runtests.c

这是symbian下的实现XML解析的源代码。里面包括解析器和解析例子代码。

/* Copyright (c) 1998, 1999, 2000 Thai Open Source Software Center Ltd
   See the file COPYING for copying permission.

   runtest.c : run the Expat test suite
*/

#ifdef HAVE_EXPAT_CONFIG_H
#include <expat_config.h>
#endif

#ifdef HAVE_CHECK_H
#include <check.h>
#else
#include "minicheck.h"
#endif

#include <assert.h>
#include <stdlib.h>
#include <stdio.h>
#include <string.h>

#include "expat.h"
#include "chardata.h"


static XML_Parser parser;


static void
basic_setup(void)
{
    parser = XML_ParserCreate(NULL);
    if (parser == NULL)
        fail("Parser not created.");
}

static void
basic_teardown(void)
{
    if (parser != NULL)
        XML_ParserFree(parser);
}

/* Generate a failure using the parser state to create an error message;
   this should be used when the parser reports an error we weren't
   expecting.
*/
static void
_xml_failure(XML_Parser parser, const char *file, int line)
{
    char buffer[1024];
    sprintf(buffer,
            "\n    %s (line %d, offset %d)\n    reported from %s, line %d",
            XML_ErrorString(XML_GetErrorCode(parser)),
            XML_GetCurrentLineNumber(parser),
            XML_GetCurrentColumnNumber(parser),
            file, line);
    _fail_unless(0, file, line, buffer);
}

#define xml_failure(parser) _xml_failure((parser), __FILE__, __LINE__)

static void
_expect_failure(char *text, enum XML_Error errorCode, char *errorMessage,
                char *file, int lineno)
{
    if (XML_Parse(parser, text, strlen(text), XML_TRUE) == XML_STATUS_OK)
        /* Hackish use of _fail_unless() macro, but let's us report
           the right filename and line number. */
        _fail_unless(0, file, lineno, errorMessage);
    if (XML_GetErrorCode(parser) != errorCode)
        _xml_failure(parser, file, lineno);
}

#define expect_failure(text, errorCode, errorMessage) \
        _expect_failure((text), (errorCode), (errorMessage), \
                        __FILE__, __LINE__)

/* Dummy handlers for when we need to set a handler to tickle a bug,
   but it doesn't need to do anything.
*/

static void XMLCALL
dummy_start_doctype_handler(void           *userData,
                            const XML_Char *doctypeName,
                            const XML_Char *sysid,
                            const XML_Char *pubid,
                            int            has_internal_subset)
{}

static void XMLCALL
dummy_end_doctype_handler(void *userData)
{}

static void XMLCALL
dummy_entity_decl_handler(void           *userData,
                          const XML_Char *entityName,
                          int            is_parameter_entity,
                          const XML_Char *value,
                          int            value_length,
                          const XML_Char *base,
                          const XML_Char *systemId,
                          const XML_Char *publicId,
                          const XML_Char *notationName)
{}

static void XMLCALL
dummy_notation_decl_handler(void *userData,
                            const XML_Char *notationName,
                            const XML_Char *base,
                            const XML_Char *systemId,
                            const XML_Char *publicId)
{}

static void XMLCALL
dummy_element_decl_handler(void *userData,
                           const XML_Char *name,
                           XML_Content *model)
{}

static void XMLCALL
dummy_attlist_decl_handler(void           *userData,
                           const XML_Char *elname,
                           const XML_Char *attname,
                           const XML_Char *att_type,
                           const XML_Char *dflt,
                           int            isrequired)
{}

static void XMLCALL
dummy_comment_handler(void *userData, const XML_Char *data)
{}

static void XMLCALL
dummy_pi_handler(void *userData, const XML_Char *target, const XML_Char *data)
{}

static void XMLCALL
dummy_start_element(void *userData,
                    const XML_Char *name, const XML_Char **atts)
{}


/*
 * Character & encoding tests.
 */

START_TEST(test_nul_byte)
{
    char text[] = "<doc>\0</doc>";

