xml.c

php-4.4.7学习linux时下载的源代码

/*
   +----------------------------------------------------------------------+
   | PHP Version 4                                                        |
   +----------------------------------------------------------------------+
   | Copyright (c) 1997-2007 The PHP Group                                |
   +----------------------------------------------------------------------+
   | This source file is subject to version 3.01 of the PHP license,      |
   | that is bundled with this package in the file LICENSE, and is        |
   | available through the world-wide-web at the following url:           |
   | http://www.php.net/license/3_01.txt                                  |
   | If you did not receive a copy of the PHP license and are unable to   |
   | obtain it through the world-wide-web, please send a note to          |
   | [e-mail] so we can mail you a copy immediately.               |
   +----------------------------------------------------------------------+
   | Authors: Stig S鎡her Bakken <[e-mail]>                            |
   |          Thies C. Arntzen <[e-mail]>                         | 
   +----------------------------------------------------------------------+
 */

/* $Id: xml.c,v 1.110.2.5.2.2 2007/01/01 09:46:49 sebastian Exp $ */

#define IS_EXT_MODULE

#ifdef HAVE_CONFIG_H
#include "config.h"
#endif

#include "php.h"

#define PHP_XML_INTERNAL
#include "php_xml.h"
#include "zend_variables.h"
#include "ext/standard/php_string.h"
#include "ext/standard/info.h"

#if HAVE_LIBEXPAT

# include "ext/standard/head.h"

/* Short-term TODO list:
 * - Implement XML_ExternalEntityParserCreate()
 * - XML_SetCommentHandler
 * - XML_SetCdataSectionHandler
 * - XML_SetParamEntityParsing
 */

/* Long-term TODO list:
 * - Fix the expat library so you can install your own memory manager
 *   functions
 */

/* Known bugs:
 * - Weird things happen with <![CDATA[]]> sections.
 */

#ifdef ZTS
int xml_globals_id;
#else
PHP_XML_API php_xml_globals xml_globals;
#endif

/* {{{ dynamically loadable module stuff */
#ifdef COMPILE_DL_XML
ZEND_GET_MODULE(xml)
#endif /* COMPILE_DL_XML */
/* }}} */

/* {{{ function prototypes */
PHP_MINIT_FUNCTION(xml);
PHP_RINIT_FUNCTION(xml);
PHP_MSHUTDOWN_FUNCTION(xml);
PHP_RSHUTDOWN_FUNCTION(xml);
PHP_MINFO_FUNCTION(xml);

static void xml_parser_dtor(zend_rsrc_list_entry *rsrc TSRMLS_DC);
static void xml_set_handler(zval **, zval **);
inline static unsigned short xml_encode_iso_8859_1(unsigned char);
inline static char xml_decode_iso_8859_1(unsigned short);
inline static unsigned short xml_encode_us_ascii(unsigned char);
inline static char xml_decode_us_ascii(unsigned short);
static XML_Char *xml_utf8_encode(const char *, int, int *, const XML_Char *);
static zval *xml_call_handler(xml_parser *, zval *, int, zval **);
static zval *_xml_xmlchar_zval(const XML_Char *, int, const XML_Char *);
static int _xml_xmlcharlen(const XML_Char *);
static void _xml_add_to_info(xml_parser *parser,char *name);
inline static char *_xml_decode_tag(xml_parser *parser, const char *tag);

void _xml_startElementHandler(void *, const char *, const char **);
void _xml_endElementHandler(void *, const char *);
void _xml_characterDataHandler(void *, const XML_Char *, int);
void _xml_processingInstructionHandler(void *, const XML_Char *, const XML_Char *);
void _xml_defaultHandler(void *, const XML_Char *, int);
void _xml_unparsedEntityDeclHandler(void *, const XML_Char *, const XML_Char *, const XML_Char *, const XML_Char *, const XML_Char *);
void _xml_notationDeclHandler(void *, const XML_Char *, const XML_Char *, const XML_Char *, const XML_Char *);
int  _xml_externalEntityRefHandler(XML_Parser, const XML_Char *, const XML_Char *, const XML_Char *, const XML_Char *);

void _xml_startNamespaceDeclHandler(void *, const XML_Char *, const XML_Char *);
void _xml_endNamespaceDeclHandler(void *, const XML_Char *);
/* }}} */

