1 (******************************************************************************)
2 (* SXSI : XPath evaluator *)
3 (* Kim Nguyen (Kim.Nguyen@nicta.com.au) *)
4 (* Copyright NICTA 2008 *)
5 (* Distributed under the terms of the LGPL (see LICENCE) *)
6 (******************************************************************************)
11 type descr = Nil | Node of node_content |String of string_content
13 val parse_xml_uri : string -> t
14 val parse_xml_string : string -> t
15 val string : t -> string
16 val descr : t -> descr
21 val print_xml_fast : out_channel -> t -> unit
22 val compare : t -> t -> int
23 val equal : t -> t -> bool
29 type string_content = string
30 type descr = Nil | Node of node_content | String of string_content
31 and node_content = int*Tag.t * descr * descr * (descr ref)
36 let string = function String s -> s | _ -> failwith "string"
39 external parse_xml_uri : string -> t = "caml_call_shredder_uri"
40 external parse_xml_string : string -> t = "caml_call_shredder_string"
42 let parse_xml_uri s = Node(0,Tag.tag "",parse_xml_uri s,Nil,ref Nil)
43 let parse_xml_string s = Node(0,Tag.tag "",parse_xml_string s,Nil,ref Nil)
44 let tstring = function Nil -> "Nil"
45 | Node (_,_,_,_,_) -> "Node"
46 | String _ -> "String"
50 let pp_str = Format.pp_print_string fmt in
51 let rec loop = function Nil -> ()
52 | String (s) -> pp_str s
53 | Node (_,t,l,r,_) when Tag.equal t Tag.pcdata -> loop l;loop r
55 pp_str ("<" ^ (Tag.to_string t));
58 | Node(_,t',atts,children,_) when Tag.equal t' Tag.attribute ->
59 (loop_attributes atts;
65 pp_str ("</"^ (Tag.to_string t)^">" )
67 | _ -> pp_str ">"; loop l;
68 pp_str ("</"^ (Tag.to_string t)^">" );
70 and loop_attributes = function
71 | Node(_,t,Node(_,_,String(s),_,_),r,_) ->
72 pp_str (" "^(Tag.to_string t)^"=\""^ s ^"\"") ;
80 function Node(i,t,l,_,_) -> print_xml fmt (Node(i,t,l,Nil,ref Nil))
81 | t -> print_xml fmt t
84 (* a bit ugly but inlining like this makes serialization faster *)
86 let print_xml_fast outc t =
87 let rec loop = function Nil -> ()
88 | String (s) -> output_string outc s
89 | Node (_,t,l,r,_) when Tag.equal t Tag.pcdata -> loop l;loop r
90 | Node (_,t,l,r,_) -> let t = Tag.to_string t in
94 Nil -> output_string outc "/>"
95 | Node(_,t',atts,children,_) when Tag.equal t' Tag.attribute ->
96 (loop_attributes atts;
98 | Nil -> output_string outc "/>"
100 output_char outc '>';
102 output_string outc "</";
103 output_string outc t;
104 output_char outc '>' )
106 output_char outc '>';
108 output_string outc "</";
109 output_string outc t;
112 and loop_attributes = function
113 | Node(_,t,Node(_,_,String(s),_,_),r,_) ->
114 output_char outc ' ';
115 output_string outc (Tag.to_string t);
116 output_string outc "=\"";
117 output_string outc s;
118 output_char outc '"';
125 let print_xml_fast outc =
126 function Node(i,t,l,_,_) -> print_xml_fast outc (Node(i,t,l,Nil,ref Nil))
127 | t -> print_xml_fast outc t
136 Format.fprintf fmt " "
143 | Nil -> prtabs fmt; Format.fprintf fmt "#"
144 | String s -> prtabs fmt; Format.fprintf fmt "(String %s)" s
145 | Node(id,t,l,r,_) ->
147 Format.fprintf fmt " (tag='";
149 Format.fprintf fmt "', id='%i')\n" id;
152 Format.fprintf fmt "\n";
155 Format.fprintf fmt "\n";
156 prtabs fmt;prtabs fmt;
157 Format.fprintf fmt "(id='%i'end )\n" id
166 let id = function Node(i,_,_,_,_) -> i
169 let tag = function Node(_,t,_,_,_) -> t
170 | _ -> failwith "tag"
172 let left = function Node(_,_,l,_,_) -> l
173 | _ -> failwith "left"
175 let right = function Node(_,_,_,r,_) -> r
176 | _ -> failwith "right"
178 let first_child = left
179 let next_sibling = right
181 let is_root = function Node (_,_,_,_,{contents=Nil}) -> true | _ -> false
182 let is_left n = match n with
183 | Node (_,_,_,_,{contents=p}) when not(is_root n) && (left p) == n -> true
186 let is_right n = match n with
187 | Node (_,_,_,_,{contents=p}) when not(is_root n) && (right p) == n -> true
191 let compare t1 t2 = match t1,t2 with
193 | String s1, String s2 -> String.compare s1 s2
194 | Nil, String _ -> -1
196 | Node(i1,_,_,_,_), Node(i2,_,_,_,_) -> i1 - i2
199 let equal t1 t2 = (compare t1 t2) == 0
201 let int_size = Sys.word_size/8
202 let ssize s = ((String.length s)/4 +1)*4
204 function Nil -> (int_size,1,0,0)
205 | String s -> (int_size + (ssize s),0,1,0)
207 let sizel,nl,sl,il = size l
208 and sizer,nr,sr,ir = size r
210 (sizel+sizer+(7*int_size),nl+nr,sl+sr,il+ir+1)
212 let s,n,st,i = size t in
222 type node_kind = [`Text | `Tree ]
224 let compare : 'a node -> 'a node -> int = fun x y -> x - y
225 let equal : 'a node -> 'a node -> bool = fun x y -> x == y
227 (* abstract type, values are pointers to a XMLTree C++ object *)
