00001 /* ==================================================================== 00002 * The Kannel Software License, Version 1.0 00003 * 00004 * Copyright (c) 2001-2008 Kannel Group 00005 * Copyright (c) 1998-2001 WapIT Ltd. 00006 * All rights reserved. 00007 * 00008 * Redistribution and use in source and binary forms, with or without 00009 * modification, are permitted provided that the following conditions 00010 * are met: 00011 * 00012 * 1. Redistributions of source code must retain the above copyright 00013 * notice, this list of conditions and the following disclaimer. 00014 * 00015 * 2. Redistributions in binary form must reproduce the above copyright 00016 * notice, this list of conditions and the following disclaimer in 00017 * the documentation and/or other materials provided with the 00018 * distribution. 00019 * 00020 * 3. The end-user documentation included with the redistribution, 00021 * if any, must include the following acknowledgment: 00022 * "This product includes software developed by the 00023 * Kannel Group (http://www.kannel.org/)." 00024 * Alternately, this acknowledgment may appear in the software itself, 00025 * if and wherever such third-party acknowledgments normally appear. 00026 * 00027 * 4. The names "Kannel" and "Kannel Group" must not be used to 00028 * endorse or promote products derived from this software without 00029 * prior written permission. For written permission, please 00030 * contact org@kannel.org. 00031 * 00032 * 5. Products derived from this software may not be called "Kannel", 00033 * nor may "Kannel" appear in their name, without prior written 00034 * permission of the Kannel Group. 00035 * 00036 * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED 00037 * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES 00038 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE 00039 * DISCLAIMED. IN NO EVENT SHALL THE KANNEL GROUP OR ITS CONTRIBUTORS 00040 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, 00041 * OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT 00042 * OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR 00043 * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, 00044 * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE 00045 * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, 00046 * EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 00047 * ==================================================================== 00048 * 00049 * This software consists of voluntary contributions made by many 00050 * individuals on behalf of the Kannel Group. For more information on 00051 * the Kannel Group, please see <http://www.kannel.org/>. 00052 * 00053 * Portions of this software are based upon software originally written at 00054 * WapIT Ltd., Helsinki, Finland for the Kannel project. 00055 */ 00056 00057 /* 00058 * xml_shared.h - Common xml tokenizer interface 00059 * This file contains mainly character set functions and binary manipulating 00060 * functions used with a binary without a string table. 00061 * 00062 * Tuomas Luttinen for Wapit Ltd and Aarno Syvänen for Wiral Ltd. 00063 */ 00064 00065 #ifndef XML_SHARED_H 00066 #define XML_SHARED_H 00067 00068 /* 00069 * Charset type is used by WML, SI and SL. 00070 */ 00071 typedef struct charset_t charset_t; 00072 00073 /* 00074 * XML binary type not containing a string table. This is used for SI and SL. 00075 */ 00076 00077 typedef struct simple_binary_t simple_binary_t; 00078 00079 #include "gwlib/gwlib.h" 00080 00081 /* 00082 * XML binary type not containing a string table. This is used for SI and SL. 00083 */ 00084 struct simple_binary_t { 00085 unsigned char wbxml_version; 00086 unsigned char public_id; 00087 unsigned long charset; 00088 Octstr *binary; 00089 int code_page; 00090 }; 00091 00092 /* 00093 * Prototypes of common functions. First functions common with wml, si and sl 00094 * compilers. 00095 * 00096 * set_charset - set the charset of the http headers into the document, if 00097 * it has no encoding set. 00098 */ 00099 void set_charset(Octstr *document, Octstr *charset); 00100 00101 /* 00102 * find_charset_encoding -- parses for a encoding argument within 00103 * the xml preabmle, ie. <?xml verion="xxx" encoding="ISO-8859-1"?> 00104 */ 00105 Octstr *find_charset_encoding(Octstr *document); 00106 00107 /* 00108 * element_check_content - a helper function for checking if an element has 00109 * content or attributes. Returns status bit for attributes (0x80) and another 00110 * for content (0x40) added into one octet. 00111 */ 00112 unsigned char element_check_content(xmlNodePtr node); 00113 00114 /* 00115 * only_blanks - checks if a text node contains only white space, when it can 00116 * be left out as a element content. 00117 */ 00118 00119 int only_blanks(const char *text); 00120 00121 /* 00122 * Parses the character set of the document given as Octstr. Returns the 00123 * MIBenum value. If the charset is not found, we default to UTF-8 value. 00124 */ 00125 int parse_charset(Octstr *os); 00126 00127 /* 00128 * Return the character sets supported by the WML compiler, as a List 00129 * of Octstrs, where each string is the MIME identifier for one charset. 00130 */ 00131 List *wml_charsets(void); 00132 00133 /* 00134 * Macro for creating an octet string from a node content. This has two 00135 * versions for different libxml node content implementation methods. 00136 */ 00137 00138 #ifdef XML_USE_BUFFER_CONTENT 00139 #define create_octstr_from_node(node) (octstr_create(node->content->content)) 00140 #else 00141 #define create_octstr_from_node(node) (octstr_create(node->content)) 00142 #endif 00143 00144 #endif 00145 00146 /* 00147 * Functions working with simple binary type (no string table) 00148 */ 00149 00150 simple_binary_t *simple_binary_create(void); 00151 00152 void simple_binary_destroy(simple_binary_t *bxml); 00153 00154 /* 00155 * Output the sibxml content field after field into octet string os. We add 00156 * string table length 0 (no string table) before the content. 00157 */ 00158 void simple_binary_output(Octstr *os, simple_binary_t *bxml); 00159 00160 void parse_end(simple_binary_t **bxml); 00161 00162 void output_char(int byte, simple_binary_t **bxml); 00163 00164 void parse_octet_string(Octstr *os, simple_binary_t **bxml); 00165 00166 /* 00167 * Add global tokens to the start and to the end of an inline string. 00168 */ 00169 void parse_inline_string(Octstr *temp, simple_binary_t **bxml); 00170 00171 void output_octet_string(Octstr *os, simple_binary_t **bxml); 00172 00173 00174 00175