/* {{{ extension definition structures */
static unsigned char third_and_fourth_args_force_ref[] = { 4, BYREF_NONE, BYREF_NONE, BYREF_FORCE, BYREF_FORCE };

function_entry xml_functions[] = {
	PHP_FE(xml_parser_create, NULL)
	PHP_FE(xml_parser_create_ns, NULL)
	PHP_FE(xml_set_object, second_arg_force_ref)
	PHP_FE(xml_set_element_handler, NULL)
	PHP_FE(xml_set_character_data_handler, NULL)
	PHP_FE(xml_set_processing_instruction_handler, NULL)
	PHP_FE(xml_set_default_handler, NULL)
	PHP_FE(xml_set_unparsed_entity_decl_handler, NULL)
	PHP_FE(xml_set_notation_decl_handler, NULL)
	PHP_FE(xml_set_external_entity_ref_handler, NULL)
	PHP_FE(xml_set_start_namespace_decl_handler, NULL)
	PHP_FE(xml_set_end_namespace_decl_handler, NULL)
	PHP_FE(xml_parse, NULL)
	PHP_FE(xml_parse_into_struct, third_and_fourth_args_force_ref)
	PHP_FE(xml_get_error_code, NULL)
	PHP_FE(xml_error_string, NULL)
	PHP_FE(xml_get_current_line_number, NULL)
	PHP_FE(xml_get_current_column_number, NULL)
	PHP_FE(xml_get_current_byte_index, NULL)
	PHP_FE(xml_parser_free, NULL)
	PHP_FE(xml_parser_set_option, NULL)
	PHP_FE(xml_parser_get_option, NULL)
	PHP_FE(utf8_encode, NULL)
	PHP_FE(utf8_decode, NULL)
	{NULL, NULL, NULL}
};

zend_module_entry xml_module_entry = {
    STANDARD_MODULE_HEADER,
	"xml",                /* extension name */
	xml_functions,        /* extension function list */
	PHP_MINIT(xml),       /* extension-wide startup function */
	PHP_MSHUTDOWN(xml),   /* extension-wide shutdown function */
	PHP_RINIT(xml),       /* per-request startup function */
	PHP_RSHUTDOWN(xml),   /* per-request shutdown function */
	PHP_MINFO(xml),       /* information function */
    NO_VERSION_YET,
	STANDARD_MODULE_PROPERTIES
};

/* All the encoding functions are set to NULL right now, since all
 * the encoding is currently done internally by expat/xmltok.
 */
xml_encoding xml_encodings[] = {
	{ "ISO-8859-1", xml_decode_iso_8859_1, xml_encode_iso_8859_1 },
	{ "US-ASCII",   xml_decode_us_ascii,   xml_encode_us_ascii   },
	{ "UTF-8",      NULL,                  NULL                  },
	{ NULL,         NULL,                  NULL                  }
};

/* True globals, no need for thread safety */
static int le_xml_parser; 

/* }}} */

/* {{{ startup, shutdown and info functions */
#ifdef ZTS
static void php_xml_init_globals(php_xml_globals *xml_globals_p TSRMLS_DC)
{
	XML(default_encoding) = "ISO-8859-1";
}
#endif

PHP_MINIT_FUNCTION(xml)
{
	le_xml_parser =	zend_register_list_destructors_ex(xml_parser_dtor, NULL, "xml", module_number);

#ifdef ZTS
	ts_allocate_id(&xml_globals_id, sizeof(php_xml_globals), (ts_allocate_ctor) php_xml_init_globals, NULL);
#else
	XML(default_encoding) = "ISO-8859-1";
#endif

