strtol( cur, NULL, 10 );
859 860 861 862 863
    }
}

/**
 * xmlNanoHTTPConnectAttempt:
864
 * @addr:  a socket address structure
865 866 867 868 869 870 871 872 873
 *
 * Attempt a connection to the given IP:port endpoint. It forces
 * non-blocking semantic on the socket, and allow 60 seconds for
 * the host to answer.
 *
 * Returns -1 in case of failure, the file descriptor number otherwise
 */

static int
874
xmlNanoHTTPConnectAttempt(struct sockaddr *addr)
875 876
{
    fd_set wfd;
877 878 879
#ifdef _WINSOCKAPI_
    fd_set xfd;
#endif
880 881
    struct timeval tv;
    int status;
882 883
    int addrlen;
    SOCKET s;
884
    
885 886 887 888 889 890 891 892 893 894 895
#ifdef SUPPORT_IP6
    if (addr->sa_family == AF_INET6) {
	s = socket (PF_INET6, SOCK_STREAM, IPPROTO_TCP);
	addrlen = sizeof (struct sockaddr_in6);
    }
    else
#endif
    {
	s = socket (PF_INET, SOCK_STREAM, IPPROTO_TCP);
	addrlen = sizeof (struct sockaddr_in);
    }
896 897 898 899
    if (s==-1) {
#ifdef DEBUG_HTTP
	perror("socket");
#endif
900
	__xmlIOErr(XML_FROM_HTTP, 0, "socket failed\n");
901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916
	return(-1);
    }
    
#ifdef _WINSOCKAPI_
    {
	u_long one = 1;

	status = ioctlsocket(s, FIONBIO, &one) == SOCKET_ERROR ? -1 : 0;
    }
#else /* _WINSOCKAPI_ */
#if defined(VMS)
    {
	int enable = 1;
	status = ioctl(s, FIONBIO, &enable);
    }
#else /* VMS */
917 918 919 920 921 922
#if defined(__BEOS__)
	{
		bool noblock = true;
		status = setsockopt(s, SOL_SOCKET, SO_NONBLOCK, &noblock, sizeof(noblock));
	}
#else /* __BEOS__ */
923 924 925 926 927 928 929 930 931 932 933 934 935 936
    if ((status = fcntl(s, F_GETFL, 0)) != -1) {
#ifdef O_NONBLOCK
	status |= O_NONBLOCK;
#else /* O_NONBLOCK */
#ifdef F_NDELAY
	status |= F_NDELAY;
#endif /* F_NDELAY */
#endif /* !O_NONBLOCK */
	status = fcntl(s, F_SETFL, status);
    }
    if (status < 0) {
#ifdef DEBUG_HTTP
	perror("nonblocking");
#endif
937
	__xmlIOErr(XML_FROM_HTTP, 0, "error setting non-blocking IO\n");
938 939 940
	closesocket(s);
	return(-1);
    }
941
#endif /* !__BEOS__ */
942 943 944
#endif /* !VMS */
#endif /* !_WINSOCKAPI_ */

945
    if (connect (s, addr, addrlen) == -1) {
946 947 948 949 950
	switch (socket_errno()) {
	    case EINPROGRESS:
	    case EWOULDBLOCK:
		break;
	    default:
951
		__xmlIOErr(XML_FROM_HTTP, 0, "error connecting to HTTP server");
952 953 954 955 956 957 958 959 960 961
		closesocket(s);
		return(-1);
	}
    }	
    
    tv.tv_sec = timeout;
    tv.tv_usec = 0;
    
    FD_ZERO(&wfd);
    FD_SET(s, &wfd);
962 963 964 965

#ifdef _WINSOCKAPI_    
    FD_ZERO(&xfd);
    FD_SET(s, &xfd);
966
    
967 968
    switch(select(s+1, NULL, &wfd, &xfd, &tv))
#else
969
    switch(select(s+1, NULL, &wfd, NULL, &tv))
970
#endif
971 972 973
    {
	case 0:
	    /* Time out */
974
	    __xmlIOErr(XML_FROM_HTTP, 0, "Connect attempt timed out");
975 976 977 978
	    closesocket(s);
	    return(-1);
	case -1:
	    /* Ermm.. ?? */
979
	    __xmlIOErr(XML_FROM_HTTP, 0, "Connect failed");
980 981 982 983
	    closesocket(s);
	    return(-1);
    }