    /* test that a NUL byte (in US-ASCII data) is an error */
    if (XML_Parse(parser, text, sizeof(text) - 1, XML_TRUE) == XML_STATUS_OK)
        fail("Parser did not report error on NUL-byte.");
    if (XML_GetErrorCode(parser) != XML_ERROR_INVALID_TOKEN)
        xml_failure(parser);
}
END_TEST


START_TEST(test_u0000_char)
{
    /* test that a NUL byte (in US-ASCII data) is an error */
    expect_failure("<doc>&#0;</doc>",
                   XML_ERROR_BAD_CHAR_REF,
                   "Parser did not report error on NUL-byte.");
}
END_TEST

START_TEST(test_bom_utf8)
{
    /* This test is really just making sure we don't core on a UTF-8 BOM. */
    char *text = "\357\273\277<e/>";

    if (XML_Parse(parser, text, strlen(text), XML_TRUE) == XML_STATUS_ERROR)
        xml_failure(parser);
}
END_TEST

START_TEST(test_bom_utf16_be)
{
    char text[] = "\376\377\0<\0e\0/\0>";

    if (XML_Parse(parser, text, sizeof(text)-1, XML_TRUE) == XML_STATUS_ERROR)
        xml_failure(parser);
}
END_TEST

START_TEST(test_bom_utf16_le)
{
    char text[] = "\377\376<\0e\0/\0>\0";

    if (XML_Parse(parser, text, sizeof(text)-1, XML_TRUE) == XML_STATUS_ERROR)
        xml_failure(parser);
}
END_TEST

static void XMLCALL
accumulate_characters(void *userData, const XML_Char *s, int len)
{
    CharData_AppendXMLChars((CharData *)userData, s, len);
}

static void XMLCALL
accumulate_attribute(void *userData, const XML_Char *name,
                     const XML_Char **atts)
{
    CharData *storage = (CharData *)userData;
    if (storage->count < 0 && atts != NULL && atts[0] != NULL) {
        /* "accumulate" the value of the first attribute we see */
        CharData_AppendXMLChars(storage, atts[1], -1);
    }
}


static void
_run_character_check(XML_Char *text, XML_Char *expected,
                     const char *file, int line)
{
    CharData storage;

    CharData_Init(&storage);
    XML_SetUserData(parser, &storage);
    XML_SetCharacterDataHandler(parser, accumulate_characters);
    if (XML_Parse(parser, text, strlen(text), XML_TRUE) == XML_STATUS_ERROR)
        _xml_failure(parser, file, line);
    CharData_CheckXMLChars(&storage, expected);
}

#define run_character_check(text, expected) \
        _run_character_check(text, expected, __FILE__, __LINE__)

static void
_run_attribute_check(XML_Char *text, XML_Char *expected,
                     const char *file, int line)
{
    CharData storage;

    CharData_Init(&storage);
    XML_SetUserData(parser, &storage);
    XML_SetStartElementHandler(parser, accumulate_attribute);
    if (XML_Parse(parser, text, strlen(text), XML_TRUE) == XML_STATUS_ERROR)
        _xml_failure(parser, file, line);
    CharData_CheckXMLChars(&storage, expected);
}

#define run_attribute_check(text, expected) \
        _run_attribute_check(text, expected, __FILE__, __LINE__)

/* Regression test for SF bug #491986. */
START_TEST(test_danish_latin1)
{
    char *text =
        "<?xml version='1.0' encoding='iso-8859-1'?>\n"
        "<e>J鴕gen 骧迤嘏</e>";
    run_character_check(text,
             "J\xC3\xB8rgen \xC3\xA6\xC3\xB8\xC3\xA5\xC3\x86\xC3\x98\xC3\x85");
}
END_TEST


/* Regression test for SF bug #514281. */
START_TEST(test_french_charref_hexidecimal)
{
    char *text =
        "<?xml version='1.0' encoding='iso-8859-1'?>\n"
        "<doc>&#xE9;&#xE8;&#xE0;&#xE7;&#xEA;&#xC8;</doc>";
    run_character_check(text,
                        "\xC3\xA9\xC3\xA8\xC3\xA0\xC3\xA7\xC3\xAA\xC3\x88");
}
END_TEST