	REGISTER_LONG_CONSTANT("XML_ERROR_NONE", XML_ERROR_NONE, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_ERROR_NO_MEMORY", XML_ERROR_NO_MEMORY, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_ERROR_SYNTAX", XML_ERROR_SYNTAX, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_ERROR_NO_ELEMENTS", XML_ERROR_NO_ELEMENTS, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_ERROR_INVALID_TOKEN", XML_ERROR_INVALID_TOKEN, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_ERROR_UNCLOSED_TOKEN", XML_ERROR_UNCLOSED_TOKEN, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_ERROR_PARTIAL_CHAR", XML_ERROR_PARTIAL_CHAR, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_ERROR_TAG_MISMATCH", XML_ERROR_TAG_MISMATCH, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_ERROR_DUPLICATE_ATTRIBUTE", XML_ERROR_DUPLICATE_ATTRIBUTE, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_ERROR_JUNK_AFTER_DOC_ELEMENT", XML_ERROR_JUNK_AFTER_DOC_ELEMENT, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_ERROR_PARAM_ENTITY_REF", XML_ERROR_PARAM_ENTITY_REF, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_ERROR_UNDEFINED_ENTITY", XML_ERROR_UNDEFINED_ENTITY, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_ERROR_RECURSIVE_ENTITY_REF", XML_ERROR_RECURSIVE_ENTITY_REF, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_ERROR_ASYNC_ENTITY", XML_ERROR_ASYNC_ENTITY, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_ERROR_BAD_CHAR_REF", XML_ERROR_BAD_CHAR_REF, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_ERROR_BINARY_ENTITY_REF", XML_ERROR_BINARY_ENTITY_REF, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_ERROR_ATTRIBUTE_EXTERNAL_ENTITY_REF", XML_ERROR_ATTRIBUTE_EXTERNAL_ENTITY_REF, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_ERROR_MISPLACED_XML_PI", XML_ERROR_MISPLACED_XML_PI, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_ERROR_UNKNOWN_ENCODING", XML_ERROR_UNKNOWN_ENCODING, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_ERROR_INCORRECT_ENCODING", XML_ERROR_INCORRECT_ENCODING, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_ERROR_UNCLOSED_CDATA_SECTION", XML_ERROR_UNCLOSED_CDATA_SECTION, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_ERROR_EXTERNAL_ENTITY_HANDLING", XML_ERROR_EXTERNAL_ENTITY_HANDLING, CONST_CS|CONST_PERSISTENT);

	REGISTER_LONG_CONSTANT("XML_OPTION_CASE_FOLDING", PHP_XML_OPTION_CASE_FOLDING, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_OPTION_TARGET_ENCODING", PHP_XML_OPTION_TARGET_ENCODING, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_OPTION_SKIP_TAGSTART", PHP_XML_OPTION_SKIP_TAGSTART, CONST_CS|CONST_PERSISTENT);
	REGISTER_LONG_CONSTANT("XML_OPTION_SKIP_WHITE", PHP_XML_OPTION_SKIP_WHITE, CONST_CS|CONST_PERSISTENT);
	
	return SUCCESS;
}


PHP_RINIT_FUNCTION(xml)
{
    return SUCCESS;
}


PHP_MSHUTDOWN_FUNCTION(xml)
{
	return SUCCESS;
}


PHP_RSHUTDOWN_FUNCTION(xml)
{
    return SUCCESS;
}


PHP_MINFO_FUNCTION(xml)
{
	php_info_print_table_start();
	php_info_print_table_row(2, "XML Support", "active");
	php_info_print_table_row(2, "XML Namespace Support", "active");
	php_info_print_table_row(2, "EXPAT Version",XML_ExpatVersion());
	php_info_print_table_end();
}
/* }}} */

/* {{{ extension-internal functions */
static zval *_xml_resource_zval(long value)
{
	zval *ret;
	TSRMLS_FETCH();

	MAKE_STD_ZVAL(ret);

	Z_TYPE_P(ret) = IS_RESOURCE;
	Z_LVAL_P(ret) = value;

	zend_list_addref(value);

	return ret;
}

static zval *_xml_string_zval(const char *str)
{
	zval *ret;
	int len = strlen(str);
	MAKE_STD_ZVAL(ret);

	Z_TYPE_P(ret) = IS_STRING;
	Z_STRLEN_P(ret) = len;
	Z_STRVAL_P(ret) = estrndup(str, len);
	return ret;
}

static zval *_xml_xmlchar_zval(const XML_Char *s, int len, const XML_Char *encoding)
{
	zval *ret;
	MAKE_STD_ZVAL(ret);
	
	if (s == NULL) {
		ZVAL_FALSE(ret);
		return ret;
	}
	if (len == 0) {
		len = _xml_xmlcharlen(s);
	}
	Z_TYPE_P(ret) = IS_STRING;
	Z_STRVAL_P(ret) = xml_utf8_decode(s, len, &Z_STRLEN_P(ret), encoding);
	return ret;
}
/* }}} */

