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 xmarkup = '\005'
77 let regexp ymarkup = '\006'
79 let regexp hreftag = xmarkup [ 'A' 'a' ]
80 let regexp href = "href=\"" uri "\""
81 let regexp hreftitle = "title=" qstring
82 let regexp hrefclose = xmarkup "/" [ 'A' 'a' ] ymarkup
84 let regexp tag_cont = ident_letter | xml_digit | "_" | "-"
85 let regexp gtagname = [ 'B' - 'Z' 'b' - 'z' ] | ident_letter tag_cont+
86 let regexp attrname = ident_letter tag_cont*
87 let regexp gopentag = xmarkup gtagname
88 let regexp closetag = xmarkup "/" gtagname ymarkup
89 let regexp attribute = attrname "=" qstring
90 let regexp generictag =
91 gopentag (utf8_blank+ attribute)* ymarkup
93 let regexp tex_token = '\\' ident
95 let regexp delim_begin = "\\["
96 let regexp delim_end = "\\]"
98 let regexp qkeyword = "'" ( ident | pident ) "'"
100 let regexp implicit = '?'
101 let regexp placeholder = '%'
102 let regexp meta = implicit number
104 let regexp csymbol = '\'' ident
106 let regexp begin_group = "@{" | "${"
107 let regexp end_group = '}'
108 let regexp wildcard = "$_"
109 let regexp ast_ident = "@" ident
110 let regexp ast_csymbol = "@" csymbol
111 let regexp meta_ident = "$" ident
112 let regexp meta_anonymous = "$_"
114 let regexp begincomment = "(**" utf8_blank
115 let regexp beginnote = "(*"
116 let regexp endcomment = "*)"
117 (* let regexp comment_char = [^'*'] | '*'[^')']
118 let regexp note = "|+" ([^'*'] | "**") comment_char* "+|" *)
123 "over"; "atop"; "frac";
124 "sqrt"; "root"; "mstyle" ; "mpadded"; "maction"
128 let level1_keywords =
129 [ "hbox"; "hvbox"; "hovbox"; "vbox";
131 "list0"; "list1"; "sep";
133 "term"; "ident"; "number"; "ref"
136 let level2_meta_keywords =
137 [ "if"; "then"; "else";
138 "fold"; "left"; "right"; "rec";
141 "anonymous"; "ident"; "number"; "term"; "fresh"
144 (* (string, int) Hashtbl.t, with multiple bindings.
145 * int is the unicode codepoint *)
146 let ligatures = Hashtbl.create 23
150 (fun (ligature, symbol) -> Hashtbl.add ligatures ligature symbol)
151 [ ("->", <:unicode<to>>); ("=>", <:unicode<Rightarrow>>);
152 (":=", <:unicode<def>>);
155 let regexp nreference =
157 uri_step ('/' uri_step)* (* path *)
160 | "def" "(" number ")"
161 | "fix" "(" number "," number "," number ")"
162 | "cfx" "(" number ")"
163 | "ind" "(" number "," number "," number ")"
164 | "con" "(" number "," number "," number ")") (* ext + reference *)
166 let error lexbuf msg =
167 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
168 raise (Error (begin_cnum, end_cnum, msg))
169 let error_at_end lexbuf msg =
170 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
171 raise (Error (begin_cnum, end_cnum, msg))
173 let loc_of_buf lexbuf =
174 HExtlib.floc_of_loc (Ulexing.loc lexbuf)
176 let return_with_loc token begin_cnum end_cnum =
177 let flocation = HExtlib.floc_of_loc (begin_cnum,end_cnum) in
180 let return lexbuf token =
181 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
182 return_with_loc token begin_cnum end_cnum
184 let return_lexeme lexbuf name = return lexbuf (name, Ulexing.utf8_lexeme lexbuf)
186 let return_symbol lexbuf s = return lexbuf ("SYMBOL", s)
187 let return_eoi lexbuf = return lexbuf ("EOI", "")
189 let remove_quotes s = String.sub s 1 (String.length s - 2)
192 let tok_func stream =
193 (* let lexbuf = Ulexing.from_utf8_stream stream in *)
194 (** XXX Obj.magic rationale.
196 * camlp5 constraints the tok_func field of Token.glexer to have type:
197 * Stream.t char -> (Stream.t 'te * flocation_function)
198 * In order to use ulex we have (in theory) to instantiate a new lexbuf each
199 * time a char Stream.t is passed, destroying the previous lexbuf which may
200 * have consumed a character from the old stream which is lost forever :-(
202 * Instead of passing to camlp5 a char Stream.t we pass a lexbuf, casting it to
203 * char Stream.t with Obj.magic where needed.
