xmlIO.c 103 KB
Newer Older
1 2 3 4 5
/*
 * xmlIO.c : implementation of the I/O interfaces used by the parser
 *
 * See Copyright for the status of this software.
 *
6
 * daniel@veillard.com
7 8 9 10
 *
 * 14 Nov 2000 ht - for VMS, truncated name of long functions to under 32 char
 */

11
#define IN_LIBXML
12
#include "libxml.h"
13 14

#include <string.h>
15
#include <stddef.h>
16
#ifdef HAVE_ERRNO_H
17
#include <errno.h>
18 19
#endif

20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

#ifdef HAVE_SYS_TYPES_H
#include <sys/types.h>
#endif
#ifdef HAVE_SYS_STAT_H
#include <sys/stat.h>
#endif
#ifdef HAVE_FCNTL_H
#include <fcntl.h>
#endif
#ifdef HAVE_UNISTD_H
#include <unistd.h>
#endif
#ifdef HAVE_STDLIB_H
#include <stdlib.h>
#endif
36
#ifdef LIBXML_ZLIB_ENABLED
37 38
#include <zlib.h>
#endif
39
#ifdef LIBXML_LZMA_ENABLED
40 41
#include <lzma.h>
#endif
42

43 44
#if defined(_WIN32) && !defined(__CYGWIN__)
#define WIN32_LEAN_AND_MEAN
45 46 47
#include <windows.h>
#endif

48 49 50 51
#if defined(_WIN32_WCE)
#include <winnls.h> /* for CP_UTF8 */
#endif

52 53 54 55 56 57 58 59
#ifndef S_ISDIR
#  ifdef _S_ISDIR
#    define S_ISDIR(x) _S_ISDIR(x)
#  elif defined(S_IFDIR)
#    ifdef S_IFMT
#      define S_ISDIR(m) (((m) & S_IFMT) == S_IFDIR)
#    elif defined(_S_IFMT)
#      define S_ISDIR(m) (((m) & _S_IFMT) == S_IFDIR)
60 61 62 63 64 65 66 67
#    endif
#  endif
#endif

#include <libxml/xmlmemory.h>
#include <libxml/parser.h>
#include <libxml/parserInternals.h>
#include <libxml/xmlIO.h>
68
#include <libxml/uri.h>
69 70 71
#include <libxml/nanohttp.h>
#include <libxml/nanoftp.h>
#include <libxml/xmlerror.h>
72 73 74
#ifdef LIBXML_CATALOG_ENABLED
#include <libxml/catalog.h>
#endif
75
#include <libxml/globals.h>
76

77 78 79
#include "buf.h"
#include "enc.h"

80
/* #define VERBOSE_FAILURE */
81
/* #define DEBUG_EXTERNAL_ENTITIES */
82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101
/* #define DEBUG_INPUT */

#ifdef DEBUG_INPUT
#define MINLEN 40
#else
#define MINLEN 4000
#endif

/*
 * Input I/O callback sets
 */
typedef struct _xmlInputCallback {
    xmlInputMatchCallback matchcallback;
    xmlInputOpenCallback opencallback;
    xmlInputReadCallback readcallback;
    xmlInputCloseCallback closecallback;
} xmlInputCallback;

#define MAX_INPUT_CALLBACK 15

102 103 104
static xmlInputCallback xmlInputCallbackTable[MAX_INPUT_CALLBACK];
static int xmlInputCallbackNr = 0;
static int xmlInputCallbackInitialized = 0;
105

106
#ifdef LIBXML_OUTPUT_ENABLED
107 108 109 110 111 112 113 114 115 116 117 118
/*
 * Output I/O callback sets
 */
typedef struct _xmlOutputCallback {
    xmlOutputMatchCallback matchcallback;
    xmlOutputOpenCallback opencallback;
    xmlOutputWriteCallback writecallback;
    xmlOutputCloseCallback closecallback;
} xmlOutputCallback;

#define MAX_OUTPUT_CALLBACK 15

119 120 121
static xmlOutputCallback xmlOutputCallbackTable[MAX_OUTPUT_CALLBACK];
static int xmlOutputCallbackNr = 0;
static int xmlOutputCallbackInitialized = 0;
122 123 124

xmlOutputBufferPtr
xmlAllocOutputBufferInternal(xmlCharEncodingHandlerPtr encoder);
125
#endif /* LIBXML_OUTPUT_ENABLED */
126

127 128
/************************************************************************
 *									*
129
 *		Tree memory error handler				*
130 131 132 133
 *									*
 ************************************************************************/

static const char *IOerr[] = {
Daniel Veillard's avatar
Daniel Veillard committed
134
    "Unknown IO error",         /* UNKNOWN */
135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176
    "Permission denied",	/* EACCES */
    "Resource temporarily unavailable",/* EAGAIN */
    "Bad file descriptor",	/* EBADF */
    "Bad message",		/* EBADMSG */
    "Resource busy",		/* EBUSY */
    "Operation canceled",	/* ECANCELED */
    "No child processes",	/* ECHILD */
    "Resource deadlock avoided",/* EDEADLK */
    "Domain error",		/* EDOM */
    "File exists",		/* EEXIST */
    "Bad address",		/* EFAULT */
    "File too large",		/* EFBIG */
    "Operation in progress",	/* EINPROGRESS */
    "Interrupted function call",/* EINTR */
    "Invalid argument",		/* EINVAL */
    "Input/output error",	/* EIO */
    "Is a directory",		/* EISDIR */
    "Too many open files",	/* EMFILE */
    "Too many links",		/* EMLINK */
    "Inappropriate message buffer length",/* EMSGSIZE */
    "Filename too long",	/* ENAMETOOLONG */
    "Too many open files in system",/* ENFILE */
    "No such device",		/* ENODEV */
    "No such file or directory",/* ENOENT */
    "Exec format error",	/* ENOEXEC */
    "No locks available",	/* ENOLCK */
    "Not enough space",		/* ENOMEM */
    "No space left on device",	/* ENOSPC */
    "Function not implemented",	/* ENOSYS */
    "Not a directory",		/* ENOTDIR */
    "Directory not empty",	/* ENOTEMPTY */
    "Not supported",		/* ENOTSUP */
    "Inappropriate I/O control operation",/* ENOTTY */
    "No such device or address",/* ENXIO */
    "Operation not permitted",	/* EPERM */
    "Broken pipe",		/* EPIPE */
    "Result too large",		/* ERANGE */
    "Read-only file system",	/* EROFS */
    "Invalid seek",		/* ESPIPE */
    "No such process",		/* ESRCH */
    "Operation timed out",	/* ETIMEDOUT */
    "Improper link",		/* EXDEV */
177
    "Attempt to load network entity %s", /* XML_IO_NETWORK_ATTEMPT */
178
    "encoder error",		/* XML_IO_ENCODER */
179 180 181 182 183 184 185
    "flush error",
    "write error",
    "no input",
    "buffer full",
    "loading error",
    "not a socket",		/* ENOTSOCK */
    "already connected",	/* EISCONN */
186
    "connection refused",	/* ECONNREFUSED */
187 188 189 190
    "unreachable network",	/* ENETUNREACH */
    "adddress in use",		/* EADDRINUSE */
    "already in use",		/* EALREADY */
    "unknown address familly",	/* EAFNOSUPPORT */
191 192
};

