1 (* Copyright (C) 2005, HELM Team.
3 * This file is part of HELM, an Hypertextual, Electronic
4 * Library of Mathematics, developed at the Computer Science
5 * Department, University of Bologna, Italy.
7 * HELM is free software; you can redistribute it and/or
8 * modify it under the terms of the GNU General Public License
9 * as published by the Free Software Foundation; either version 2
10 * of the License, or (at your option) any later version.
12 * HELM is distributed in the hope that it will be useful,
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 * GNU General Public License for more details.
17 * You should have received a copy of the GNU General Public License
18 * along with HELM; if not, write to the Free Software
19 * Foundation, Inc., 59 Temple Place - Suite 330, Boston,
22 * For details, see the HELM World-Wide-Web page,
23 * http://helm.cs.unibo.it/
30 exception Error of int * int * string
32 module StringSet = Set.Make(String)
35 let regexp number = xml_digit+
36 let regexp utf8_blank = " " | "\r\n" | "\n" | "\t" | [160] (* this is a nbsp *)
37 let regexp percentage =
38 ('-' | "") [ '0' - '9' ] + '%'
39 let regexp floatwithunit =
40 ('-' | "") [ '0' - '9' ] + ["."] [ '0' - '9' ] + ([ 'a' - 'z' ] + | "" )
41 let regexp color = "#" [ '0' - '9' 'a' - 'f' 'A' - 'F' ] [ '0' - '9' 'a' - 'f'
42 'A' - 'F' ] [ '0' - '9' 'a' - 'f' 'A' - 'F' ] [ '0' - '9' 'a' - 'f' 'A' - 'F' ]
43 [ '0' - '9' 'a' - 'f' 'A' - 'F' ] [ '0' - '9' 'a' - 'f' 'A' - 'F' ]
45 (* ZACK: breaks unicode's binder followed by an ascii letter without blank *)
46 (* let regexp ident_letter = xml_letter *)
48 let regexp ident_letter = [ 'a' - 'z' 'A' - 'Z' ]
50 (* must be in sync with "is_ligature_char" below *)
51 let regexp ligature_char = [ "'`~!?@*()[]<>-+=|:;.,/\"" ]
52 let regexp ligature = ligature_char ligature_char+
54 let regexp we_proved = "we" utf8_blank+ "proved"
55 let regexp we_have = "we" utf8_blank+ "have"
56 let regexp let_rec = "let" utf8_blank+ "rec"
57 let regexp let_corec = "let" utf8_blank+ "corec"
58 let regexp ident_decoration = '\'' | '?' | '`'
59 let regexp ident_cont = ident_letter | xml_digit | '_'
60 let regexp ident_start = ident_letter
61 let regexp ident = ident_letter ident_cont* ident_decoration*
62 let regexp variable_ident = '_' '_' number
63 let regexp pident = '_' ident
65 let regexp uri_step = [ 'a' - 'z' 'A' - 'Z' '0' - '9' '_' '-' ''' ]+
68 ("cic:/" | "theory:/") (* schema *)
69 (* ident ('/' ident)* |+ path +| *)
70 uri_step ('/' uri_step)* (* path *)
71 ('.' ident)+ (* ext *)
72 (* ("#xpointer(" number ('/' number)+ ")")? (* xpointer *) *)
73 ("(" number (',' number)* ")")? (* reference spec *)
75 let regexp qstring = '"' [^ '"']* '"'
76 let regexp hreftag = "<" [ 'A' 'a' ]
77 let regexp href = "href=\"" uri "\""
78 let regexp hreftitle = "title=" qstring
79 let regexp hrefclose = "</" [ 'A' 'a' ] ">"
81 let regexp tex_token = '\\' ident
83 let regexp delim_begin = "\\["
84 let regexp delim_end = "\\]"
86 let regexp qkeyword = "'" ( ident | pident ) "'"
88 let regexp implicit = '?'