984 985 986 987 988
    if ( FD_ISSET(s, &wfd)
#ifdef _WINSOCKAPI_
                           || FD_ISSET(s, &xfd)
#endif
                                                ) {
989 990
	SOCKLEN_T len;
	len = sizeof(status);
991
#ifdef SO_ERROR
992 993
	if (getsockopt(s, SOL_SOCKET, SO_ERROR, (char*)&status, &len) < 0 ) {
	    /* Solaris error code */
994
	    __xmlIOErr(XML_FROM_HTTP, 0, "getsockopt failed\n");
995 996
	    return (-1);
	}
997
#endif
998
	if ( status ) {
999
	    __xmlIOErr(XML_FROM_HTTP, 0, "Error connecting to remote host");
1000 1001 1002 1003 1004 1005
	    closesocket(s);
	    errno = status;
	    return (-1);
	}
    } else {
	/* pbm */
1006
	__xmlIOErr(XML_FROM_HTTP, 0, "select failed\n");
1007
	closesocket(s);
1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028
	return (-1);
    }
    
    return(s);
}
 
/**
 * xmlNanoHTTPConnectHost:
 * @host:  the host name
 * @port:  the port number
 *
 * Attempt a connection to the given host:port endpoint. It tries
 * the multiple IP provided by the DNS if available.
 *
 * Returns -1 in case of failure, the file descriptor number otherwise
 */

static int
xmlNanoHTTPConnectHost(const char *host, int port)
{
    struct hostent *h;
1029
    struct sockaddr *addr = NULL;
1030
    struct in_addr ia;
1031
    struct sockaddr_in sockin;
1032

1033 1034
#ifdef SUPPORT_IP6
    struct in6_addr ia6;
1035
    struct sockaddr_in6 sockin6;
1036 1037 1038
#endif
    int i;
    int s;
1039

1040 1041 1042 1043 1044
    memset (&sockin, 0, sizeof(sockin));
#ifdef SUPPORT_IP6
    memset (&sockin6, 0, sizeof(sockin6));
    if (have_ipv6 ())
#if !defined(HAVE_GETADDRINFO) && defined(RES_USE_INET6)
1045
    {
1046 1047 1048 1049 1050
	if (!(_res.options & RES_INIT))
	    res_init();
	_res.options |= RES_USE_INET6;
    }
#elif defined(HAVE_GETADDRINFO)
1051
    {
1052 1053 1054 1055 1056 1057 1058 1059 1060
	int status;
	struct addrinfo hints, *res, *result;

	result = NULL;
	memset (&hints, 0,sizeof(hints));
	hints.ai_socktype = SOCK_STREAM;

	status = getaddrinfo (host, NULL, &hints, &result);
	if (status) {
1061
	    __xmlIOErr(XML_FROM_HTTP, 0, "getaddrinfo failed\n");
1062 1063 1064 1065
	    return (-1);
	}

	for (res = result; res; res = res->ai_next) {
1066
	    if (res->ai_family == AF_INET || res->ai_family == AF_INET6) {
1067 1068 1069 1070 1071
		if (res->ai_family == AF_INET6) {
		    memcpy (&sockin6, res->ai_addr, res->ai_addrlen);
		    sockin6.sin6_port = htons (port);
		    addr = (struct sockaddr *)&sockin6;
		}
1072
		else {
1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084
		    memcpy (&sockin, res->ai_addr, res->ai_addrlen);
		    sockin.sin_port = htons (port);
		    addr = (struct sockaddr *)&sockin;
		}

		s = xmlNanoHTTPConnectAttempt (addr);
		if (s != -1) {
		    freeaddrinfo (result);
		    return (s);
		}
	    }
	}
1085 1086 1087
	if (result)
	    freeaddrinfo (result);
	return (-1);
1088
    } else
1089
#endif
1090 1091 1092 1093
#endif
    {   
	h = gethostbyname (host);
	if (h == NULL) {
1094 1095 1096 1097 1098 1099 1100