230 external parse_xml_uri : string -> t = "caml_call_shredder_uri"
231 let parse_xml_uri uri = parse_xml_uri uri
233 external parse_xml_string : string -> t = "caml_call_shredder_string"
234 let parse_xml_string uri = parse_xml_string uri
239 type t (* pointer to the text collection *)
241 external nullt : unit -> [`Text ] node = "caml_xml_tree_nullt"
243 external get_text : t -> [`Text] node -> string = "caml_text_collection_get_text"
244 let is_empty _ (n : [`Text] node) = equal nil n
251 external serialize : string -> unit = "caml_xml_tree_serialize"
252 external unserialize : string -> t = "caml_xml_tree_unserialize"
254 external root : t -> [`Tree] node = "caml_xml_tree_root"
255 external nullt : unit -> [`Tree ] node = "caml_xml_tree_nullt"
258 let is_nil x = equal x nil
260 external parent : t -> [`Tree] node -> [`Tree] node = "caml_xml_tree_first_child"
261 external parent_doc : t -> [`Text ] node -> [`Tree ] node = "caml_xml_tree_parent_doc"
262 external first_child : t -> [`Tree] node -> [`Tree] node = "caml_xml_tree_first_child"
263 external next_sibling : t -> [`Tree] node -> [`Tree] node = "caml_xml_tree_next_sibling"
265 external is_leaf : t -> [`Tree] node -> bool = "caml_xml_tree_is_leaf"
267 external tag : t -> [`Tree ] node -> Tag.t = "caml_xml_tree_tag"
269 external text_collection : t -> Text.t = "caml_xml_tree_text_collection"
271 let is_last t n = equal nil (next_sibling t n)
273 external prev_text : t -> [`Tree] node -> [`Text ] node = "caml_xml_tree_prev_text"
274 external my_text : t -> [`Tree] node -> [`Text ] node = "caml_xml_tree_my_text"
275 external next_text : t -> [`Tree] node -> [`Text ] node = "caml_xml_tree_next_text"
277 external text_xml_id : t -> [`Text ] node -> int = "caml_xml_tree_text_xml_id"
278 external node_xml_id : t -> [`Tree ] node -> int = "caml_xml_tree_node_xml_id"
283 module Binary : BINARY = struct
286 [ `Node of [`Tree ] node
287 | `String of [`Text ] node * [`Tree ] node ]
288 type string_content = [ `Text ] node
291 | Node of node_content
292 | String of string_content
296 type t = { doc : doc;
301 let node_of_t t = { doc= t; node= Node(`Node (root t)) }
304 let parse_xml_uri str = node_of_t (parse_xml_uri str)
305 let parse_xml_string str = node_of_t (parse_xml_string str)
307 let compare a b = match a.node,b.node with
308 | Node(`Node i),Node(`Node j) -> compare i j
309 | _, Node(`Node( _ )) -> 1
310 | Node(`String (i,_)),Node(`String (j,_)) -> compare i j
311 | Node(`Node( _ )),Node(`String (_,_)) -> -1
312 | _, Node(`String (_,_)) -> 1
313 | String i, String j -> compare i j
314 | Node _ , String _ -> -1
319 let equal a b = (compare a b) == 0
321 let string t = match t.node with
322 | String i -> Text.get_text (text_collection t.doc) i
325 let norm (n : [`Tree ] node ) = if is_nil n then Nil else Node (`Node n)
332 | Nil | String _ -> failwith "first_child"
334 let fs = first_child n.doc t in
335 let txt = prev_text n.doc t in
336 if Text.is_empty (text_collection n.doc) txt
338 else Node (`String (txt, fs))
340 | Node(`String (i,_)) -> String i
342 { n with node = node'}
347 | Nil | String _ -> failwith "next_sibling"
348 | Node (`String (_,ns)) -> norm ns
350 let ns = next_sibling n.doc t in
351 let txt = next_text n.doc t in
352 if Text.is_empty (text_collection n.doc) txt
354 else Node (`String (txt, ns))
356 { n with node = node'}
359 let left = first_child
360 let right = next_sibling
362 function { doc=d; node=Node(`Node n)} -> text_xml_id d n
363 | { doc=d; node=Node(`String (i,_) )} -> node_xml_id d i
367 function { node=Node(`String _) } -> Tag.pcdata
368 | { doc=d; node=Node(`Node n)} -> tag d n
369 | _ -> failwith "Tag"
373 let print_xml_fast outc t =
374 let rec loop ?(print_right=true) t = match t.node with
376 | String (s) -> output_string outc (string t)
377 | Node _ when Tag.equal (tag t) Tag.pcdata -> loop (left t); loop (right t)
380 let tg = Tag.to_string (tag t) in
384 output_char outc '<';
385 output_string outc tg;
387 Nil -> output_string outc "/>"
388 | Node(_) when Tag.equal (tag l) Tag.attribute ->
389 (loop_attributes (left l);
390 match (right l).node with
391 | Nil -> output_string outc "/>"
393 output_char outc '>';
395 output_string outc "</";
396 output_string outc tg;
397 output_char outc '>' )
399 output_char outc '>';
401 output_string outc "</";
402 output_string outc tg;
404 );if print_right then loop r
405 and loop_attributes a = match a.node with
406 | Node(_) -> let value = string (left(left a)) in
407 output_char outc ' ';
408 output_string outc (Tag.to_string (tag a));
409 output_string outc "=\"";
410 output_string outc value;
411 output_char outc '"';
412 loop_attributes (right a)
415 loop ~print_right:false t