193
#if defined(_WIN32) || defined (__DJGPP__) && !defined (__CYGWIN__)
194 195 196 197 198 199 200 201 202
/**
 * __xmlIOWin32UTF8ToWChar:
 * @u8String:  uft-8 string
 *
 * Convert a string from utf-8 to wchar (WINDOWS ONLY!)
 */
static wchar_t *
__xmlIOWin32UTF8ToWChar(const char *u8String)
{
203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219
    wchar_t *wString = NULL;

    if (u8String) {
        int wLen =
            MultiByteToWideChar(CP_UTF8, MB_ERR_INVALID_CHARS, u8String,
                                -1, NULL, 0);
        if (wLen) {
            wString = xmlMalloc(wLen * sizeof(wchar_t));
            if (wString) {
                if (MultiByteToWideChar
                    (CP_UTF8, 0, u8String, -1, wString, wLen) == 0) {
                    xmlFree(wString);
                    wString = NULL;
                }
            }
        }
    }
220

221
    return wString;
222 223 224
}
#endif

225 226 227 228 229 230 231 232 233 234 235 236 237
/**
 * xmlIOErrMemory:
 * @extra:  extra informations
 *
 * Handle an out of memory condition
 */
static void
xmlIOErrMemory(const char *extra)
{
    __xmlSimpleError(XML_FROM_IO, XML_ERR_NO_MEMORY, NULL, NULL, extra);
}

/**
238
 * __xmlIOErr:
239
 * @code:  the error number
240
 * @
241 242
 * @extra:  extra informations
 *
243
 * Handle an I/O error
244
 */
245 246
void
__xmlIOErr(int domain, int code, const char *extra)
247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377
{
    unsigned int idx;

    if (code == 0) {
#ifdef HAVE_ERRNO_H
	if (errno == 0) code = 0;
#ifdef EACCES
        else if (errno == EACCES) code = XML_IO_EACCES;
#endif
#ifdef EAGAIN
        else if (errno == EAGAIN) code = XML_IO_EAGAIN;
#endif
#ifdef EBADF
        else if (errno == EBADF) code = XML_IO_EBADF;
#endif
#ifdef EBADMSG
        else if (errno == EBADMSG) code = XML_IO_EBADMSG;
#endif
#ifdef EBUSY
        else if (errno == EBUSY) code = XML_IO_EBUSY;
#endif
#ifdef ECANCELED
        else if (errno == ECANCELED) code = XML_IO_ECANCELED;
#endif
#ifdef ECHILD
        else if (errno == ECHILD) code = XML_IO_ECHILD;
#endif
#ifdef EDEADLK
        else if (errno == EDEADLK) code = XML_IO_EDEADLK;
#endif
#ifdef EDOM
        else if (errno == EDOM) code = XML_IO_EDOM;
#endif
#ifdef EEXIST
        else if (errno == EEXIST) code = XML_IO_EEXIST;
#endif
#ifdef EFAULT
        else if (errno == EFAULT) code = XML_IO_EFAULT;
#endif
#ifdef EFBIG
        else if (errno == EFBIG) code = XML_IO_EFBIG;
#endif
#ifdef EINPROGRESS
        else if (errno == EINPROGRESS) code = XML_IO_EINPROGRESS;
#endif
#ifdef EINTR
        else if (errno == EINTR) code = XML_IO_EINTR;
#endif
#ifdef EINVAL
        else if (errno == EINVAL) code = XML_IO_EINVAL;
#endif
#ifdef EIO
        else if (errno == EIO) code = XML_IO_EIO;
#endif
#ifdef EISDIR
        else if (errno == EISDIR) code = XML_IO_EISDIR;
#endif
#ifdef EMFILE
        else if (errno == EMFILE) code = XML_IO_EMFILE;
#endif
#ifdef EMLINK
        else if (errno == EMLINK) code = XML_IO_EMLINK;
#endif
#ifdef EMSGSIZE
        else if (errno == EMSGSIZE) code = XML_IO_EMSGSIZE;
#endif
#ifdef ENAMETOOLONG
        else if (errno == ENAMETOOLONG) code = XML_IO_ENAMETOOLONG;
#endif
#ifdef ENFILE
        else if (errno == ENFILE) code = XML_IO_ENFILE;
#endif
#ifdef ENODEV
        else if (errno == ENODEV) code = XML_IO_ENODEV;
#endif
#ifdef ENOENT
        else if (errno == ENOENT) code = XML_IO_ENOENT;
#endif
#ifdef ENOEXEC
        else if (errno == ENOEXEC) code = XML_IO_ENOEXEC;
#endif
#ifdef ENOLCK
        else if (errno == ENOLCK) code = XML_IO_ENOLCK;
#endif
#ifdef ENOMEM
        else if (errno == ENOMEM) code = XML_IO_ENOMEM;
#endif
#ifdef ENOSPC
        else if (errno == ENOSPC) code = XML_IO_ENOSPC;
#endif
#ifdef ENOSYS
        else if (errno == ENOSYS) code = XML_IO_ENOSYS;
#endif
#ifdef ENOTDIR
        else if (errno == ENOTDIR) code = XML_IO_ENOTDIR;
#endif
#ifdef ENOTEMPTY
        else if (errno == ENOTEMPTY) code = XML_IO_ENOTEMPTY;
#endif
#ifdef ENOTSUP
        else if (errno == ENOTSUP) code = XML_IO_ENOTSUP;
#endif
#ifdef ENOTTY
        else if (errno == ENOTTY) code = XML_IO_ENOTTY;
#endif
#ifdef ENXIO
        else if (errno == ENXIO) code = XML_IO_ENXIO;
#endif
#ifdef EPERM
        else if (errno == EPERM) code = XML_IO_EPERM;
#endif
#ifdef EPIPE
        else if (errno == EPIPE) code = XML_IO_EPIPE;
#endif
#ifdef ERANGE
        else if (errno == ERANGE) code = XML_IO_ERANGE;
#endif
#ifdef EROFS
        else if (errno == EROFS) code = XML_IO_EROFS;
#endif
#ifdef ESPIPE
        else if (errno == ESPIPE) code = XML_IO_ESPIPE;
#endif
#ifdef ESRCH
        else if (errno == ESRCH) code = XML_IO_ESRCH;
#endif
#ifdef ETIMEDOUT
        else if (errno == ETIMEDOUT) code = XML_IO_ETIMEDOUT;
#endif
#ifdef EXDEV
        else if (errno == EXDEV) code = XML_IO_EXDEV;
378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404
#endif
#ifdef ENOTSOCK
        else if (errno == ENOTSOCK) code = XML_IO_ENOTSOCK;
#endif
#ifdef EISCONN
        else if (errno == EISCONN) code = XML_IO_EISCONN;
#endif
#ifdef ECONNREFUSED
        else if (errno == ECONNREFUSED) code = XML_IO_ECONNREFUSED;
#endif
#ifdef ETIMEDOUT
        else if (errno == ETIMEDOUT) code = XML_IO_ETIMEDOUT;
#endif
#ifdef ENETUNREACH
        else if (errno == ENETUNREACH) code = XML_IO_ENETUNREACH;
#endif
#ifdef EADDRINUSE
        else if (errno == EADDRINUSE) code = XML_IO_EADDRINUSE;
#endif
#ifdef EINPROGRESS
        else if (errno == EINPROGRESS) code = XML_IO_EINPROGRESS;
#endif
#ifdef EALREADY
        else if (errno == EALREADY) code = XML_IO_EALREADY;
#endif
#ifdef EAFNOSUPPORT
        else if (errno == EAFNOSUPPORT) code = XML_IO_EAFNOSUPPORT;
405 406 407 408 409 410 411
#endif
        else code = XML_IO_UNKNOWN;
#endif /* HAVE_ERRNO_H */
    }
    idx = 0;
    if (code >= XML_IO_UNKNOWN) idx = code - XML_IO_UNKNOWN;
    if (idx >= (sizeof(IOerr) / sizeof(IOerr[0]))) idx = 0;
412