/* {{{ xml_parser_dtor() */
static void xml_parser_dtor(zend_rsrc_list_entry *rsrc TSRMLS_DC)
{
	xml_parser *parser = (xml_parser *)rsrc->ptr;

	/* please leave this commented - or ask [e-mail] before doing it (again) */
#ifdef ZEND_ENGINE_2
	if (parser->object) {
		zval_ptr_dtor(&parser->object);
	}
#endif
	
	if (parser->parser) {
		XML_ParserFree(parser->parser);
	}
	if (parser->ltags) {
		int inx;
		for (inx = 0; inx < parser->level; inx++)
			efree(parser->ltags[ inx ]);
		efree(parser->ltags);
	}
	if (parser->startElementHandler) {
		zval_ptr_dtor(&parser->startElementHandler);
	}
	if (parser->endElementHandler) {
		zval_ptr_dtor(&parser->endElementHandler);
	}
	if (parser->characterDataHandler) {
		zval_ptr_dtor(&parser->characterDataHandler);
	}
	if (parser->processingInstructionHandler) {
		zval_ptr_dtor(&parser->processingInstructionHandler);
	}
	if (parser->defaultHandler) {
		zval_ptr_dtor(&parser->defaultHandler);
	}
	if (parser->unparsedEntityDeclHandler) {
		zval_ptr_dtor(&parser->unparsedEntityDeclHandler);
	}
	if (parser->notationDeclHandler) {
		zval_ptr_dtor(&parser->notationDeclHandler);
	}
	if (parser->externalEntityRefHandler) {
		zval_ptr_dtor(&parser->externalEntityRefHandler);
	}
	if (parser->unknownEncodingHandler) {
		zval_ptr_dtor(&parser->unknownEncodingHandler);
	}
	if (parser->startNamespaceDeclHandler) {
		zval_ptr_dtor(&parser->startNamespaceDeclHandler);
	}
	if (parser->endNamespaceDeclHandler) {
		zval_ptr_dtor(&parser->endNamespaceDeclHandler);
	}
	if (parser->baseURI) {
		efree(parser->baseURI);
	}

	efree(parser);
}
/* }}} */

/* {{{ xml_set_handler() */
static void xml_set_handler(zval **handler, zval **data)
{
	/* IS_ARRAY might indicate that we're using array($obj, 'method') syntax */
	if (Z_TYPE_PP(data) != IS_ARRAY) {
		convert_to_string_ex(data);
	}

	/* If we have already a handler, release it */
	if (*handler) {
		zval_ptr_dtor(handler);
	}

	zval_add_ref(data);
	*handler = *data;
}
/* }}} */

/* {{{ xml_call_handler() */
static zval *xml_call_handler(xml_parser *parser, zval *handler, int argc, zval **argv)
{
	TSRMLS_FETCH();

	if (parser && handler) {
		zval *retval;
		int i;	
		int result;

		MAKE_STD_ZVAL(retval);
		ZVAL_FALSE(retval);

		result = call_user_function(EG(function_table), &parser->object, handler, retval, argc, argv TSRMLS_CC);

		if (result == FAILURE) {
			zval **method;
			zval **obj;

			if (Z_TYPE_P(handler) == IS_STRING) {
				php_error_docref(NULL TSRMLS_CC, E_WARNING, "Unable to call handler %s()", Z_STRVAL_P(handler));
			} else if (zend_hash_index_find(Z_ARRVAL_P(handler), 0, (void **) &obj) == SUCCESS &&
					   zend_hash_index_find(Z_ARRVAL_P(handler), 1, (void **) &method) == SUCCESS &&
					   Z_TYPE_PP(obj) == IS_OBJECT &&
					   Z_TYPE_PP(method) == IS_STRING) {
				php_error_docref(NULL TSRMLS_CC, E_WARNING, "Unable to call handler %s::%s()", Z_OBJCE_PP(obj)->name, Z_STRVAL_PP(method));
			} else 
				php_error_docref(NULL TSRMLS_CC, E_WARNING, "Unable to call handler");

			zval_dtor(retval);
			efree(retval);
		}

		for (i = 0; i < argc; i++) {
			zval_ptr_dtor(&(argv[i]));
		}

		if (result == FAILURE) {
			return NULL;
		} else {
			return retval;
		}
	}
	return NULL;