89 let regexp placeholder = '%'
90 let regexp meta = implicit number
92 let regexp csymbol = '\'' ident
94 let regexp begin_group = "@{" | "${"
95 let regexp end_group = '}'
96 let regexp wildcard = "$_"
97 let regexp ast_ident = "@" ident
98 let regexp ast_csymbol = "@" csymbol
99 let regexp meta_ident = "$" ident
100 let regexp meta_anonymous = "$_"
102 let regexp begincomment = "(**" utf8_blank
103 let regexp beginnote = "(*"
104 let regexp endcomment = "*)"
105 (* let regexp comment_char = [^'*'] | '*'[^')']
106 let regexp note = "|+" ([^'*'] | "**") comment_char* "+|" *)
111 "over"; "atop"; "frac";
112 "sqrt"; "root"; "mstyle" ; "mpadded"; "maction"
116 let level1_keywords =
117 [ "hbox"; "hvbox"; "hovbox"; "vbox";
119 "list0"; "list1"; "sep";
121 "term"; "ident"; "number";
124 let level2_meta_keywords =
125 [ "if"; "then"; "elCicNotationParser.se";
126 "fold"; "left"; "right"; "rec";
129 "anonymous"; "ident"; "number"; "term"; "fresh"
132 (* (string, int) Hashtbl.t, with multiple bindings.
133 * int is the unicode codepoint *)
134 let ligatures = Hashtbl.create 23
138 (fun (ligature, symbol) -> Hashtbl.add ligatures ligature symbol)
139 [ ("->", <:unicode<to>>); ("=>", <:unicode<Rightarrow>>);
140 (":=", <:unicode<def>>);
143 let regexp nreference =
145 uri_step ('/' uri_step)* (* path *)
148 | "def" "(" number ")"
149 | "fix" "(" number "," number "," number ")"
150 | "cfx" "(" number ")"
151 | "ind" "(" number "," number "," number ")"
152 | "con" "(" number "," number "," number ")") (* ext + reference *)
154 let error lexbuf msg =
155 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
156 raise (Error (begin_cnum, end_cnum, msg))
157 let error_at_end lexbuf msg =
158 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
159 raise (Error (begin_cnum, end_cnum, msg))
161 let loc_of_buf lexbuf =
162 HExtlib.floc_of_loc (Ulexing.loc lexbuf)
164 let return_with_loc token begin_cnum end_cnum =
165 let flocation = HExtlib.floc_of_loc (begin_cnum,end_cnum) in
168 let return lexbuf token =
169 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
170 return_with_loc token begin_cnum end_cnum
172 let return_lexeme lexbuf name = return lexbuf (name, Ulexing.utf8_lexeme lexbuf)
174 let return_symbol lexbuf s = return lexbuf ("SYMBOL", s)
175 let return_eoi lexbuf = return lexbuf ("EOI", "")
177 let remove_quotes s = String.sub s 1 (String.length s - 2)
180 let tok_func stream =
181 (* let lexbuf = Ulexing.from_utf8_stream stream in *)
182 (** XXX Obj.magic rationale.
184 * camlp5 constraints the tok_func field of Token.glexer to have type:
185 * Stream.t char -> (Stream.t 'te * flocation_function)
186 * In order to use ulex we have (in theory) to instantiate a new lexbuf each
187 * time a char Stream.t is passed, destroying the previous lexbuf which may
188 * have consumed a character from the old stream which is lost forever :-(
190 * Instead of passing to camlp5 a char Stream.t we pass a lexbuf, casting it to
191 * char Stream.t with Obj.magic where needed.
193 let lexbuf = Obj.magic stream in
194 Token.make_stream_and_location
199 | Ulexing.Error -> error_at_end lexbuf "Unexpected character"
200 | Ulexing.InvalidCodepoint p ->
201 error_at_end lexbuf (sprintf "Invalid code point: %d" p))
204 Token.tok_func = tok_func;
205 Token.tok_using = (fun _ -> ());
206 Token.tok_removing = (fun _ -> ());
207 Token.tok_match = Token.default_match;
208 Token.tok_text = Token.lexer_text;
209 Token.tok_comm = None;
212 let expand_macro lexbuf =
214 Ulexing.utf8_sub_lexeme lexbuf 1 (Ulexing.lexeme_length lexbuf - 1)
217 ("SYMBOL", Utf8Macro.expand macro)
218 with Utf8Macro.Macro_not_found _ ->
219 (* FG: unexpanded TeX macros are terminated by a space for rendering *)