413 414 415 416 417 418 419 420 421 422 423 424 425 426
    __xmlSimpleError(domain, code, NULL, IOerr[idx], extra);
}

/**
 * xmlIOErr:
 * @code:  the error number
 * @extra:  extra informations
 *
 * Handle an I/O error
 */
static void
xmlIOErr(int code, const char *extra)
{
    __xmlIOErr(XML_FROM_IO, code, extra);
427
}
428

429
/**
430 431
 * __xmlLoaderErr:
 * @ctx: the parser context
432 433 434 435
 * @extra:  extra informations
 *
 * Handle a resource access error
 */
436 437
void
__xmlLoaderErr(void *ctx, const char *msg, const char *filename)
438
{
439
    xmlParserCtxtPtr ctxt = (xmlParserCtxtPtr) ctx;
440
    xmlStructuredErrorFunc schannel = NULL;
441 442 443 444
    xmlGenericErrorFunc channel = NULL;
    void *data = NULL;
    xmlErrorLevel level = XML_ERR_ERROR;

445 446 447
    if ((ctxt != NULL) && (ctxt->disableSAX != 0) &&
        (ctxt->instate == XML_PARSER_EOF))
	return;
448 449 450 451 452 453 454 455
    if ((ctxt != NULL) && (ctxt->sax != NULL)) {
        if (ctxt->validate) {
	    channel = ctxt->sax->error;
	    level = XML_ERR_ERROR;
	} else {
	    channel = ctxt->sax->warning;
	    level = XML_ERR_WARNING;
	}
Daniel Veillard's avatar
Daniel Veillard committed
456 457
	if (ctxt->sax->initialized == XML_SAX2_MAGIC)
	    schannel = ctxt->sax->serror;
458 459
	data = ctxt->userData;
    }
460
    __xmlRaiseError(schannel, channel, data, ctxt, NULL, XML_FROM_IO,
461 462 463
                    XML_IO_LOAD_ERROR, level, NULL, 0,
		    filename, NULL, NULL, 0, 0,
		    msg, filename);
464

465 466
}

467 468
/************************************************************************
 *									*
469
 *		Tree memory error handler				*
470 471
 *									*
 ************************************************************************/
472
/**
473
 * xmlNormalizeWindowsPath:
474
 * @path: the input file path
475
 *
476 477
 * This function is obsolete. Please see xmlURIFromPath in uri.c for
 * a better solution.
478 479
 *
 * Returns a canonicalized version of the path
480 481 482 483
 */
xmlChar *
xmlNormalizeWindowsPath(const xmlChar *path)
{
484
    return xmlCanonicPath(path);
485 486
}

487 488 489 490
/**
 * xmlCleanupInputCallbacks:
 *
 * clears the entire input callback table. this includes the
491
 * compiled-in I/O.
492 493 494 495 496 497 498 499 500
 */
void
xmlCleanupInputCallbacks(void)
{
    int i;

    if (!xmlInputCallbackInitialized)
        return;

501
    for (i = xmlInputCallbackNr - 1; i >= 0; i--) {
502 503 504 505 506 507 508
        xmlInputCallbackTable[i].matchcallback = NULL;
        xmlInputCallbackTable[i].opencallback = NULL;
        xmlInputCallbackTable[i].readcallback = NULL;
        xmlInputCallbackTable[i].closecallback = NULL;
    }

    xmlInputCallbackNr = 0;
509
    xmlInputCallbackInitialized = 0;
510 511
}

512
/**
513
 * xmlPopInputCallbacks:
514 515
 *
 * Clear the top input callback from the input stack. this includes the
516
 * compiled-in I/O.
517 518 519 520 521 522 523 524 525 526 527
 *
 * Returns the number of input callback registered or -1 in case of error.
 */
int
xmlPopInputCallbacks(void)
{
    if (!xmlInputCallbackInitialized)
        return(-1);

    if (xmlInputCallbackNr <= 0)
        return(-1);
528

529 530 531 532 533 534 535 536 537
    xmlInputCallbackNr--;
    xmlInputCallbackTable[xmlInputCallbackNr].matchcallback = NULL;
    xmlInputCallbackTable[xmlInputCallbackNr].opencallback = NULL;
    xmlInputCallbackTable[xmlInputCallbackNr].readcallback = NULL;
    xmlInputCallbackTable[xmlInputCallbackNr].closecallback = NULL;

    return(xmlInputCallbackNr);
}

538
#ifdef LIBXML_OUTPUT_ENABLED
539 540 541 542
/**
 * xmlCleanupOutputCallbacks:
 *
 * clears the entire output callback table. this includes the
543
 * compiled-in I/O callbacks.
544 545 546 547 548 549 550 551 552
 */
void
xmlCleanupOutputCallbacks(void)
{
    int i;

    if (!xmlOutputCallbackInitialized)
        return;

553
    for (i = xmlOutputCallbackNr - 1; i >= 0; i--) {
554 555 556 557 558 559 560
        xmlOutputCallbackTable[i].matchcallback = NULL;
        xmlOutputCallbackTable[i].opencallback = NULL;
        xmlOutputCallbackTable[i].writecallback = NULL;
        xmlOutputCallbackTable[i].closecallback = NULL;
    }

    xmlOutputCallbackNr = 0;
561
    xmlOutputCallbackInitialized = 0;
562
}
563
#endif /* LIBXML_OUTPUT_ENABLED */
564

565 566 567 568 569 570
/************************************************************************
 *									*
 *		Standard I/O for file accesses				*
 *									*
 ************************************************************************/

571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592
#if defined(_WIN32) || defined (__DJGPP__) && !defined (__CYGWIN__)

