xmllint.c.svn-base

这是一个用于解析xml文件的类库。使用这个类库

/*
 * xmllint.c : a small tester program for XML input.
 *
 * See Copyright for the status of this software.
 *
 * [e-mail]
 */

#include "libxml.h"

#include <string.h>
#include <stdarg.h>
#include <assert.h>

//#if defined (_WIN32) && !defined(__CYGWIN__)
//#if defined (_MSC_VER) || defined(__BORLANDC__)
//#include <winsock2.h>
//#pragma comment(lib, "ws2_32.lib")
//#define gettimeofday(p1,p2)
//#endif /* _MSC_VER */
//#endif /* _WIN32 */

#ifdef HAVE_SYS_TIME_H
#include <sys/time.h>
#endif
#ifdef HAVE_TIME_H
#include <time.h>
#endif

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

#ifdef HAVE_SYS_TIMEB_H
#include <sys/timeb.h>
#endif

#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_SYS_MMAN_H
#include <sys/mman.h>
/* seems needed for Solaris */
#ifndef MAP_FAILED
#define MAP_FAILED ((void *) -1)
#endif
#endif
#ifdef HAVE_STDLIB_H
#include <stdlib.h>
#endif
#ifdef HAVE_LIBREADLINE
#include <readline/readline.h>
#ifdef HAVE_LIBHISTORY
#include <readline/history.h>
#endif
#endif

#include <libxml/xmlmemory.h>
#include <libxml/parser.h>
#include <libxml/parserInternals.h>
#include <libxml/HTMLparser.h>
#include <libxml/HTMLtree.h>
#include <libxml/tree.h>
#include <libxml/xpath.h>
#include <libxml/debugXML.h>
#include <libxml/xmlerror.h>
#ifdef LIBXML_XINCLUDE_ENABLED
#include <libxml/xinclude.h>
#endif
#ifdef LIBXML_CATALOG_ENABLED
#include <libxml/catalog.h>
#endif
#include <libxml/globals.h>
#include <libxml/xmlreader.h>
#ifdef LIBXML_SCHEMAS_ENABLED
#include <libxml/relaxng.h>
#include <libxml/xmlschemas.h>
#endif
#ifdef LIBXML_PATTERN_ENABLED
#include <libxml/pattern.h>
#endif

#ifndef XML_XML_DEFAULT_CATALOG
#define XML_XML_DEFAULT_CATALOG "file:///etc/xml/catalog"
#endif

#ifdef LIBXML_DEBUG_ENABLED
static int shell = 0;
static int debugent = 0;
#endif
static int debug = 0;
static int maxmem = 0;
#ifdef LIBXML_TREE_ENABLED
static int copy = 0;
#endif /* LIBXML_TREE_ENABLED */
static int recovery = 0;
static int noent = 0;
static int noblanks = 0;
static int noout = 0;
static int nowrap = 0;
#ifdef LIBXML_OUTPUT_ENABLED
static int format = 0;
static const char *output = NULL;
static int compress = 0;
#endif /* LIBXML_OUTPUT_ENABLED */
#ifdef LIBXML_VALID_ENABLED
static int valid = 0;
static int postvalid = 0;
static char * dtdvalid = NULL;
static char * dtdvalidfpi = NULL;
#endif
#ifdef LIBXML_SCHEMAS_ENABLED
static char * relaxng = NULL;
static xmlRelaxNGPtr relaxngschemas = NULL;
static char * schema = NULL;
static xmlSchemaPtr wxschemas = NULL;
#endif
static int repeat = 0;
static int insert = 0;
#if defined(LIBXML_HTML_ENABLED) || defined(LIBXML_VALID_ENABLED)
static int html = 0;
static int xmlout = 0;
#endif
static int htmlout = 0;
#ifdef LIBXML_PUSH_ENABLED
static int push = 0;
#endif /* LIBXML_PUSH_ENABLED */
#ifdef HAVE_SYS_MMAN_H
static int memory = 0;
#endif
static int testIO = 0;
static char *encoding = NULL;
#ifdef LIBXML_XINCLUDE_ENABLED
static int xinclude = 0;
#endif
static int dtdattrs = 0;
static int loaddtd = 0;
static int progresult = 0;
static int timing = 0;
static int generate = 0;
static int dropdtd = 0;
#ifdef LIBXML_CATALOG_ENABLED
static int catalogs = 0;
static int nocatalogs = 0;
#endif
#ifdef LIBXML_READER_ENABLED
static int stream = 0;
static int walker = 0;
#endif /* LIBXML_READER_ENABLED */
static int chkregister = 0;
static int nbregister = 0;
#ifdef LIBXML_SAX1_ENABLED
static int sax1 = 0;
#endif /* LIBXML_SAX1_ENABLED */
#ifdef LIBXML_PATTERN_ENABLED
static const char *pattern = NULL;
static xmlPatternPtr patternc = NULL;
#endif
static int options = 0;