220 "SYMBOL", (Ulexing.utf8_lexeme lexbuf ^ " ")
222 let remove_quotes s = String.sub s 1 (String.length s - 2)
223 let remove_left_quote s = String.sub s 1 (String.length s - 1)
225 let rec level2_pattern_token_group counter buffer =
228 if (counter > 0) then
229 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
230 snd (Ulexing.loc lexbuf)
232 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
233 ignore (level2_pattern_token_group (counter + 1) buffer lexbuf) ;
234 level2_pattern_token_group counter buffer lexbuf
236 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
237 level2_pattern_token_group counter buffer lexbuf
239 let read_unparsed_group token_name lexbuf =
240 let buffer = Buffer.create 16 in
241 let begin_cnum, _ = Ulexing.loc lexbuf in
242 let end_cnum = level2_pattern_token_group 0 buffer lexbuf in
243 return_with_loc (token_name, Buffer.contents buffer) begin_cnum end_cnum
245 let handle_keywords lexbuf k name =
246 let s = Ulexing.utf8_lexeme lexbuf in
248 return lexbuf ("", s)
250 return lexbuf (name, s)
253 let rec level2_meta_token =
255 | utf8_blank+ -> level2_meta_token lexbuf
256 | hreftag -> return lexbuf ("ATAG","")
257 | hrefclose -> return lexbuf ("ATAGEND","")
259 handle_keywords lexbuf (fun x -> List.mem x level2_meta_keywords) "IDENT"
260 | variable_ident -> return lexbuf ("IDENT", Ulexing.utf8_lexeme lexbuf)
262 handle_keywords lexbuf (fun x -> List.mem x level2_meta_keywords) "PIDENT"
263 | "@{" -> read_unparsed_group "UNPARSED_AST" lexbuf
265 return lexbuf ("UNPARSED_AST",
266 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
268 return lexbuf ("UNPARSED_AST",
269 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
270 | eof -> return_eoi lexbuf
272 let rec comment_token acc depth =
275 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
276 comment_token acc (depth + 1) lexbuf
278 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
281 else comment_token acc (depth - 1) lexbuf
283 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
284 comment_token acc depth lexbuf
286 (** @param k continuation to be invoked when no ligature has been found *)
287 let ligatures_token k =
290 let lexeme = Ulexing.utf8_lexeme lexbuf in
291 (match List.rev (Hashtbl.find_all ligatures lexeme) with
292 | [] -> (* ligature not found, rollback and try default lexer *)
293 Ulexing.rollback lexbuf;
295 | default_lig :: _ -> (* ligatures found, use the default one *)
296 return_symbol lexbuf default_lig)
297 | eof -> return_eoi lexbuf
298 | _ -> (* not a ligature, rollback and try default lexer *)
299 Ulexing.rollback lexbuf;
304 type t = Stdpp.location
305 let compare = Pervasives.compare
308 module LocalizeEnv = Map.Make (LocalizeOD)
315 let update_table loc desc href loctable =
316 if desc <> None || href <> None
318 (let s,e = HExtlib.loc_of_floc loc in
319 prerr_endline (Printf.sprintf "*** [%d,%d] \"%s\",\"%s\""
320 s e (so_pp href) (so_pp desc));
321 LocalizeEnv.add loc (href,desc) loctable)
325 let level2_ast_token loctable status =
326 let rec aux desc href =
328 | let_rec -> return lexbuf ("LETREC","")
329 | let_corec -> return lexbuf ("LETCOREC","")
330 | we_proved -> return lexbuf ("WEPROVED","")
331 | we_have -> return lexbuf ("WEHAVE","")
332 | utf8_blank+ -> ligatures_token (aux desc href) lexbuf
334 let s = Ulexing.utf8_lexeme lexbuf in
335 return lexbuf ("META", String.sub s 1 (String.length s - 1))
336 | implicit -> return lexbuf ("IMPLICIT", "")
337 | placeholder -> return lexbuf ("PLACEHOLDER", "")
338 | hreftag -> aux_in_tag None None lexbuf
339 | hrefclose -> aux None None lexbuf
340 | ident -> loctable :=
341 update_table (loc_of_buf lexbuf) desc href !loctable;
342 handle_keywords lexbuf (fun x -> StringSet.mem x status) "IDENT"