/*
 * Okay, I got fed up by the non-portability of this error message
 * extraction code. it work on Linux, if it work on your platform
 * and one want to enable it, send me the defined(foobar) needed
 */
#if defined(HAVE_NETDB_H) && defined(HOST_NOT_FOUND) && defined(linux)
1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126
	    const char *h_err_txt = "";

	    switch (h_errno) {
		case HOST_NOT_FOUND:
		    h_err_txt = "Authoritive host not found";
		    break;

		case TRY_AGAIN:
		    h_err_txt =
			"Non-authoritive host not found or server failure.";
		    break;

		case NO_RECOVERY:
		    h_err_txt =
			"Non-recoverable errors:  FORMERR, REFUSED, or NOTIMP.";
		    break;

		case NO_ADDRESS:
		    h_err_txt =
			"Valid name, no data record of requested type.";
		    break;

		default:
		    h_err_txt = "No error text defined.";
		    break;
	    }
1127
	    __xmlIOErr(XML_FROM_HTTP, 0, h_err_txt);
1128
#else
1129
	    __xmlIOErr(XML_FROM_HTTP, 0, "Failed to resolve host");
1130
#endif
1131 1132
	    return (-1);
	}
1133

1134 1135 1136 1137 1138 1139 1140 1141
	for (i = 0; h->h_addr_list[i]; i++) {
	    if (h->h_addrtype == AF_INET) {
		/* A records (IPv4) */
		memcpy (&ia, h->h_addr_list[i], h->h_length);
		sockin.sin_family = h->h_addrtype;
		sockin.sin_addr = ia;
		sockin.sin_port = htons (port);
		addr = (struct sockaddr *) &sockin;
1142
#ifdef SUPPORT_IP6
1143 1144 1145 1146 1147 1148 1149 1150 1151 1152
	    } else if (have_ipv6 () && (h->h_addrtype == AF_INET6)) {
		/* AAAA records (IPv6) */
		memcpy (&ia6, h->h_addr_list[i], h->h_length);
		sockin6.sin6_family = h->h_addrtype;
		sockin6.sin6_addr = ia6;
		sockin6.sin6_port = htons (port);
		addr = (struct sockaddr *) &sockin6;
#endif
	    } else
		break;              /* for */
1153

1154 1155 1156 1157 1158
	    s = xmlNanoHTTPConnectAttempt (addr);
	    if (s != -1)
		return (s);
	}
    }
1159 1160
#ifdef DEBUG_HTTP
    xmlGenericError(xmlGenericErrorContext,
1161 1162
                    "xmlNanoHTTPConnectHost:  unable to connect to '%s'.\n",
                    host);
1163
#endif
1164
    return (-1);
1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183
}


/**
 * xmlNanoHTTPOpen:
 * @URL:  The URL to load
 * @contentType:  if available the Content-Type information will be
 *                returned at that location
 *
 * This function try to open a connection to the indicated resource
 * via HTTP GET.
 *
 * Returns NULL in case of failure, otherwise a request handler.
 *     The contentType, if provided must be freed by the caller
 */

void*
xmlNanoHTTPOpen(const char *URL, char **contentType) {
    if (contentType != NULL) *contentType = NULL;
1184
    return(xmlNanoHTTPMethod(URL, NULL, NULL, contentType, NULL, 0));
1185 1186 1187 1188 1189 1190 1191
}

/**
 * xmlNanoHTTPOpenRedir:
 * @URL:  The URL to load
 * @contentType:  if available the Content-Type information will be
 *                returned at that location
1192
 * @redir: if available the redirected URL will be returned
1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204
 *
 * This function try to open a connection to the indicated resource
 * via HTTP GET.
 *
 * Returns NULL in case of failure, otherwise a request handler.
 *     The contentType, if provided must be freed by the caller
 */

void*
xmlNanoHTTPOpenRedir(const char *URL, char **contentType, char **redir) {
    if (contentType != NULL) *contentType = NULL;
    if (redir != NULL) *redir = NULL;
1205
    return(xmlNanoHTTPMethodRedir(URL, NULL, NULL, contentType, redir, NULL,0));
1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228
}