/************************************************************************
 * 									*
 * Memory allocation consumption debugging				*
 * 									*
 ************************************************************************/

static void
OOM(void)
{
    fprintf(stderr, "Ran out of memory needs > %d bytes\n", maxmem);
    progresult = 9;
}

static void
myFreeFunc(void *mem)
{
    xmlMemFree(mem);
}
static void *
myMallocFunc(size_t size)
{
    void *ret;

    ret = xmlMemMalloc(size);
    if (ret != NULL) {
        if (xmlMemUsed() > maxmem) {
            OOM();
            xmlMemFree(ret);
            return (NULL);
        }
    }
    return (ret);
}
static void *
myReallocFunc(void *mem, size_t size)
{
    void *ret;

    ret = xmlMemRealloc(mem, size);
    if (ret != NULL) {
        if (xmlMemUsed() > maxmem) {
            OOM();
            xmlMemFree(ret);
            return (NULL);
        }
    }
    return (ret);
}
static char *
myStrdupFunc(const char *str)
{
    char *ret;

    ret = xmlMemoryStrdup(str);
    if (ret != NULL) {
        if (xmlMemUsed() > maxmem) {
            OOM();
            xmlFree(ret);
            return (NULL);
        }
    }
    return (ret);
}
/************************************************************************
 * 									*
 * Internal timing routines to remove the necessity to have		*
 * unix-specific function calls.					*
 * 									*
 ************************************************************************/

#ifndef HAVE_GETTIMEOFDAY 
#ifdef HAVE_SYS_TIMEB_H
#ifdef HAVE_SYS_TIME_H
#ifdef HAVE_FTIME

static int
my_gettimeofday(struct timeval *tvp, void *tzp)
{
	struct timeb timebuffer;

	ftime(&timebuffer);
	if (tvp) {
		tvp->tv_sec = timebuffer.time;
		tvp->tv_usec = timebuffer.millitm * 1000L;
	}
	return (0);
}
#define HAVE_GETTIMEOFDAY 1
#define gettimeofday my_gettimeofday

#endif /* HAVE_FTIME */
#endif /* HAVE_SYS_TIME_H */
#endif /* HAVE_SYS_TIMEB_H */
#endif /* !HAVE_GETTIMEOFDAY */

#if defined(HAVE_GETTIMEOFDAY)
static struct timeval begin, end;

/*
 * startTimer: call where you want to start timing
 */
static void
startTimer(void)
{
    gettimeofday(&begin, NULL);
}

/*
 * endTimer: call where you want to stop timing and to print out a
 *           message about the timing performed; format is a printf
 *           type argument
 */
static void
endTimer(const char *fmt, ...)
{
    long msec;
    va_list ap;

    gettimeofday(&end, NULL);
    msec = end.tv_sec - begin.tv_sec;
    msec *= 1000;
    msec += (end.tv_usec - begin.tv_usec) / 1000;

#ifndef HAVE_STDARG_H
#error "endTimer required stdarg functions"
#endif
    va_start(ap, fmt);
    vfprintf(stderr, fmt, ap);
    va_end(ap);

    fprintf(stderr, " took %ld ms\n", msec);
}
#elif defined(HAVE_TIME_H)
/*
 * No gettimeofday function, so we have to make do with calling clock.
 * This is obviously less accurate, but there's little we can do about
 * that.
 */
#ifndef CLOCKS_PER_SEC
#define CLOCKS_PER_SEC 100
#endif

static clock_t begin, end;
static void
startTimer(void)
{
    begin = clock();
}
static void
endTimer(const char *fmt, ...)
{
    long msec;
    va_list ap;

    end = clock();
    msec = ((end - begin) * 1000) / CLOCKS_PER_SEC;

#ifndef HAVE_STDARG_H
#error "endTimer required stdarg functions"
#endif
    va_start(ap, fmt);
    vfprintf(stderr, fmt, ap);
    va_end(ap);
    fprintf(stderr, " took %ld ms\n", msec);
}
#else

/*
 * We don't have a gettimeofday or time.h, so we just don't do timing
 */
static void
startTimer(void)
{
    /*
     * Do nothing
     */
}
static void
endTimer(char *format, ...)
{
    /*
     * We cannot do anything because we don't have a timing function
     */
#ifdef HAVE_STDARG_H
    va_start(ap, format);
    vfprintf(stderr, format, ap);
    va_end(ap);
    fprintf(stderr, " was not timed\n", msec);
#else
    /* We don't have gettimeofday, time or stdarg.h, what crazy world is
     * this ?!
     */
#endif
}
#endif
/************************************************************************
 * 									*
 * 			HTML ouput					*
 * 									*
 ************************************************************************/
char buffer[50000];

static void
xmlHTMLEncodeSend(void) {
    char *result;

    result = (char *) xmlEncodeEntitiesReentrant(NULL, BAD_CAST buffer);
    if (result) {
	xmlGenericError(xmlGenericErrorContext, "%s", result);
	xmlFree(result);
    }
    buffer[0] = 0;
}

