1 (***********************************************************************)
5 (* Kim Nguyen, LRI UMR8623 *)
6 (* Université Paris-Sud & CNRS *)
8 (* Copyright 2010-2012 Université Paris-Sud and Centre National de la *)
9 (* Recherche Scientifique. All rights reserved. This file is *)
10 (* distributed under the terms of the GNU Lesser General Public *)
11 (* License, with the special exception on linking described in file *)
14 (***********************************************************************)
17 Time-stamp: <Last modified on 2013-02-07 09:59:37 CET by Kim Nguyen>
24 mutable data : string;
25 mutable first_child : node;
26 mutable next_sibling : node;
41 let dummy_tag = QName.of_string "#dummy"
54 (* TODO add other intersting stuff *)
63 mutable stack : node list;
64 text_buffer : Buffer.t;
65 mutable current_preorder : int;
68 let print_node_ptr fmt n =
69 Format.fprintf fmt "<%s>"
70 (if n == nil then "NIL" else
71 if n == dummy then "DUMMY" else
72 "NODE " ^ string_of_int n.preorder)
74 let debug_node fmt node =
75 Format.fprintf fmt "{ tag=%s; preorder=%i; data=%s; first_child=%a; next_sibling=%a; parent=%a }"
76 (QName.to_string node.tag)
79 print_node_ptr node.first_child
80 print_node_ptr node.next_sibling
81 print_node_ptr node.parent
84 let debug_ctx fmt ctx =
85 Format.fprintf fmt "Current context: { preorder = %i\n; stack = \n%a\n }\n-------------\n"
87 (Pretty.print_list ~sep:";\n" debug_node) ctx.stack
90 let push n ctx = ctx.stack <- n :: ctx.stack
93 [] -> failwith "XML parse error"
94 | e :: l -> ctx.stack <- l; e
96 let top ctx = match ctx.stack with
97 | [] -> failwith "XML parse error"
101 let i = ctx.current_preorder in
102 ctx.current_preorder <- 1 + i;
105 let is_left n = n.next_sibling == dummy
108 let text_string = QName.to_string QName.text
109 let attr_map_string = QName.to_string QName.attribute_map
111 let rec start_element_handler parser_ ctx tag attr_list =
113 let parent = top ctx in
114 let n = { tag = QName.of_string tag;
118 next_sibling = dummy;
122 if parent.first_child == dummy then parent.first_child <- n
123 else parent.next_sibling <- n;
128 start_element_handler parser_ ctx attr_map_string [];
129 List.iter (do_attribute parser_ ctx) attr_list;
130 end_element_handler parser_ ctx attr_map_string
132 and do_attribute parser_ ctx (att, value) =
133 let att_tag = " " ^ att in
134 start_element_handler parser_ ctx att_tag [];
135 start_element_handler parser_ ctx text_string [];
136 let n = top ctx in n.data <- value;
137 end_element_handler parser_ ctx text_string;
138 end_element_handler parser_ ctx att_tag
140 and consume_closing ctx n =
141 if n.next_sibling != dummy then
142 let _ = pop ctx in consume_closing ctx (top ctx)
144 and end_element_handler parser_ ctx tag =
146 let node = top ctx in
147 if node.first_child == dummy then node.first_child <- nil
149 node.next_sibling <- nil;
150 consume_closing ctx node
153 and do_text parser_ ctx =
154 if Buffer.length ctx.text_buffer != 0 then
155 let s = Buffer.contents ctx.text_buffer in
156 Buffer.clear ctx.text_buffer;
157 start_element_handler parser_ ctx text_string [];
158 let node = top ctx in
160 end_element_handler parser_ ctx text_string
164 let character_data_handler parser_ ctx text =
165 Buffer.add_string ctx.text_buffer text
167 let create_parser () =
168 let ctx = { text_buffer = Buffer.create 512;
169 current_preorder = 0;
171 let parser_ = Expat.parser_create ~encoding:None in
172 Expat.set_start_element_handler parser_ (start_element_handler parser_ ctx);
173 Expat.set_end_element_handler parser_ (end_element_handler parser_ ctx);
174 Expat.set_character_data_handler parser_ (character_data_handler parser_ ctx);
175 push { tag = QName.document;
179 next_sibling = dummy;
184 let node = top ctx in
185 node.next_sibling <- nil;
186 consume_closing ctx node;
189 root.next_sibling <- nil;
191 | _ -> raise (Expat.Expat_error Expat.UNCLOSED_TOKEN)
196 let parser_, finalize = create_parser () in
197 Expat.parse parser_ s;
201 let buffer = String.create 4096 in
202 let parser_, finalize = create_parser () in
204 let read = input fd buffer 0 4096 in
206 let () = Expat.parse_sub parser_ buffer 0 read in
208 in loop (); finalize ()
213 let load_xml_file = Parser.parse_file
214 let load_xml_string = Parser.parse_string
217 let output_escape_string out s =
218 for i = 0 to String.length s - 1 do
220 | '<' -> output_string out "<"
221 | '>' -> output_string out ">"
222 | '&' -> output_string out "&"
223 | '"' -> output_string out """
224 | '\'' -> output_string out "'"
225 | c -> output_char out c
228 let rec print_attributes out tree_ node =
229 if node != nil then begin
230 output_string out (QName.to_string node.tag);
231 output_string out "=\"";
232 output_escape_string out node.first_child.data;
234 print_attributes out tree_ node.next_sibling
237 let rec print_xml out tree_ node =
240 if node.tag == QName.text then
241 output_escape_string out node.data
243 let tag = QName.to_string node.tag in
245 output_string out tag;
247 if node.first_child.tag == QName.attribute_map then
249 print_attributes out tree_ node.first_child.first_child
251 node.first_child.next_sibling
255 if fchild == nil then output_string out "/>"
258 print_xml out tree_ fchild;
259 output_string out "</";
260 output_string out tag;
264 print_xml out tree_ node.next_sibling
268 let first_child _ n = n.first_child
269 let next_sibling _ n = n.next_sibling
270 let parent _ n = n.parent
272 let data _ n = n.data
273 let preorder _ n = n.preorder