/**
 * xmlNanoHTTPRead:
 * @ctx:  the HTTP context
 * @dest:  a buffer
 * @len:  the buffer length
 *
 * This function tries to read @len bytes from the existing HTTP connection
 * and saves them in @dest. This is a blocking call.
 *
 * Returns the number of byte read. 0 is an indication of an end of connection.
 *         -1 indicates a parameter error.
 */
int
xmlNanoHTTPRead(void *ctx, void *dest, int len) {
    xmlNanoHTTPCtxtPtr ctxt = (xmlNanoHTTPCtxtPtr) ctx;

    if (ctx == NULL) return(-1);
    if (dest == NULL) return(-1);
    if (len <= 0) return(0);

    while (ctxt->inptr - ctxt->inrptr < len) {
1229
        if (xmlNanoHTTPRecv(ctxt) <= 0) break;
1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254
    }
    if (ctxt->inptr - ctxt->inrptr < len)
        len = ctxt->inptr - ctxt->inrptr;
    memcpy(dest, ctxt->inrptr, len);
    ctxt->inrptr += len;
    return(len);
}

/**
 * xmlNanoHTTPClose:
 * @ctx:  the HTTP context
 *
 * This function closes an HTTP context, it ends up the connection and
 * free all data related to it.
 */
void
xmlNanoHTTPClose(void *ctx) {
    xmlNanoHTTPCtxtPtr ctxt = (xmlNanoHTTPCtxtPtr) ctx;

    if (ctx == NULL) return;

    xmlNanoHTTPFreeCtxt(ctxt);
}

/**
1255
 * xmlNanoHTTPMethodRedir:
1256 1257 1258 1259
 * @URL:  The URL to load
 * @method:  the HTTP method to use
 * @input:  the input string if any
 * @contentType:  the Content-Type information IN and OUT
1260
 * @redir:  the redirected URL OUT
1261
 * @headers:  the extra headers
1262
 * @ilen:  input length
1263 1264 1265 1266 1267 1268
 *
 * This function try to open a connection to the indicated resource
 * via HTTP using the given @method, adding the given extra headers
 * and the input buffer for the request content.
 *
 * Returns NULL in case of failure, otherwise a request handler.
1269
 *     The contentType, or redir, if provided must be freed by the caller
1270 1271 1272
 */

void*
1273
xmlNanoHTTPMethodRedir(const char *URL, const char *method, const char *input,
1274 1275
                  char **contentType, char **redir,
		  const char *headers, int ilen ) {
1276 1277
    xmlNanoHTTPCtxtPtr ctxt;
    char *bp, *p;
1278
    int blen, ret;
1279 1280 1281
    int head;
    int nbRedirects = 0;
    char *redirURL = NULL;
William M. Brack's avatar
William M. Brack committed
1282 1283 1284
#ifdef DEBUG_HTTP
    int xmt_bytes;
#endif
1285 1286 1287 1288 1289 1290 1291 1292 1293 1294
    
    if (URL == NULL) return(NULL);
    if (method == NULL) method = "GET";
    xmlNanoHTTPInit();

retry:
    if (redirURL == NULL)
	ctxt = xmlNanoHTTPNewCtxt(URL);
    else {
	ctxt = xmlNanoHTTPNewCtxt(redirURL);
1295
	ctxt->location = xmlMemStrdup(redirURL);
1296 1297
    }