/**
 * xmlHTMLPrintFileInfo:
 * @input:  an xmlParserInputPtr input
 * 
 * Displays the associated file and line informations for the current input
 */

static void
xmlHTMLPrintFileInfo(xmlParserInputPtr input) {
    int len;
    xmlGenericError(xmlGenericErrorContext, "<p>");

    len = strlen(buffer);
    if (input != NULL) {
	if (input->filename) {
	    _snprintf(&buffer[len], sizeof(buffer) - len, "%s:%d: ", input->filename,
		    input->line);
	} else {
	    _snprintf(&buffer[len], sizeof(buffer) - len, "Entity: line %d: ", input->line);
	}
    }
    xmlHTMLEncodeSend();
}

/**
 * xmlHTMLPrintFileContext:
 * @input:  an xmlParserInputPtr input
 * 
 * Displays current context within the input content for error tracking
 */

static void
xmlHTMLPrintFileContext(xmlParserInputPtr input) {
    const xmlChar *cur, *base;
    int len;
    int n;

    if (input == NULL) return;
    xmlGenericError(xmlGenericErrorContext, "<pre>\n");
    cur = input->cur;
    base = input->base;
    while ((cur > base) && ((*cur == '\n') || (*cur == '\r'))) {
	cur--;
    }
    n = 0;
    while ((n++ < 80) && (cur > base) && (*cur != '\n') && (*cur != '\r'))
        cur--;
    if ((*cur == '\n') || (*cur == '\r')) cur++;
    base = cur;
    n = 0;
    while ((*cur != 0) && (*cur != '\n') && (*cur != '\r') && (n < 79)) {
	len = strlen(buffer);
        _snprintf(&buffer[len], sizeof(buffer) - len, "%c", 
		    (unsigned char) *cur++);
	n++;
    }
    len = strlen(buffer);
    _snprintf(&buffer[len], sizeof(buffer) - len, "\n");
    cur = input->cur;
    while ((*cur == '\n') || (*cur == '\r'))
	cur--;
    n = 0;
    while ((cur != base) && (n++ < 80)) {
	len = strlen(buffer);
        _snprintf(&buffer[len], sizeof(buffer) - len, " ");
        base++;
    }
    len = strlen(buffer);
    _snprintf(&buffer[len], sizeof(buffer) - len, "^\n");
    xmlHTMLEncodeSend();
    xmlGenericError(xmlGenericErrorContext, "</pre>");
}

/**
 * xmlHTMLError:
 * @ctx:  an XML parser context
 * @msg:  the message to display/transmit
 * @...:  extra parameters for the message display
 * 
 * Display and format an error messages, gives file, line, position and
 * extra parameters.
 */
static void
xmlHTMLError(void *ctx, const char *msg, ...)
{
    xmlParserCtxtPtr ctxt = (xmlParserCtxtPtr) ctx;
    xmlParserInputPtr input;
    va_list args;
    int len;

    buffer[0] = 0;
    input = ctxt->input;
    if ((input != NULL) && (input->filename == NULL) && (ctxt->inputNr > 1)) {
        input = ctxt->inputTab[ctxt->inputNr - 2];
    }
        
    xmlHTMLPrintFileInfo(input);

    xmlGenericError(xmlGenericErrorContext, "<b>error</b>: ");
    va_start(args, msg);
    len = strlen(buffer);
    _vsnprintf(&buffer[len],  sizeof(buffer) - len, msg, args);
    va_end(args);
    xmlHTMLEncodeSend();
    xmlGenericError(xmlGenericErrorContext, "</p>\n");

    xmlHTMLPrintFileContext(input);
    xmlHTMLEncodeSend();
}

/**
 * xmlHTMLWarning:
 * @ctx:  an XML parser context
 * @msg:  the message to display/transmit
 * @...:  extra parameters for the message display
 * 
 * Display and format a warning messages, gives file, line, position and
 * extra parameters.
 */
static void
xmlHTMLWarning(void *ctx, const char *msg, ...)
{
    xmlParserCtxtPtr ctxt = (xmlParserCtxtPtr) ctx;
    xmlParserInputPtr input;
    va_list args;
    int len;

    buffer[0] = 0;
    input = ctxt->input;
    if ((input != NULL) && (input->filename == NULL) && (ctxt->inputNr > 1)) {
        input = ctxt->inputTab[ctxt->inputNr - 2];
    }
        

    xmlHTMLPrintFileInfo(input);
        
    xmlGenericError(xmlGenericErrorContext, "<b>warning</b>: ");
    va_start(args, msg);
    len = strlen(buffer);    
    _vsnprintf(&buffer[len],  sizeof(buffer) - len, msg, args);