/**
 *  xmlWrapOpenUtf8:
 * @path:  the path in utf-8 encoding
 * @mode:  type of access (0 - read, 1 - write)
 *
 * function opens the file specified by @path
 *
 */
static FILE*
xmlWrapOpenUtf8(const char *path,int mode)
{
    FILE *fd = NULL;
    wchar_t *wPath;

    wPath = __xmlIOWin32UTF8ToWChar(path);
    if(wPath)
    {
       fd = _wfopen(wPath, mode ? L"wb" : L"rb");
       xmlFree(wPath);
    }
593
    /* maybe path in native encoding */
594 595 596 597 598 599
    if(fd == NULL)
       fd = fopen(path, mode ? "wb" : "rb");

    return fd;
}

600
#ifdef LIBXML_ZLIB_ENABLED
601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625
static gzFile
xmlWrapGzOpenUtf8(const char *path, const char *mode)
{
    gzFile fd;
    wchar_t *wPath;

    fd = gzopen (path, mode);
    if (fd)
        return fd;

    wPath = __xmlIOWin32UTF8ToWChar(path);
    if(wPath)
    {
	int d, m = (strstr(mode, "r") ? O_RDONLY : O_RDWR);
#ifdef _O_BINARY
        m |= (strstr(mode, "b") ? _O_BINARY : 0);
#endif
	d = _wopen(wPath, m);
	if (d >= 0)
	    fd = gzdopen(d, mode);
        xmlFree(wPath);
    }

    return fd;
}
Rob Richards's avatar
Rob Richards committed
626
#endif
627

628 629 630 631 632 633 634 635 636
/**
 *  xmlWrapStatUtf8:
 * @path:  the path in utf-8 encoding
 * @info:  structure that stores results
 *
 * function obtains information about the file or directory
 *
 */
static int
637
xmlWrapStatUtf8(const char *path, struct _stat *info) {
638 639 640 641
    int retval = -1;
    wchar_t *wPath;

    wPath = __xmlIOWin32UTF8ToWChar(path);
642 643
    if (wPath) {
       retval = _wstat(wPath, info);
644 645
       xmlFree(wPath);
    }
646
    /* maybe path in native encoding */
647
    if(retval < 0)
648
       retval = _stat(path, info);
649 650 651 652 653
    return retval;
}

#endif

654
/**
655
 * xmlCheckFilename:
656 657 658 659 660 661 662 663 664
 * @path:  the path to check
 *
 * function checks to see if @path is a valid source
 * (file, socket...) for XML.
 *
 * if stat is not available on the target machine,
 * returns 1.  if stat fails, returns 0 (if calling
 * stat on the filename fails, it can't be right).
 * if stat succeeds and the file is a directory,
665
 * returns 2.  otherwise returns 1.
666 667
 */

668
int
669 670
xmlCheckFilename (const char *path)
{
671
#ifdef HAVE_STAT
672 673 674
#if defined(_WIN32) || defined (__DJGPP__) && !defined (__CYGWIN__)
    struct _stat stat_buffer;
#else
675
    struct stat stat_buffer;
676
#endif
677
#endif
678 679
    if (path == NULL)
	return(0);
680

681
#ifdef HAVE_STAT
682
#if defined(_WIN32) || defined (__DJGPP__) && !defined (__CYGWIN__)
683 684 685 686 687 688 689 690
    /*
     * On Windows stat and wstat do not work with long pathname,
     * which start with '\\?\'
     */
    if ((path[0] == '\\') && (path[1] == '\\') && (path[2] == '?') &&
	(path[3] == '\\') )
	    return 1;

691
    if (xmlWrapStatUtf8(path, &stat_buffer) == -1)
692 693
        return 0;
#else
694 695
    if (stat(path, &stat_buffer) == -1)
        return 0;
696
#endif
697
#ifdef S_ISDIR
698
    if (S_ISDIR(stat_buffer.st_mode))
699
        return 2;
700
#endif
701
#endif /* HAVE_STAT */
702 703 704
    return 1;
}

705
/**
706
 * xmlInputReadCallbackNop:
707
 *
708
 * No Operation xmlInputReadCallback function, does nothing.
709 710 711
 *
 * Returns zero
 */
712
int
713 714 715
xmlInputReadCallbackNop(void *context ATTRIBUTE_UNUSED,
                        char *buffer ATTRIBUTE_UNUSED,
                        int len ATTRIBUTE_UNUSED) {
716 717 718 719 720 721 722 723 724 725 726 727 728
    return(0);
}

/**
 * xmlFdRead:
 * @context:  the I/O context
 * @buffer:  where to drop data
 * @len:  number of bytes to read
 *
 * Read @len bytes to @buffer from the I/O channel.
 *
 * Returns the number of bytes written
 */
729
static int
730
xmlFdRead (void * context, char * buffer, int len) {
731 732
    int ret;

733
    ret = read((int) (ptrdiff_t) context, &buffer[0], len);
734 735
    if (ret < 0) xmlIOErr(0, "read()");
    return(ret);
736 737
}

738
#ifdef LIBXML_OUTPUT_ENABLED
739 740 741 742 743 744 745 746 747 748
/**
 * xmlFdWrite:
 * @context:  the I/O context
 * @buffer:  where to get data
 * @len:  number of bytes to write
 *
 * Write @len bytes from @buffer to the I/O channel.
 *
 * Returns the number of bytes written
 */
749
static int
750
xmlFdWrite (void * context, const char * buffer, int len) {
751
    int ret = 0;
752

753
    if (len > 0) {
754
	ret = write((int) (ptrdiff_t) context, &buffer[0], len);
755 756
	if (ret < 0) xmlIOErr(0, "write()");
    }
757
    return(ret);
758
}
759
#endif /* LIBXML_OUTPUT_ENABLED */
760 761 762 763 764 765

/**
 * xmlFdClose:
 * @context:  the I/O context
 *
 * Close an I/O channel
766 767
 *
 * Returns 0 in case of success and error code otherwise
768
 */
769
static int
770
xmlFdClose (void * context) {
771
    int ret;
772
    ret = close((int) (ptrdiff_t) context);
773 774
    if (ret < 0) xmlIOErr(0, "close()");
    return(ret);
775 776 777 778 779 780 781 782 783 784
}

/**
 * xmlFileMatch:
 * @filename:  the URI for matching
 *
 * input from FILE *
 *
 * Returns 1 if matches, 0 otherwise
 */
785
int
786
xmlFileMatch (const char *filename ATTRIBUTE_UNUSED) {
787 788 789 790
    return(1);
}

/**
791
 * xmlFileOpen_real:
792 793 794 795 796 797 798
 * @filename:  the URI for matching
 *
 * input from FILE *, supports compressed input
 * if @filename is " " then the standard input is used
 *
 * Returns an I/O context or NULL in case of error
 */
799 800
static void *
xmlFileOpen_real (const char *filename) {
801
    const char *path = filename;
802
    FILE *fd;
803

804 805 806
    if (filename == NULL)
        return(NULL);

