1 /*
2    neon XML parser interface
3    Copyright (C) 1999-2007, Joe Orton <joe@manyfish.co.uk>
4 
5    This library is free software; you can redistribute it and/or
6    modify it under the terms of the GNU Library General Public
7    License as published by the Free Software Foundation; either
8    version 2 of the License, or (at your option) any later version.
9 
10    This library is distributed in the hope that it will be useful,
11    but WITHOUT ANY WARRANTY; without even the implied warranty of
12    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
13    Library General Public License for more details.
14 
15    You should have received a copy of the GNU Library General Public
16    License along with this library; if not, write to the Free
17    Software Foundation, Inc., 59 Temple Place - Suite 330, Boston,
18    MA 02111-1307, USA
19 
20 */
21 
22 #ifndef NE_XML_H
23 #define NE_XML_H
24 
25 #include <sys/types.h> /* for size_t */
26 
27 #include "ne_defs.h"
28 
29 NE_BEGIN_DECLS
30 
31 /* The neon XML interface filters a streamed XML tree through a stack
32  * of SAX "handlers".  A handler is made up of three callbacks
33  * (start-element, char-data, end-element).  Each start-element event
34  * is passed to each handler in the stack in turn until one until one
35  * accepts the element.  This handler then receives subsequent
36  * char-data and end-element events for the element.
37  *
38  * For each new start-element event, the search up the handler stack
39  * begins with the handler for the parent element (for the root
40  * element, at the base of the stack).
41  *
42  * For each accepted element, a "state" integer is stored, which is
43  * passed to the corresponding char-data and end-element callbacks for
44  * the element.  This integer is also passed to the start-element
45  * callback of child elements so they can determine context.
46  *
47  * If no handler in the stack accepts a particular element, it (and
48  * its children, if any) is ignored. */
49 
50 #define NE_XML_DECLINE (0)
51 #define NE_XML_ABORT (-1)
52 
53 /* A start-element callback for element with given namespace/name.
54  * The callback may return:
55  *   <0  =>  abort the parse (NE_XML_ABORT)
56  *    0  =>  decline this element (NE_XML_DECLINE)
57  *   >0  =>  accept this element; value is state for this element.
58  *
59  * The 'parent' integer is the state returned by the handler of the
60  * parent element.   The attributes array gives name/value pairs
61  * in atts[n] and atts[n+1] from n=0 up to atts[n]==NULL. */
62 typedef int ne_xml_startelm_cb(void *userdata, int parent,
63                                const char *nspace, const char *name,
64                                const char **atts);
65 
66 /* state for the root element */
67 #define NE_XML_STATEROOT (0)
68 
69 /* Character data callback; may return non-zero to abort the parse. */
70 typedef int ne_xml_cdata_cb(void *userdata, int state,
71                             const char *cdata, size_t len);
72 /* End element callback; may return non-zero to abort the parse. */
73 typedef int ne_xml_endelm_cb(void *userdata, int state,
74                              const char *nspace, const char *name);
75 
76 typedef struct ne_xml_parser_s ne_xml_parser;
77 
78 /* Create an XML parser. */
79 ne_xml_parser *ne_xml_create(void);
80 
81 /* Push a new handler on the stack of parser 'p'. 'cdata' and/or
82  * 'endelm' may be NULL; startelm must be non-NULL. */
83 void ne_xml_push_handler(ne_xml_parser *p,
84                          ne_xml_startelm_cb *startelm,
85                          ne_xml_cdata_cb *cdata,
86                          ne_xml_endelm_cb *endelm,
87                          void *userdata);
88 
89 /* ne_xml_failed returns non-zero if there was an error during
90  * parsing, or zero if the parse completed successfully.  The return
91  * value is equal to that of the last ne_xml_parse() call for this
92  * parser object. */
93 int ne_xml_failed(ne_xml_parser *p);
94 
95 /* Set error string for parser.  (The string may be truncated
96  * internally). */
97 void ne_xml_set_error(ne_xml_parser *p, const char *msg);
98 
99 /* Return the error string (and never NULL).  After ne_xml_failed
100  * returns >0, this will describe the parse error.  Otherwise it will
101  * be a default error string. */
102 const char *ne_xml_get_error(ne_xml_parser *p);
103 
104 /* Parse the given block of input of length len.  Parser must be
105  * called with len=0 to signify the end of the document (for that
106  * case, the block argument is ignored).  Returns zero on success, or
107  * non-zero on error: for an XML syntax error, a positive number is
108  * returned; if parsing is aborted by a caller-supplied callback, that
109  * callback's return value is returned. */
110 int ne_xml_parse(ne_xml_parser *p, const char *block, size_t len);
111 
112 /* As ne_xml_parse, casting (ne_xml_parser *)userdata internally.
113  * (This function can be passed to ne_add_response_body_reader) */
114 int ne_xml_parse_v(void *userdata, const char *block, size_t len);
115 
116 /* Return current line of document during parsing or after parsing is
117  * complete. */
118 int ne_xml_currentline(ne_xml_parser *p);
119 
120 /* From a start_element callback which was passed 'attrs' using given
121  * parser, return attribute of given name and namespace.  If nspace is
122  * NULL, no namespace resolution is performed.  Note that this call is
123  * context-specific; if called outside a start_element callback,
124  * behaviour is undefined. */
125 const char *ne_xml_get_attr(ne_xml_parser *parser,
126 			    const char **attrs, const char *nspace,
127 			    const char *name);
128 
129 /* From a start_element callback, resolve a given XML Namespace
130  * prefix, if defined.  Given a non-NULL prefix, returns the namespace
131  * URI which corresponds to the prefix 'prefix' (of length 'length'),
132  * or NULL if no such namespace prefix is defined.  Given a NULL
133  * prefix, returns the default namespace URI or the empty string if
134  * none is defined.  Note that this call is context-specific; if
135  * called outside a start_element callback, behaviour is undefined. */
136 const char *ne_xml_resolve_nspace(ne_xml_parser *parser,
137                                   const char *prefix, size_t length);
138 
139 /* Return the encoding of the document being parsed.  May return NULL
140  * if no encoding is defined or if the XML declaration has not yet
141  * been parsed. */
142 const char *ne_xml_doc_encoding(const ne_xml_parser *p);
143 
144 /* Destroy the parser object. */
145 void ne_xml_destroy(ne_xml_parser *p);
146 
147 /* A utility interface for mapping {nspace, name} onto an integer. */
148 struct ne_xml_idmap {
149     const char *nspace, *name;
150     int id;
151 };
152 
153 /* Return the size of an idmap array */
154 #define NE_XML_MAPLEN(map) (sizeof(map) / sizeof(struct ne_xml_idmap))
155 
156 /* Return the 'id' corresponding to {nspace, name}, or zero. */
157 int ne_xml_mapid(const struct ne_xml_idmap map[], size_t maplen,
158                  const char *nspace, const char *name);
159 
160 /* media type, appropriate for adding to a Content-Type header */
161 #define NE_XML_MEDIA_TYPE "application/xml"
162 
163 NE_END_DECLS
164 
165 #endif /* NE_XML_H */
166