START_TEST(test_french_charref_decimal)
{
    char *text =
        "<?xml version='1.0' encoding='iso-8859-1'?>\n"
        "<doc>&#233;&#232;&#224;&#231;&#234;&#200;</doc>";
    run_character_check(text,
                        "\xC3\xA9\xC3\xA8\xC3\xA0\xC3\xA7\xC3\xAA\xC3\x88");
}
END_TEST

START_TEST(test_french_latin1)
{
    char *text =
        "<?xml version='1.0' encoding='iso-8859-1'?>\n"
        "<doc>\xE9\xE8\xE0\xE7\xEa\xC8</doc>";
    run_character_check(text,
                        "\xC3\xA9\xC3\xA8\xC3\xA0\xC3\xA7\xC3\xAA\xC3\x88");
}
END_TEST

START_TEST(test_french_utf8)
{
    char *text =
        "<?xml version='1.0' encoding='utf-8'?>\n"
        "<doc>\xC3\xA9</doc>";
    run_character_check(text, "\xC3\xA9");
}
END_TEST

/* Regression test for SF bug #600479.
   XXX There should be a test that exercises all legal XML Unicode
   characters as PCDATA and attribute value content, and XML Name
   characters as part of element and attribute names.
*/
START_TEST(test_utf8_false_rejection)
{
    char *text = "<doc>\xEF\xBA\xBF</doc>";
    run_character_check(text, "\xEF\xBA\xBF");
}
END_TEST

/* Regression test for SF bug #477667.
   This test assures that any 8-bit character followed by a 7-bit
   character will not be mistakenly interpreted as a valid UTF-8
   sequence.
*/
START_TEST(test_illegal_utf8)
{
    char text[100];
    int i;

    for (i = 128; i <= 255; ++i) {
        sprintf(text, "<e>%ccd</e>", i);
        if (XML_Parse(parser, text, strlen(text), XML_TRUE) == XML_STATUS_OK) {
            sprintf(text,
                    "expected token error for '%c' (ordinal %d) in UTF-8 text",
                    i, i);
            fail(text);
        }
        else if (XML_GetErrorCode(parser) != XML_ERROR_INVALID_TOKEN)
            xml_failure(parser);
        /* Reset the parser since we use the same parser repeatedly. */
        XML_ParserReset(parser, NULL);
    }
}
END_TEST

START_TEST(test_utf16)
{
    /* <?xml version="1.0" encoding="UTF-16"?>
       <doc a='123'>some text</doc>
    */
    char text[] =
        "\000<\000?\000x\000m\000\154\000 \000v\000e\000r\000s\000i\000o"
        "\000n\000=\000'\0001\000.\000\060\000'\000 \000e\000n\000c\000o"
        "\000d\000i\000n\000g\000=\000'\000U\000T\000F\000-\0001\000\066"
        "\000'\000?\000>\000\n"
        "\000<\000d\000o\000c\000 \000a\000=\000'\0001\0002\0003\000'"
        "\000>\000s\000o\000m\000e\000 \000t\000e\000x\000t\000<\000/"
        "\000d\000o\000c\000>";
    if (XML_Parse(parser, text, sizeof(text)-1, XML_TRUE) == XML_STATUS_ERROR)
        xml_failure(parser);
}
END_TEST

START_TEST(test_utf16_le_epilog_newline)
{
    int first_chunk_bytes = 17;
    char text[] = 
        "\xFF\xFE"                      /* BOM */
        "<\000e\000/\000>\000"          /* document element */
        "\r\000\n\000\r\000\n\000";     /* epilog */

    if (first_chunk_bytes >= sizeof(text) - 1)
        fail("bad value of first_chunk_bytes");
    if (  XML_Parse(parser, text, first_chunk_bytes, XML_FALSE)
          == XML_STATUS_ERROR)
        xml_failure(parser);
    else {
        enum XML_Status rc;
        rc = XML_Parse(parser, text + first_chunk_bytes,
                       sizeof(text) - first_chunk_bytes - 1, XML_TRUE);
        if (rc == XML_STATUS_ERROR)
            xml_failure(parser);
    }
}
END_TEST

/* Regression test for SF bug #481609, #774028. */
START_TEST(test_latin1_umlauts)
{
    char *text =
        "<?xml version='1.0' encoding='iso-8859-1'?>\n"
        "<e a='