807 808 809 810 811
    if (!strcmp(filename, "-")) {
	fd = stdin;
	return((void *) fd);
    }

812
    if (!xmlStrncasecmp(BAD_CAST filename, BAD_CAST "file://localhost/", 17)) {
813
#if defined (_WIN32) || defined (__DJGPP__) && !defined(__CYGWIN__)
814 815
	path = &filename[17];
#else
816
	path = &filename[16];
817
#endif
818
    } else if (!xmlStrncasecmp(BAD_CAST filename, BAD_CAST "file:///", 8)) {
819
#if defined (_WIN32) || defined (__DJGPP__) && !defined(__CYGWIN__)
820 821
	path = &filename[8];
#else
822
	path = &filename[7];
823 824 825 826 827 828 829
#endif
    } else if (!xmlStrncasecmp(BAD_CAST filename, BAD_CAST "file:/", 6)) {
        /* lots of generators seems to lazy to read RFC 1738 */
#if defined (_WIN32) || defined (__DJGPP__) && !defined(__CYGWIN__)
	path = &filename[6];
#else
	path = &filename[5];
830
#endif
831
    }
832

833 834
    /* Do not check DDNAME on zOS ! */
#if !defined(__MVS__)
835 836
    if (!xmlCheckFilename(path))
        return(NULL);
837
#endif
838

839
#if defined(_WIN32) || defined (__DJGPP__) && !defined (__CYGWIN__)
840
    fd = xmlWrapOpenUtf8(path, 0);
841 842
#else
    fd = fopen(path, "r");
843
#endif /* WIN32 */
844
    if (fd == NULL) xmlIOErr(0, path);
845 846 847
    return((void *) fd);
}

848 849 850 851 852 853
/**
 * xmlFileOpen:
 * @filename:  the URI for matching
 *
 * Wrapper around xmlFileOpen_real that try it with an unescaped
 * version of @filename, if this fails fallback to @filename
854 855
 *
 * Returns a handler or NULL in case or failure
856 857 858 859 860
 */
void *
xmlFileOpen (const char *filename) {
    char *unescaped;
    void *retval;
861

862 863 864 865 866 867 868
    retval = xmlFileOpen_real(filename);
    if (retval == NULL) {
	unescaped = xmlURIUnescapeString(filename, 0, NULL);
	if (unescaped != NULL) {
	    retval = xmlFileOpen_real(unescaped);
	    xmlFree(unescaped);
	}
869
    }
870

871 872 873
    return retval;
}

874
#ifdef LIBXML_OUTPUT_ENABLED
875 876 877 878 879 880 881 882 883
/**
 * xmlFileOpenW:
 * @filename:  the URI for matching
 *
 * output to from FILE *,
 * if @filename is "-" then the standard output is used
 *
 * Returns an I/O context or NULL in case of error
 */
884
static void *
885 886
xmlFileOpenW (const char *filename) {
    const char *path = NULL;
887
    FILE *fd;
888 889 890 891 892 893

    if (!strcmp(filename, "-")) {
	fd = stdout;
	return((void *) fd);
    }

894
    if (!xmlStrncasecmp(BAD_CAST filename, BAD_CAST "file://localhost/", 17))
895
#if defined (_WIN32) || defined (__DJGPP__) && !defined(__CYGWIN__)
896 897
	path = &filename[17];
#else
898
	path = &filename[16];
899
#endif
900
    else if (!xmlStrncasecmp(BAD_CAST filename, BAD_CAST "file:///", 8)) {
901
#if defined (_WIN32) || defined (__DJGPP__) && !defined(__CYGWIN__)
902 903
	path = &filename[8];
#else
904
	path = &filename[7];
905
#endif
906
    } else
907 908 909 910 911
	path = filename;

    if (path == NULL)
	return(NULL);

912
#if defined(_WIN32) || defined (__DJGPP__) && !defined (__CYGWIN__)
913
    fd = xmlWrapOpenUtf8(path, 1);
914 915
#elif(__MVS__)
    fd = fopen(path, "w");
916
#else
917
    fd = fopen(path, "wb");
918
#endif /* WIN32 */
919

920
    if (fd == NULL) xmlIOErr(0, path);
921 922
    return((void *) fd);
}
923
#endif /* LIBXML_OUTPUT_ENABLED */
924 925 926 927 928 929 930 931 932

/**
 * xmlFileRead:
 * @context:  the I/O context
 * @buffer:  where to drop data
 * @len:  number of bytes to write
 *
 * Read @len bytes to @buffer from the I/O channel.
 *
933
 * Returns the number of bytes written or < 0 in case of failure
934
 */
935
int
936
xmlFileRead (void * context, char * buffer, int len) {
937
    int ret;
938
    if ((context == NULL) || (buffer == NULL))
939
        return(-1);
940 941 942
    ret = fread(&buffer[0], 1,  len, (FILE *) context);
    if (ret < 0) xmlIOErr(0, "fread()");
    return(ret);
943 944
}

945
#ifdef LIBXML_OUTPUT_ENABLED
946 947 948 949 950 951 952 953 954 955
/**
 * xmlFileWrite:
 * @context:  the I/O context
 * @buffer:  where to drop data
 * @len:  number of bytes to write
 *
 * Write @len bytes from @buffer to the I/O channel.
 *
 * Returns the number of bytes written
 */
956
static int
957
xmlFileWrite (void * context, const char * buffer, int len) {
958 959
    int items;

960
    if ((context == NULL) || (buffer == NULL))
961
        return(-1);
962
    items = fwrite(&buffer[0], len, 1, (FILE *) context);
963
    if ((items == 0) && (ferror((FILE *) context))) {
964
        xmlIOErr(0, "fwrite()");
965 966
	return(-1);
    }
967
    return(items * len);
968
}
969
#endif /* LIBXML_OUTPUT_ENABLED */
970 971 972 973 974 975

/**
 * xmlFileClose:
 * @context:  the I/O context
 *
 * Close an I/O channel
976 977
 *
 * Returns 0 or -1 in case of error
978
 */
979
int
980
xmlFileClose (void * context) {
981
    FILE *fil;
982
    int ret;
983

984 985
    if (context == NULL)
        return(-1);
986
    fil = (FILE *) context;
987 988 989 990
    if ((fil == stdout) || (fil == stderr)) {
        ret = fflush(fil);
	if (ret < 0)
	    xmlIOErr(0, "fflush()");
991
	return(0);
992 993
    }
    if (fil == stdin)
994
	return(0);
995 996 997 998
    ret = ( fclose((FILE *) context) == EOF ) ? -1 : 0;
    if (ret < 0)
        xmlIOErr(0, "fclose()");
    return(ret);
999 1000 1001 1002 1003 1004 1005 1006
}

/**
 * xmlFileFlush:
 * @context:  the I/O context
 *
 * Flush an I/O channel
 */
1007
static int
1008
xmlFileFlush (void * context) {
1009
    int ret;
1010 1011 1012

    if (context == NULL)
        return(-1);
1013 1014 1015 1016
    ret = ( fflush((FILE *) context) == EOF ) ? -1 : 0;
    if (ret < 0)
        xmlIOErr(0, "fflush()");
    return(ret);
1017 1018
}