1298 1299 1300 1301
    if ( ctxt == NULL ) {
	return ( NULL );
    }

1302
    if ((ctxt->protocol == NULL) || (strcmp(ctxt->protocol, "http"))) {
1303
	__xmlIOErr(XML_FROM_HTTP, XML_HTTP_URL_SYNTAX, "Not a valid HTTP URI");
1304 1305 1306 1307 1308
        xmlNanoHTTPFreeCtxt(ctxt);
	if (redirURL != NULL) xmlFree(redirURL);
        return(NULL);
    }
    if (ctxt->hostname == NULL) {
1309 1310
	__xmlIOErr(XML_FROM_HTTP, XML_HTTP_UNKNOWN_HOST,
	           "Failed to identify host in URI");
1311
        xmlNanoHTTPFreeCtxt(ctxt);
1312
	if (redirURL != NULL) xmlFree(redirURL);
1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324
        return(NULL);
    }
    if (proxy) {
	blen = strlen(ctxt->hostname) * 2 + 16;
	ret = xmlNanoHTTPConnectHost(proxy, proxyPort);
    }
    else {
	blen = strlen(ctxt->hostname);
	ret = xmlNanoHTTPConnectHost(ctxt->hostname, ctxt->port);
    }
    if (ret < 0) {
        xmlNanoHTTPFreeCtxt(ctxt);
1325
	if (redirURL != NULL) xmlFree(redirURL);
1326 1327 1328 1329
        return(NULL);
    }
    ctxt->fd = ret;

1330
    if (input == NULL)
1331
	ilen = 0;
1332 1333 1334
    else
	blen += 36;

1335
    if (headers != NULL)
1336
	blen += strlen(headers) + 2;
1337 1338
    if (contentType && *contentType)
	blen += strlen(*contentType) + 16;
1339
    blen += strlen(method) + strlen(ctxt->path) + 24;
1340
    bp = (char*)xmlMallocAtomic(blen);
1341 1342
    if ( bp == NULL ) {
        xmlNanoHTTPFreeCtxt( ctxt );
1343
	xmlHTTPErrMemory("allocating header buffer");
1344 1345 1346 1347 1348
	return ( NULL );
    }

    p = bp;

1349 1350
    if (proxy) {
	if (ctxt->port != 80) {
1351 1352
	    p += snprintf( p, blen - (p - bp), "%s http://%s:%d%s", 
			method, ctxt->hostname,
1353
		 	ctxt->port, ctxt->path );
1354
	}
1355 1356
	else 
	    p += snprintf( p, blen - (p - bp), "%s http://%s%s", method,
1357
	    		ctxt->hostname, ctxt->path);
1358 1359
    }
    else
1360
	p += snprintf( p, blen - (p - bp), "%s %s", method, ctxt->path);
1361

1362 1363
    p += snprintf( p, blen - (p - bp), " HTTP/1.0\r\nHost: %s\r\n", 
		    ctxt->hostname);
1364 1365

    if (contentType != NULL && *contentType) 
1366
	p += snprintf(p, blen - (p - bp), "Content-Type: %s\r\n", *contentType);
1367 1368

    if (headers != NULL)
1369
	p += snprintf( p, blen - (p - bp), "%s", headers );
1370

1371
    if (input != NULL)
1372
	snprintf(p, blen - (p - bp), "Content-Length: %d\r\n\r\n", ilen );
1373
    else
1374
	snprintf(p, blen - (p - bp), "\r\n");
1375

1376 1377 1378 1379 1380 1381 1382 1383 1384
#ifdef DEBUG_HTTP
    xmlGenericError(xmlGenericErrorContext,
	    "-> %s%s", proxy? "(Proxy) " : "", bp);
    if ((blen -= strlen(bp)+1) < 0)
	xmlGenericError(xmlGenericErrorContext,
		"ERROR: overflowed buffer by %d bytes\n", -blen);
#endif
    ctxt->outptr = ctxt->out = bp;
    ctxt->state = XML_NANO_HTTP_WRITE;
1385 1386
    blen = strlen( ctxt->out );
#ifdef DEBUG_HTTP
William M. Brack's avatar
William M. Brack committed
1387
    xmt_bytes = xmlNanoHTTPSend(ctxt, ctxt->out, blen );
1388 1389 1390 1391 1392 1393
    if ( xmt_bytes != blen )
        xmlGenericError( xmlGenericErrorContext,
			"xmlNanoHTTPMethodRedir:  Only %d of %d %s %s\n",
			xmt_bytes, blen,
			"bytes of HTTP headers sent to host",
			ctxt->hostname );
William M. Brack's avatar
William M. Brack committed
1394 1395
#else
    xmlNanoHTTPSend(ctxt, ctxt->out, blen );
1396 1397 1398
#endif