344 return lexbuf ("IDENT", Ulexing.utf8_lexeme lexbuf)
345 | pident -> loctable :=
346 update_table (loc_of_buf lexbuf) desc href !loctable;
347 handle_keywords lexbuf (fun x -> StringSet.mem x status) "PIDENT"
348 | number -> loctable :=
349 update_table (loc_of_buf lexbuf) desc href !loctable;
350 return lexbuf ("NUMBER", Ulexing.utf8_lexeme lexbuf)
351 | tex_token -> loctable :=
352 update_table (loc_of_buf lexbuf) desc href !loctable;
353 return lexbuf (expand_macro lexbuf)
354 | nreference -> return lexbuf ("NREF", Ulexing.utf8_lexeme lexbuf)
355 | uri -> return lexbuf ("URI", Ulexing.utf8_lexeme lexbuf)
357 return lexbuf ("QSTRING", remove_quotes (Ulexing.utf8_lexeme lexbuf))
359 return lexbuf ("CSYMBOL", remove_left_quote (Ulexing.utf8_lexeme lexbuf))
360 | "${" -> read_unparsed_group "UNPARSED_META" lexbuf
361 | "@{" -> read_unparsed_group "UNPARSED_AST" lexbuf
362 | '(' -> return lexbuf ("LPAREN", "")
363 | ')' -> return lexbuf ("RPAREN", "")
365 return lexbuf ("UNPARSED_META",
366 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
367 | meta_anonymous -> return lexbuf ("UNPARSED_META", "anonymous")
369 let _comment = comment_token (Ulexing.utf8_lexeme lexbuf) 0 lexbuf in
371 Ulexing.utf8_sub_lexeme lexbuf 2 (Ulexing.lexeme_length lexbuf - 4)
373 return lexbuf ("NOTE", comment) *)
374 ligatures_token (aux desc href) lexbuf
375 | begincomment -> return lexbuf ("BEGINCOMMENT","")
376 | endcomment -> return lexbuf ("ENDCOMMENT","")
377 | eof -> return_eoi lexbuf
379 update_table (loc_of_buf lexbuf) desc href !loctable;
380 return_symbol lexbuf (Ulexing.utf8_lexeme lexbuf)
382 and aux_in_tag desc href = lexer
383 | utf8_blank+ -> ligatures_token (aux_in_tag desc href) lexbuf
386 (Some (Ulexing.utf8_sub_lexeme lexbuf 6 (Ulexing.lexeme_length lexbuf - 7)))
390 (Some (Ulexing.utf8_sub_lexeme lexbuf 7 (Ulexing.lexeme_length lexbuf - 8)))
392 | ">" -> aux desc href lexbuf
393 | _ -> aux None None lexbuf
396 let rec level1_pattern_token =
398 | utf8_blank+ -> ligatures_token level1_pattern_token lexbuf
399 | number -> return lexbuf ("NUMBER", Ulexing.utf8_lexeme lexbuf)
400 | hreftag -> return lexbuf ("ATAG", "")
401 | hrefclose -> return lexbuf ("ATAGEND","")
402 | ident ->handle_keywords lexbuf (fun x -> List.mem x level1_keywords) "IDENT"
403 | variable_ident -> return lexbuf ("IDENT", Ulexing.utf8_lexeme lexbuf)
404 | pident->handle_keywords lexbuf (fun x->List.mem x level1_keywords) "PIDENT"
405 | color -> return lexbuf ("COLOR", Ulexing.utf8_lexeme lexbuf)
407 return lexbuf ("PERCENTAGE", Ulexing.utf8_lexeme lexbuf)
409 return lexbuf ("FLOATWITHUNIT", Ulexing.utf8_lexeme lexbuf)
410 | tex_token -> return lexbuf (expand_macro lexbuf)
412 return lexbuf ("QKEYWORD", remove_quotes (Ulexing.utf8_lexeme lexbuf))
413 | '(' -> return lexbuf ("LPAREN", "")
414 | ')' -> return lexbuf ("RPAREN", "")
415 | eof -> return_eoi lexbuf
416 | _ -> return_symbol lexbuf (Ulexing.utf8_lexeme lexbuf)
418 let level1_pattern_token = ligatures_token level1_pattern_token
419 let level2_ast_token loctable status =
420 ligatures_token (level2_ast_token loctable status)
422 (* API implementation *)
424 level1_pattern_lexer : (string * string) Token.glexer;
425 level2_ast_lexer : (string * string) Token.glexer;
426 level2_meta_lexer : (string * string) Token.glexer
429 let mk_lexers loctable keywords =
430 let initial_keywords =
431 [ "CProp"; "Prop"; "Type"; "Set"; "let"; "match";
432 "with"; "in"; "and"; "to"; "as"; "on"; "return"; "done" ]
435 List.fold_right StringSet.add (initial_keywords @ keywords) StringSet.empty
438 level1_pattern_lexer = mk_lexer level1_pattern_token;
439 level2_ast_lexer = mk_lexer (level2_ast_token loctable status);
440 level2_meta_lexer = mk_lexer level2_meta_token