1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040
#ifdef LIBXML_OUTPUT_ENABLED
/**
 * xmlBufferWrite:
 * @context:  the xmlBuffer
 * @buffer:  the data to write
 * @len:  number of bytes to write
 *
 * Write @len bytes from @buffer to the xml buffer
 *
 * Returns the number of bytes written
 */
static int
xmlBufferWrite (void * context, const char * buffer, int len) {
    int ret;

    ret = xmlBufferAdd((xmlBufferPtr) context, (const xmlChar *) buffer, len);
    if (ret != 0)
        return(-1);
    return(len);
}
#endif

1041
#ifdef LIBXML_ZLIB_ENABLED
1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054
/************************************************************************
 *									*
 *		I/O for compressed file accesses			*
 *									*
 ************************************************************************/
/**
 * xmlGzfileMatch:
 * @filename:  the URI for matching
 *
 * input from compressed file test
 *
 * Returns 1 if matches, 0 otherwise
 */
1055
static int
1056
xmlGzfileMatch (const char *filename ATTRIBUTE_UNUSED) {
1057 1058 1059 1060
    return(1);
}

/**
1061
 * xmlGzfileOpen_real:
1062 1063 1064 1065 1066 1067 1068
 * @filename:  the URI for matching
 *
 * input from compressed file open
 * if @filename is " " then the standard input is used
 *
 * Returns an I/O context or NULL in case of error
 */
1069
static void *
1070
xmlGzfileOpen_real (const char *filename) {
1071 1072 1073 1074
    const char *path = NULL;
    gzFile fd;

    if (!strcmp(filename, "-")) {
1075 1076
        int duped_fd = dup(fileno(stdin));
        fd = gzdopen(duped_fd, "rb");
1077
        if (fd == Z_NULL && duped_fd >= 0) {
1078 1079 1080
            close(duped_fd);  /* gzdOpen() does not close on failure */
        }

1081 1082 1083
	return((void *) fd);
    }

1084
    if (!xmlStrncasecmp(BAD_CAST filename, BAD_CAST "file://localhost/", 17))
1085
#if defined (_WIN32) || defined (__DJGPP__) && !defined(__CYGWIN__)
1086 1087
	path = &filename[17];
#else
1088
	path = &filename[16];
1089
#endif
1090
    else if (!xmlStrncasecmp(BAD_CAST filename, BAD_CAST "file:///", 8)) {
1091
#if defined (_WIN32) || defined (__DJGPP__) && !defined(__CYGWIN__)
1092 1093
	path = &filename[8];
#else
1094
	path = &filename[7];
1095
#endif
1096
    } else
1097 1098 1099 1100 1101 1102 1103
	path = filename;

    if (path == NULL)
	return(NULL);
    if (!xmlCheckFilename(path))
        return(NULL);

1104
#if defined(_WIN32) || defined (__DJGPP__) && !defined (__CYGWIN__)
1105
    fd = xmlWrapGzOpenUtf8(path, "rb");
1106
#else
1107
    fd = gzopen(path, "rb");
1108
#endif
1109 1110 1111
    return((void *) fd);
}

1112 1113 1114 1115
/**
 * xmlGzfileOpen:
 * @filename:  the URI for matching
 *
1116 1117
 * Wrapper around xmlGzfileOpen if the open fais, it will
 * try to unescape @filename
1118 1119 1120 1121 1122
 */
static void *
xmlGzfileOpen (const char *filename) {
    char *unescaped;
    void *retval;
1123 1124 1125 1126 1127 1128 1129 1130

    retval = xmlGzfileOpen_real(filename);
    if (retval == NULL) {
	unescaped = xmlURIUnescapeString(filename, 0, NULL);
	if (unescaped != NULL) {
	    retval = xmlGzfileOpen_real(unescaped);
	}
	xmlFree(unescaped);
1131 1132 1133 1134
    }
    return retval;
}

1135
#ifdef LIBXML_OUTPUT_ENABLED
1136 1137 1138 1139 1140 1141 1142 1143 1144 1145
/**
 * xmlGzfileOpenW:
 * @filename:  the URI for matching
 * @compression:  the compression factor (0 - 9 included)
 *
 * input from compressed file open
 * if @filename is " " then the standard input is used
 *
 * Returns an I/O context or NULL in case of error
 */
1146
static void *
1147 1148 1149 1150 1151
xmlGzfileOpenW (const char *filename, int compression) {
    const char *path = NULL;
    char mode[15];
    gzFile fd;

1152
    snprintf(mode, sizeof(mode), "wb%d", compression);
1153
    if (!strcmp(filename, "-")) {
1154 1155
        int duped_fd = dup(fileno(stdout));
        fd = gzdopen(duped_fd, "rb");
1156
        if (fd == Z_NULL && duped_fd >= 0) {
1157 1158 1159
            close(duped_fd);  /* gzdOpen() does not close on failure */
        }

1160 1161 1162
	return((void *) fd);
    }

1163
    if (!xmlStrncasecmp(BAD_CAST filename, BAD_CAST "file://localhost/", 17))
1164
#if defined (_WIN32) || defined (__DJGPP__) && !defined(__CYGWIN__)
1165 1166
	path = &filename[17];
#else
1167
	path = &filename[16];
1168
#endif
1169
    else if (!xmlStrncasecmp(BAD_CAST filename, BAD_CAST "file:///", 8)) {
1170
#if defined (_WIN32) || defined (__DJGPP__) && !defined(__CYGWIN__)
1171 1172
	path = &filename[8];
#else
1173
	path = &filename[7];
1174
#endif
1175
    } else
1176 1177 1178 1179 1180
	path = filename;

    if (path == NULL)
	return(NULL);

1181
#if defined(_WIN32) || defined (__DJGPP__) && !defined (__CYGWIN__)
1182
    fd = xmlWrapGzOpenUtf8(path, mode);
1183
#else
1184
    fd = gzopen(path, mode);
1185
#endif
1186 1187
    return((void *) fd);
}
1188
#endif /* LIBXML_OUTPUT_ENABLED */
1189 1190 1191 1192 1193 1194 1195 1196 1197

/**
 * xmlGzfileRead:
 * @context:  the I/O context
 * @buffer:  where to drop data
 * @len:  number of bytes to write
 *
 * Read @len bytes to @buffer from the compressed I/O channel.
 *
Nick Wellnhofer's avatar
Nick Wellnhofer committed
1198
 * Returns the number of bytes read.
1199
 */
1200
static int
1201
xmlGzfileRead (void * context, char * buffer, int len) {
1202 1203 1204 1205 1206
    int ret;

    ret = gzread((gzFile) context, &buffer[0], len);
    if (ret < 0) xmlIOErr(0, "gzread()");
    return(ret);
1207 1208
}