205 let lexbuf = Obj.magic stream in
206 Token.make_stream_and_location
211 | Ulexing.Error -> error_at_end lexbuf "Unexpected character"
212 | Ulexing.InvalidCodepoint p ->
213 error_at_end lexbuf (sprintf "Invalid code point: %d" p))
216 Token.tok_func = tok_func;
217 Token.tok_using = (fun _ -> ());
218 Token.tok_removing = (fun _ -> ());
219 Token.tok_match = Token.default_match;
220 Token.tok_text = Token.lexer_text;
221 Token.tok_comm = None;
224 let expand_macro lexbuf =
226 Ulexing.utf8_sub_lexeme lexbuf 1 (Ulexing.lexeme_length lexbuf - 1)
229 ("SYMBOL", Utf8Macro.expand macro)
230 with Utf8Macro.Macro_not_found _ ->
231 (* FG: unexpanded TeX macros are terminated by a space for rendering *)
232 "SYMBOL", (Ulexing.utf8_lexeme lexbuf ^ " ")
234 let remove_quotes s = String.sub s 1 (String.length s - 2)
235 let remove_left_quote s = String.sub s 1 (String.length s - 1)
237 let rec level2_pattern_token_group counter buffer =
240 if (counter > 0) then
241 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
242 snd (Ulexing.loc lexbuf)
244 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
245 ignore (level2_pattern_token_group (counter + 1) buffer lexbuf) ;
246 level2_pattern_token_group counter buffer lexbuf
248 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
249 level2_pattern_token_group counter buffer lexbuf
251 let read_unparsed_group token_name lexbuf =
252 let buffer = Buffer.create 16 in
253 let begin_cnum, _ = Ulexing.loc lexbuf in
254 let end_cnum = level2_pattern_token_group 0 buffer lexbuf in
255 return_with_loc (token_name, Buffer.contents buffer) begin_cnum end_cnum
257 let handle_keywords lexbuf k name =
258 let s = Ulexing.utf8_lexeme lexbuf in
259 prerr_endline (Printf.sprintf "handling \"%s\" as a keyword or %s?" s name);
261 return lexbuf ("", s)
263 return lexbuf (name, s)
266 let rec level2_meta_token =
268 | utf8_blank+ -> level2_meta_token lexbuf
269 | hreftag -> return lexbuf ("ATAG","")
270 | hrefclose -> return lexbuf ("ATAGEND","")
272 | closetag -> level2_meta_token lexbuf
274 handle_keywords lexbuf (fun x -> List.mem x level2_meta_keywords) "IDENT"
275 | variable_ident -> return lexbuf ("IDENT", Ulexing.utf8_lexeme lexbuf)
277 handle_keywords lexbuf (fun x -> List.mem x level2_meta_keywords) "PIDENT"
278 | "@{" -> read_unparsed_group "UNPARSED_AST" lexbuf
280 return lexbuf ("UNPARSED_AST",
281 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
283 return lexbuf ("UNPARSED_AST",
284 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
285 | eof -> return_eoi lexbuf
287 let rec comment_token acc depth =
290 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
291 comment_token acc (depth + 1) lexbuf
293 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
296 else comment_token acc (depth - 1) lexbuf
298 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
299 comment_token acc depth lexbuf
301 (** @param k continuation to be invoked when no ligature has been found *)
302 let ligatures_token k =
305 let lexeme = Ulexing.utf8_lexeme lexbuf in
306 (match List.rev (Hashtbl.find_all ligatures lexeme) with
307 | [] -> (* ligature not found, rollback and try default lexer *)
308 Ulexing.rollback lexbuf;
310 | default_lig :: _ -> (* ligatures found, use the default one *)
311 return_symbol lexbuf default_lig)
312 | eof -> return_eoi lexbuf
313 | _ -> (* not a ligature, rollback and try default lexer *)
314 Ulexing.rollback lexbuf;
319 type t = Stdpp.location