    if ( input != NULL ) {
William M. Brack's avatar
William M. Brack committed
1399
#ifdef DEBUG_HTTP
1400 1401 1402 1403 1404 1405 1406 1407
        xmt_bytes = xmlNanoHTTPSend( ctxt, input, ilen );

	if ( xmt_bytes != ilen )
	    xmlGenericError( xmlGenericErrorContext,
	    		"xmlNanoHTTPMethodRedir:  Only %d of %d %s %s\n",
			xmt_bytes, ilen,
			"bytes of HTTP content sent to host",
			ctxt->hostname );
William M. Brack's avatar
William M. Brack committed
1408 1409
#else
	xmlNanoHTTPSend( ctxt, input, ilen );
1410 1411 1412
#endif
    }

1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436
    ctxt->state = XML_NANO_HTTP_READ;
    head = 1;

    while ((p = xmlNanoHTTPReadLine(ctxt)) != NULL) {
        if (head && (*p == 0)) {
	    head = 0;
	    ctxt->content = ctxt->inrptr;
	    xmlFree(p);
	    break;
	}
	xmlNanoHTTPScanAnswer(ctxt, p);

#ifdef DEBUG_HTTP
	xmlGenericError(xmlGenericErrorContext, "<- %s\n", p);
#endif
        xmlFree(p);
    }

    if ((ctxt->location != NULL) && (ctxt->returnValue >= 300) &&
        (ctxt->returnValue < 400)) {
#ifdef DEBUG_HTTP
	xmlGenericError(xmlGenericErrorContext,
		"\nRedirect to: %s\n", ctxt->location);
#endif
1437
	while ( xmlNanoHTTPRecv(ctxt) > 0 ) ;
1438 1439
        if (nbRedirects < XML_NANO_HTTP_MAX_REDIR) {
	    nbRedirects++;
1440 1441
	    if (redirURL != NULL)
		xmlFree(redirURL);
1442 1443 1444 1445 1446
	    redirURL = xmlMemStrdup(ctxt->location);
	    xmlNanoHTTPFreeCtxt(ctxt);
	    goto retry;
	}
	xmlNanoHTTPFreeCtxt(ctxt);
1447
	if (redirURL != NULL) xmlFree(redirURL);
1448 1449
#ifdef DEBUG_HTTP
	xmlGenericError(xmlGenericErrorContext,
1450
		"xmlNanoHTTPMethodRedir: Too many redirects, aborting ...\n");
1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461
#endif
	return(NULL);
    }

    if (contentType != NULL) {
	if (ctxt->contentType != NULL)
	    *contentType = xmlMemStrdup(ctxt->contentType);
	else
	    *contentType = NULL;
    }

1462 1463 1464 1465 1466 1467 1468 1469 1470
    if ((redir != NULL) && (redirURL != NULL)) {
	*redir = redirURL;
    } else {
	if (redirURL != NULL)
	    xmlFree(redirURL);
	if (redir != NULL)
	    *redir = NULL;
    }

1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484
#ifdef DEBUG_HTTP
    if (ctxt->contentType != NULL)
	xmlGenericError(xmlGenericErrorContext,
		"\nCode %d, content-type '%s'\n\n",
	       ctxt->returnValue, ctxt->contentType);
    else
	xmlGenericError(xmlGenericErrorContext,
		"\nCode %d, no content-type\n\n",
	       ctxt->returnValue);
#endif