1209
#ifdef LIBXML_OUTPUT_ENABLED
1210 1211 1212 1213 1214 1215 1216 1217 1218 1219
/**
 * xmlGzfileWrite:
 * @context:  the I/O context
 * @buffer:  where to drop data
 * @len:  number of bytes to write
 *
 * Write @len bytes from @buffer to the compressed I/O channel.
 *
 * Returns the number of bytes written
 */
1220
static int
1221
xmlGzfileWrite (void * context, const char * buffer, int len) {
1222 1223 1224 1225 1226
    int ret;

    ret = gzwrite((gzFile) context, (char *) &buffer[0], len);
    if (ret < 0) xmlIOErr(0, "gzwrite()");
    return(ret);
1227
}
1228
#endif /* LIBXML_OUTPUT_ENABLED */
1229 1230 1231 1232 1233 1234 1235

/**
 * xmlGzfileClose:
 * @context:  the I/O context
 *
 * Close a compressed I/O channel
 */
1236
static int
1237
xmlGzfileClose (void * context) {
1238 1239 1240 1241 1242
    int ret;

    ret =  (gzclose((gzFile) context) == Z_OK ) ? 0 : -1;
    if (ret < 0) xmlIOErr(0, "gzclose()");
    return(ret);
1243
}
1244
#endif /* LIBXML_ZLIB_ENABLED */
1245

1246
#ifdef LIBXML_LZMA_ENABLED
1247 1248 1249 1250 1251
/************************************************************************
 *									*
 *		I/O for compressed file accesses			*
 *									*
 ************************************************************************/
1252
#include "xzlib.h"
1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280
/**
 * xmlXzfileMatch:
 * @filename:  the URI for matching
 *
 * input from compressed file test
 *
 * Returns 1 if matches, 0 otherwise
 */
static int
xmlXzfileMatch (const char *filename ATTRIBUTE_UNUSED) {
    return(1);
}

/**
 * xmlXzFileOpen_real:
 * @filename:  the URI for matching
 *
 * input from compressed file open
 * if @filename is " " then the standard input is used
 *
 * Returns an I/O context or NULL in case of error
 */
static void *
xmlXzfileOpen_real (const char *filename) {
    const char *path = NULL;
    xzFile fd;

    if (!strcmp(filename, "-")) {
Patrick Monnerat's avatar
Patrick Monnerat committed
1281
        fd = __libxml2_xzdopen(dup(fileno(stdin)), "rb");
1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299
	return((void *) fd);
    }

    if (!xmlStrncasecmp(BAD_CAST filename, BAD_CAST "file://localhost/", 17)) {
	path = &filename[16];
    } else if (!xmlStrncasecmp(BAD_CAST filename, BAD_CAST "file:///", 8)) {
	path = &filename[7];
    } else if (!xmlStrncasecmp(BAD_CAST filename, BAD_CAST "file:/", 6)) {
        /* lots of generators seems to lazy to read RFC 1738 */
	path = &filename[5];
    } else
	path = filename;

    if (path == NULL)
	return(NULL);
    if (!xmlCheckFilename(path))
        return(NULL);

1300
    fd = __libxml2_xzopen(path, "rb");
1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343
    return((void *) fd);
}

/**
 * xmlXzfileOpen:
 * @filename:  the URI for matching
 *
 * Wrapper around xmlXzfileOpen_real that try it with an unescaped
 * version of @filename, if this fails fallback to @filename
 *
 * Returns a handler or NULL in case or failure
 */
static void *
xmlXzfileOpen (const char *filename) {
    char *unescaped;
    void *retval;

    retval = xmlXzfileOpen_real(filename);
    if (retval == NULL) {
	unescaped = xmlURIUnescapeString(filename, 0, NULL);
	if (unescaped != NULL) {
	    retval = xmlXzfileOpen_real(unescaped);
	}
	xmlFree(unescaped);
    }

    return retval;
}

/**
 * xmlXzfileRead:
 * @context:  the I/O context
 * @buffer:  where to drop data
 * @len:  number of bytes to write
 *
 * Read @len bytes to @buffer from the compressed I/O channel.
 *
 * Returns the number of bytes written
 */
static int
xmlXzfileRead (void * context, char * buffer, int len) {
    int ret;

1344
    ret = __libxml2_xzread((xzFile) context, &buffer[0], len);
1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358
    if (ret < 0) xmlIOErr(0, "xzread()");
    return(ret);
}

/**
 * xmlXzfileClose:
 * @context:  the I/O context
 *
 * Close a compressed I/O channel
 */
static int
xmlXzfileClose (void * context) {
    int ret;

1359
    ret =  (__libxml2_xzclose((xzFile) context) == LZMA_OK ) ? 0 : -1;
1360 1361 1362
    if (ret < 0) xmlIOErr(0, "xzclose()");
    return(ret);
}
1363
#endif /* LIBXML_LZMA_ENABLED */
1364

1365 1366 1367 1368 1369 1370
#ifdef LIBXML_HTTP_ENABLED
/************************************************************************
 *									*
 *			I/O for HTTP file accesses			*
 *									*
 ************************************************************************/
1371

1372
#ifdef LIBXML_OUTPUT_ENABLED
1373 1374 1375 1376 1377 1378 1379 1380 1381 1382
typedef struct xmlIOHTTPWriteCtxt_
{
    int			compression;

    char *		uri;

    void *		doc_buff;

} xmlIOHTTPWriteCtxt, *xmlIOHTTPWriteCtxtPtr;

1383
#ifdef LIBXML_ZLIB_ENABLED
1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425

#define DFLT_WBITS		( -15 )
#define DFLT_MEM_LVL		( 8 )
#define GZ_MAGIC1		( 0x1f )
#define GZ_MAGIC2		( 0x8b )
#define LXML_ZLIB_OS_CODE	( 0x03 )
#define INIT_HTTP_BUFF_SIZE	( 32768 )
#define DFLT_ZLIB_RATIO		( 5 )

/*
**  Data structure and functions to work with sending compressed data
**  via HTTP.
*/

typedef struct xmlZMemBuff_
{
   unsigned long	size;
   unsigned long	crc;

   unsigned char *	zbuff;
   z_stream		zctrl;

} xmlZMemBuff, *xmlZMemBuffPtr;

/**
 * append_reverse_ulong
 * @buff:  Compressed memory buffer
 * @data:  Unsigned long to append
 *
 * Append a unsigned long in reverse byte order to the end of the
 * memory buffer.
 */
static void
append_reverse_ulong( xmlZMemBuff * buff, unsigned long data ) {

    int		idx;

    if ( buff == NULL )
	return;

    /*
    **  This is plagiarized from putLong in gzio.c (zlib source) where
1426
    **  the number "4" is hardcoded.  If zlib is ever patched to
1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448
    **  support 64 bit file sizes, this code would need to be patched
    **  as well.
    */

    for ( idx = 0; idx < 4; idx++ ) {
	*buff->zctrl.next_out = ( data & 0xff );
	data >>= 8;
	buff->zctrl.next_out++;
    }

    return;
}

/**
 *
 * xmlFreeZMemBuff
 * @buff:  The memory buffer context to clear
 *
 * Release all the resources associated with the compressed memory buffer.
 */
static void
xmlFreeZMemBuff( xmlZMemBuffPtr buff ) {
William M. Brack's avatar
William M. Brack committed
1449 1450