320 let compare = Pervasives.compare
323 module LocalizeEnv = Map.Make (LocalizeOD)
330 let update_table loc desc href loctable =
331 if desc <> None || href <> None
333 (let s,e = HExtlib.loc_of_floc loc in
335 prerr_endline (Printf.sprintf "*** [%d,%d] \"%s\",\"%s\""
336 s e (so_pp href) (so_pp desc));
338 Printf.printf "*** [%d,%d] \"%s\",\"%s\""
339 s e (so_pp href) (so_pp desc);
340 LocalizeEnv.add loc (href,desc) loctable)
344 let level2_ast_token loctable status =
345 prerr_endline ("X keywords = " ^ (String.concat ", "
346 (StringSet.elements status)));
347 (* start = starting point of last open A tag (default -1 = no tag) *)
348 let rec aux desc href start =
350 | let_rec -> return lexbuf ("LETREC","")
351 | let_corec -> return lexbuf ("LETCOREC","")
352 | we_proved -> return lexbuf ("WEPROVED","")
353 | we_have -> return lexbuf ("WEHAVE","")
354 | utf8_blank+ -> ligatures_token (aux desc href start) lexbuf
356 let s = Ulexing.utf8_lexeme lexbuf in
357 return lexbuf ("META", String.sub s 1 (String.length s - 1))
358 | implicit -> return lexbuf ("IMPLICIT", "")
359 | placeholder -> return lexbuf ("PLACEHOLDER", "")
361 let start = Ulexing.lexeme_start lexbuf in
362 aux_attr None None start lexbuf
363 | hrefclose -> return lexbuf ("LEXING_ERROR", Ulexing.utf8_lexeme lexbuf)
364 (* ignore other tags *)
366 | closetag -> ligatures_token (aux desc href start) lexbuf
370 handle_keywords lexbuf (fun x -> StringSet.mem x status) "IDENT"
371 in aux_close_tag desc href start idtok lexbuf
372 (* ("IDENT", Ulexing.utf8_lexeme lexbuf) lexbuf *)
374 handle_keywords lexbuf (fun x -> StringSet.mem x status) "IDENT"
376 return lexbuf ("IDENT", Ulexing.utf8_lexeme lexbuf)
377 | pident -> if start <> ~-1 then
378 aux_close_tag desc href start ("PIDENT", Ulexing.utf8_lexeme lexbuf) lexbuf
380 handle_keywords lexbuf (fun x -> StringSet.mem x status) "PIDENT"
381 | number -> let token = "NUMBER", Ulexing.utf8_lexeme lexbuf
384 aux_close_tag desc href start token lexbuf
385 else return lexbuf token
386 | tex_token -> let token = expand_macro lexbuf
389 aux_close_tag desc href start token lexbuf
390 else return lexbuf token
391 | nreference -> return lexbuf ("NREF", Ulexing.utf8_lexeme lexbuf)
392 | uri -> return lexbuf ("URI", Ulexing.utf8_lexeme lexbuf)
394 return lexbuf ("QSTRING", remove_quotes (Ulexing.utf8_lexeme lexbuf))
396 return lexbuf ("CSYMBOL", remove_left_quote (Ulexing.utf8_lexeme lexbuf))
397 | "${" -> read_unparsed_group "UNPARSED_META" lexbuf
398 | "@{" -> read_unparsed_group "UNPARSED_AST" lexbuf
399 | '(' -> return lexbuf ("LPAREN", "")
400 | ')' -> return lexbuf ("RPAREN", "")
402 return lexbuf ("UNPARSED_META",
403 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
404 | meta_anonymous -> return lexbuf ("UNPARSED_META", "anonymous")
406 let _comment = comment_token (Ulexing.utf8_lexeme lexbuf) 0 lexbuf in
408 Ulexing.utf8_sub_lexeme lexbuf 2 (Ulexing.lexeme_length lexbuf - 4)
410 return lexbuf ("NOTE", comment) *)
411 ligatures_token (aux desc href start) lexbuf
412 | begincomment -> return lexbuf ("BEGINCOMMENT","")
413 | endcomment -> return lexbuf ("ENDCOMMENT","")
414 | eof -> return_eoi lexbuf
415 | _ -> let token = "SYMBOL", (Ulexing.utf8_lexeme lexbuf)
418 aux_close_tag desc href start token lexbuf
419 else return lexbuf token
421 and aux_attr desc href start = lexer
422 | utf8_blank+ -> ligatures_token (aux_attr desc href start) lexbuf
425 (Some (Ulexing.utf8_sub_lexeme lexbuf 6 (Ulexing.lexeme_length lexbuf - 7)))