    return((void *) ctxt);
}

1485 1486 1487 1488 1489 1490 1491
/**
 * xmlNanoHTTPMethod:
 * @URL:  The URL to load
 * @method:  the HTTP method to use
 * @input:  the input string if any
 * @contentType:  the Content-Type information IN and OUT
 * @headers:  the extra headers
1492
 * @ilen:  input length
1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503
 *
 * This function try to open a connection to the indicated resource
 * via HTTP using the given @method, adding the given extra headers
 * and the input buffer for the request content.
 *
 * Returns NULL in case of failure, otherwise a request handler.
 *     The contentType, if provided must be freed by the caller
 */

void*
xmlNanoHTTPMethod(const char *URL, const char *method, const char *input,
1504
                  char **contentType, const char *headers, int ilen) {
1505
    return(xmlNanoHTTPMethodRedir(URL, method, input, contentType,
1506
		                  NULL, headers, ilen));
1507 1508
}

1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523
/**
 * xmlNanoHTTPFetch:
 * @URL:  The URL to load
 * @filename:  the filename where the content should be saved
 * @contentType:  if available the Content-Type information will be
 *                returned at that location
 *
 * This function try to fetch the indicated resource via HTTP GET
 * and save it's content in the file.
 *
 * Returns -1 in case of failure, 0 incase of success. The contentType,
 *     if provided must be freed by the caller
 */
int
xmlNanoHTTPFetch(const char *URL, const char *filename, char **contentType) {
1524 1525
    void *ctxt = NULL;
    char *buf = NULL;
1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545
    int fd;
    int len;
    
    ctxt = xmlNanoHTTPOpen(URL, contentType);
    if (ctxt == NULL) return(-1);

    if (!strcmp(filename, "-")) 
        fd = 0;
    else {
        fd = open(filename, O_CREAT | O_WRONLY, 00644);
	if (fd < 0) {
	    xmlNanoHTTPClose(ctxt);
	    if ((contentType != NULL) && (*contentType != NULL)) {
	        xmlFree(*contentType);
		*contentType = NULL;
	    }
	    return(-1);
	}
    }

1546 1547
    xmlNanoHTTPFetchContent( ctxt, &buf, &len );
    if ( len > 0 ) {
1548 1549 1550 1551 1552 1553 1554 1555
	write(fd, buf, len);
    }

    xmlNanoHTTPClose(ctxt);
    close(fd);
    return(0);
}

1556
#ifdef LIBXML_OUTPUT_ENABLED
1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568
/**
 * xmlNanoHTTPSave:
 * @ctxt:  the HTTP context
 * @filename:  the filename where the content should be saved
 *
 * This function saves the output of the HTTP transaction to a file
 * It closes and free the context at the end
 *
 * Returns -1 in case of failure, 0 incase of success.
 */
int
xmlNanoHTTPSave(void *ctxt, const char *filename) {
1569
    char *buf = NULL;
1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584
    int fd;
    int len;
    
    if (ctxt == NULL) return(-1);

    if (!strcmp(filename, "-")) 
        fd = 0;
    else {
        fd = open(filename, O_CREAT | O_WRONLY);
	if (fd < 0) {
	    xmlNanoHTTPClose(ctxt);
	    return(-1);
	}
    }

1585 1586
    xmlNanoHTTPFetchContent( ctxt, &buf, &len );
    if ( len > 0 ) {
1587 1588 1589 1590 1591 1592
	write(fd, buf, len);
    }

    xmlNanoHTTPClose(ctxt);
    return(0);
}
1593
#endif /* LIBXML_OUTPUT_ENABLED */
1594 1595 1596 1597 1598

/**
 * xmlNanoHTTPReturnCode:
 * @ctx:  the HTTP context
 *
1599 1600
 * Get the latest HTTP return code received
 *
1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615
 * Returns the HTTP return code for the request.
 */
int
xmlNanoHTTPReturnCode(void *ctx) {
    xmlNanoHTTPCtxtPtr ctxt = (xmlNanoHTTPCtxtPtr) ctx;

    if (ctxt == NULL) return(-1);