#ifdef DEBUG_HTTP
1451
    int z_err;
William M. Brack's avatar
William M. Brack committed
1452
#endif
1453 1454 1455 1456 1457 1458

    if ( buff == NULL )
	return;

    xmlFree( buff->zbuff );
#ifdef DEBUG_HTTP
William M. Brack's avatar
William M. Brack committed
1459
    z_err = deflateEnd( &buff->zctrl );
1460 1461 1462 1463
    if ( z_err != Z_OK )
	xmlGenericError( xmlGenericErrorContext,
			"xmlFreeZMemBuff:  Error releasing zlib context:  %d\n",
			z_err );
William M. Brack's avatar
William M. Brack committed
1464 1465
#else
    deflateEnd( &buff->zctrl );
1466
#endif
1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477

    xmlFree( buff );
    return;
}

/**
 * xmlCreateZMemBuff
 *@compression:	Compression value to use
 *
 * Create a memory buffer to hold the compressed XML document.  The
 * compressed document in memory will end up being identical to what
1478
 * would be created if gzopen/gzwrite/gzclose were being used to
1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495
 * write the document to disk.  The code for the header/trailer data to
 * the compression is plagiarized from the zlib source files.
 */
static void *
xmlCreateZMemBuff( int compression ) {

    int			z_err;
    int			hdr_lgth;
    xmlZMemBuffPtr	buff = NULL;

    if ( ( compression < 1 ) || ( compression > 9 ) )
	return ( NULL );

    /*  Create the control and data areas  */

    buff = xmlMalloc( sizeof( xmlZMemBuff ) );
    if ( buff == NULL ) {
1496
	xmlIOErrMemory("creating buffer context");
1497 1498 1499 1500 1501 1502 1503 1504
	return ( NULL );
    }

    (void)memset( buff, 0, sizeof( xmlZMemBuff ) );
    buff->size = INIT_HTTP_BUFF_SIZE;
    buff->zbuff = xmlMalloc( buff->size );
    if ( buff->zbuff == NULL ) {
	xmlFreeZMemBuff( buff );
1505
	xmlIOErrMemory("creating buffer");
1506 1507 1508 1509 1510 1511
	return ( NULL );
    }

    z_err = deflateInit2( &buff->zctrl, compression, Z_DEFLATED,
			    DFLT_WBITS, DFLT_MEM_LVL, Z_DEFAULT_STRATEGY );
    if ( z_err != Z_OK ) {
1512
	xmlChar msg[500];
1513 1514
	xmlFreeZMemBuff( buff );
	buff = NULL;
1515
	xmlStrPrintf(msg, 500,
1516
		    "xmlCreateZMemBuff:  %s %d\n",
1517 1518 1519
		    "Error initializing compression context.  ZLIB error:",
		    z_err );
	xmlIOErr(XML_IO_WRITE, (const char *) msg);
1520 1521 1522 1523
	return ( NULL );
    }

    /*  Set the header data.  The CRC will be needed for the trailer  */
1524
    buff->crc = crc32( 0L, NULL, 0 );
1525 1526
    hdr_lgth = snprintf( (char *)buff->zbuff, buff->size,
			"%c%c%c%c%c%c%c%c%c%c",
1527
			GZ_MAGIC1, GZ_MAGIC2, Z_DEFLATED,
1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564
			0, 0, 0, 0, 0, 0, LXML_ZLIB_OS_CODE );
    buff->zctrl.next_out  = buff->zbuff + hdr_lgth;
    buff->zctrl.avail_out = buff->size - hdr_lgth;

    return ( buff );
}

/**
 * xmlZMemBuffExtend
 * @buff:  Buffer used to compress and consolidate data.
 * @ext_amt:   Number of bytes to extend the buffer.
 *
 * Extend the internal buffer used to store the compressed data by the
 * specified amount.
 *
 * Returns 0 on success or -1 on failure to extend the buffer.  On failure
 * the original buffer still exists at the original size.
 */
static int
xmlZMemBuffExtend( xmlZMemBuffPtr buff, size_t ext_amt ) {

    int			rc = -1;
    size_t		new_size;
    size_t		cur_used;

    unsigned char *	tmp_ptr = NULL;

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

    else if ( ext_amt == 0 )
	return ( 0 );

    cur_used = buff->zctrl.next_out - buff->zbuff;
    new_size = buff->size + ext_amt;

#ifdef DEBUG_HTTP
1565
    if ( cur_used > new_size )
1566 1567 1568
	xmlGenericError( xmlGenericErrorContext,
			"xmlZMemBuffExtend:  %s\n%s %d bytes.\n",
			"Buffer overwrite detected during compressed memory",
1569
			"buffer extension.  Overflowed by",
1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581
			(cur_used - new_size ) );
#endif

    tmp_ptr = xmlRealloc( buff->zbuff, new_size );
    if ( tmp_ptr != NULL ) {
	rc = 0;
	buff->size  = new_size;
	buff->zbuff = tmp_ptr;
	buff->zctrl.next_out  = tmp_ptr + cur_used;
	buff->zctrl.avail_out = new_size - cur_used;
    }
    else {
1582 1583
	xmlChar msg[500];
	xmlStrPrintf(msg, 500,
1584
		    "xmlZMemBuffExtend:  %s %lu bytes.\n",
1585
		    "Allocation failure extending output buffer to",
1586
		    (unsigned long) new_size );
1587
	xmlIOErr(XML_IO_WRITE, (const char *) msg);
1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627
    }

    return ( rc );
}

/**
 * xmlZMemBuffAppend
 * @buff:  Buffer used to compress and consolidate data
 * @src:   Uncompressed source content to append to buffer
 * @len:   Length of source data to append to buffer
 *
 * Compress and append data to the internal buffer.  The data buffer
 * will be expanded if needed to store the additional data.
 *
 * Returns the number of bytes appended to the buffer or -1 on error.
 */
static int
xmlZMemBuffAppend( xmlZMemBuffPtr buff, const char * src, int len ) {

    int		z_err;
    size_t	min_accept;

    if ( ( buff == NULL ) || ( src == NULL ) )
	return ( -1 );

    buff->zctrl.avail_in = len;
    buff->zctrl.next_in  = (unsigned char *)src;
    while ( buff->zctrl.avail_in > 0 ) {
	/*
	**  Extend the buffer prior to deflate call if a reasonable amount
	**  of output buffer space is not available.
	*/
	min_accept = buff->zctrl.avail_in / DFLT_ZLIB_RATIO;
	if ( buff->zctrl.avail_out <= min_accept ) {
	    if ( xmlZMemBuffExtend( buff, buff->size ) == -1 )
		return ( -1 );
	}

	z_err = deflate( &buff->zctrl, Z_NO_FLUSH );
	if ( z_err != Z_OK ) {
Daniel Veillard's avatar