429 (Some (Ulexing.utf8_sub_lexeme lexbuf 7 (Ulexing.lexeme_length lexbuf - 8)))
431 | ymarkup -> aux desc href start lexbuf
432 | _ -> return lexbuf ("LEXING_ERROR", Ulexing.utf8_lexeme lexbuf)
434 (* and aux_in_tag desc href = lexer
435 | ident -> loctable :=
436 update_table (loc_of_buf lexbuf) desc href !loctable;
437 handle_keywords lexbuf (fun x -> StringSet.mem x status) "IDENT"
439 return lexbuf ("IDENT", Ulexing.utf8_lexeme lexbuf)
440 | pident -> loctable :=
441 update_table (loc_of_buf lexbuf) desc href !loctable;
442 handle_keywords lexbuf (fun x -> StringSet.mem x status) "PIDENT"
443 | number -> loctable :=
444 update_table (loc_of_buf lexbuf) desc href !loctable;
445 return lexbuf ("NUMBER", Ulexing.utf8_lexeme lexbuf)
446 | tex_token -> loctable :=
447 update_table (loc_of_buf lexbuf) desc href !loctable;
448 return lexbuf (expand_macro lexbuf)
450 update_table (loc_of_buf lexbuf) desc href !loctable;
451 return_symbol lexbuf (Ulexing.utf8_lexeme lexbuf)
453 and aux_close_tag desc href start token = lexer
454 | hrefclose -> let _,b = Ulexing.loc lexbuf in
455 loctable := update_table (HExtlib.floc_of_loc (start,b)) desc href !loctable;
457 (Printf.sprintf "adding loc (%d,%d) to table" start b);
458 return_with_loc token start b
459 | _ -> return lexbuf ("LEXING_ERROR", Ulexing.utf8_lexeme lexbuf)
462 let rec level1_pattern_token =
464 | utf8_blank+ -> ligatures_token level1_pattern_token lexbuf
465 | number -> return lexbuf ("NUMBER", Ulexing.utf8_lexeme lexbuf)
466 (* | hreftag -> return lexbuf ("ATAG", "")
467 | hrefclose -> return lexbuf ("ATAGEND","")
469 | closetag -> ligatures_token level1_pattern_token lexbuf *)
471 return lexbuf ("QKEYWORD", remove_quotes (Ulexing.utf8_lexeme lexbuf))
472 | csymbol -> prerr_endline ("lexing csymbol " ^ (Ulexing.utf8_lexeme lexbuf));
473 return lexbuf ("CSYMBOL", remove_left_quote (Ulexing.utf8_lexeme lexbuf))
474 | ident -> handle_keywords lexbuf (fun x -> List.mem x level1_keywords) "IDENT"
475 | variable_ident -> return lexbuf ("IDENT", Ulexing.utf8_lexeme lexbuf)
476 | pident-> handle_keywords lexbuf (fun x->List.mem x level1_keywords) "PIDENT"
477 | color -> return lexbuf ("COLOR", Ulexing.utf8_lexeme lexbuf)
479 return lexbuf ("PERCENTAGE", Ulexing.utf8_lexeme lexbuf)
481 return lexbuf ("FLOATWITHUNIT", Ulexing.utf8_lexeme lexbuf)
482 | tex_token -> return lexbuf (expand_macro lexbuf)
483 | '(' -> return lexbuf ("LPAREN", "")
484 | ')' -> return lexbuf ("RPAREN", "")
485 | eof -> return_eoi lexbuf
486 | _ -> return_symbol lexbuf (Ulexing.utf8_lexeme lexbuf)
488 let level1_pattern_token = ligatures_token level1_pattern_token
489 let level2_ast_token loctable status =
490 ligatures_token (level2_ast_token loctable status)
492 (* API implementation *)
494 level1_pattern_lexer : (string * string) Token.glexer;
495 level2_ast_lexer : (string * string) Token.glexer;
496 level2_meta_lexer : (string * string) Token.glexer;
499 let initial_keywords =
500 [ "CProp"; "Prop"; "Type"; "Set"; "let"; "match";
501 "with"; "in"; "and"; "to"; "as"; "on"; "return"; "done" ]
504 let mk_lexers loctable keywords =
505 prerr_endline ("mk_lexers keywords: " ^ (String.concat ", " keywords));
507 List.fold_right StringSet.add (initial_keywords @ keywords) StringSet.empty
510 level1_pattern_lexer = mk_lexer level1_pattern_token;
511 level2_ast_lexer = mk_lexer (level2_ast_token loctable status);
512 level2_meta_lexer = mk_lexer level2_meta_token;