    return(ctxt->returnValue);
}

/**
 * xmlNanoHTTPAuthHeader:
 * @ctx:  the HTTP context
 *
1616 1617
 * Get the authentication header of an HTTP context
 *
1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629
 * Returns the stashed value of the WWW-Authenticate or Proxy-Authenticate
 * header.
 */
const char *
xmlNanoHTTPAuthHeader(void *ctx) {
    xmlNanoHTTPCtxtPtr ctxt = (xmlNanoHTTPCtxtPtr) ctx;

    if (ctxt == NULL) return(NULL);

    return(ctxt->authHeader);
}

1630
/**
1631
 * xmlNanoHTTPContentLength:
1632 1633
 * @ctx:  the HTTP context
 *
1634 1635
 * Provides the specified content length from the HTTP header.
 *
1636 1637 1638 1639 1640 1641
 * Return the specified content length from the HTTP header.  Note that
 * a value of -1 indicates that the content length element was not included in
 * the response header.
 */
int
xmlNanoHTTPContentLength( void * ctx ) {
1642
    xmlNanoHTTPCtxtPtr	ctxt = (xmlNanoHTTPCtxtPtr)ctx;
1643 1644 1645 1646

    return ( ( ctxt == NULL ) ? -1 : ctxt->ContentLength );
}

1647 1648 1649 1650 1651 1652 1653 1654 1655 1656
/**
 * xmlNanoHTTPRedir:
 * @ctx:  the HTTP context
 *
 * Provides the specified redirection URL if available from the HTTP header.
 *
 * Return the specified redirection URL or NULL if not redirected.
 */
const char *
xmlNanoHTTPRedir( void * ctx ) {
1657
    xmlNanoHTTPCtxtPtr	ctxt = (xmlNanoHTTPCtxtPtr)ctx;
1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671

    return ( ( ctxt == NULL ) ? NULL : ctxt->location );
}

/**
 * xmlNanoHTTPEncoding:
 * @ctx:  the HTTP context
 *
 * Provides the specified encoding if specified in the HTTP headers.
 *
 * Return the specified encoding or NULL if not available
 */
const char *
xmlNanoHTTPEncoding( void * ctx ) {
1672
    xmlNanoHTTPCtxtPtr	ctxt = (xmlNanoHTTPCtxtPtr)ctx;
1673 1674 1675 1676

    return ( ( ctxt == NULL ) ? NULL : ctxt->encoding );
}

1677 1678 1679 1680 1681 1682 1683 1684 1685 1686
/**
 * xmlNanoHTTPMimeType:
 * @ctx:  the HTTP context
 *
 * Provides the specified Mime-Type if specified in the HTTP headers.
 *
 * Return the specified Mime-Type or NULL if not available
 */
const char *
xmlNanoHTTPMimeType( void * ctx ) {
1687
    xmlNanoHTTPCtxtPtr	ctxt = (xmlNanoHTTPCtxtPtr)ctx;
1688 1689 1690 1691

    return ( ( ctxt == NULL ) ? NULL : ctxt->mimeType );
}

1692
/**
1693
 * xmlNanoHTTPFetchContent:
1694 1695 1696 1697
 * @ctx:  the HTTP context
 * @ptr:  pointer to set to the content buffer.
 * @len:  integer pointer to hold the length of the content
 *
1698 1699
 * Check if all the content was read
 *
1700 1701 1702 1703 1704 1705
 * Returns 0 if all the content was read and available, returns
 * -1 if received content length was less than specified or an error 
 * occurred.
 */
int
xmlNanoHTTPFetchContent( void * ctx, char ** ptr, int * len ) {
1706
    xmlNanoHTTPCtxtPtr	ctxt = (xmlNanoHTTPCtxtPtr)ctx;
1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749

    int			rc = 0;
    int			cur_lgth;
    int			rcvd_lgth;
    int			dummy_int;
    char *		dummy_ptr = NULL;

    /*  Dummy up return input parameters if not provided  */

    if ( len == NULL )
        len = &dummy_int;

    if ( ptr == NULL )
        ptr = &dummy_ptr;

    /*  But can't work without the context pointer  */

    if ( ( ctxt == NULL ) || ( ctxt->content == NULL