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 tagname = ident_letter tag_cont*
86 let regexp opentag = xmarkup tagname
87 let regexp closetag = xmarkup "/" tagname ymarkup
88 let regexp attribute = tagname "=" qstring
89 let regexp generictag =
90 opentag (utf8_blank+ attribute)* ymarkup
92 let regexp tex_token = '\\' ident
94 let regexp delim_begin = "\\["
95 let regexp delim_end = "\\]"
97 let regexp qkeyword = "'" ( ident | pident ) "'"
99 let regexp implicit = '?'
100 let regexp placeholder = '%'
101 let regexp meta = implicit number
103 let regexp csymbol = '\'' ident
105 let regexp begin_group = "@{" | "${"
106 let regexp end_group = '}'
107 let regexp wildcard = "$_"
108 let regexp ast_ident = "@" ident
109 let regexp ast_csymbol = "@" csymbol
110 let regexp meta_ident = "$" ident
111 let regexp meta_anonymous = "$_"
113 let regexp begincomment = "(**" utf8_blank
114 let regexp beginnote = "(*"
115 let regexp endcomment = "*)"
116 (* let regexp comment_char = [^'*'] | '*'[^')']
117 let regexp note = "|+" ([^'*'] | "**") comment_char* "+|" *)
122 "over"; "atop"; "frac";
123 "sqrt"; "root"; "mstyle" ; "mpadded"; "maction"
127 let level1_keywords =
128 [ "hbox"; "hvbox"; "hovbox"; "vbox";
130 "list0"; "list1"; "sep";
132 "term"; "ident"; "number";
135 let level2_meta_keywords =
136 [ "if"; "then"; "elCicNotationParser.se";
137 "fold"; "left"; "right"; "rec";
140 "anonymous"; "ident"; "number"; "term"; "fresh"
143 (* (string, int) Hashtbl.t, with multiple bindings.
144 * int is the unicode codepoint *)
145 let ligatures = Hashtbl.create 23
149 (fun (ligature, symbol) -> Hashtbl.add ligatures ligature symbol)
150 [ ("->", <:unicode<to>>); ("=>", <:unicode<Rightarrow>>);
151 (":=", <:unicode<def>>);
154 let regexp nreference =
156 uri_step ('/' uri_step)* (* path *)
159 | "def" "(" number ")"
160 | "fix" "(" number "," number "," number ")"
161 | "cfx" "(" number ")"
162 | "ind" "(" number "," number "," number ")"
163 | "con" "(" number "," number "," number ")") (* ext + reference *)
165 let error lexbuf msg =
166 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
167 raise (Error (begin_cnum, end_cnum, msg))
168 let error_at_end lexbuf msg =
169 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
170 raise (Error (begin_cnum, end_cnum, msg))
172 let loc_of_buf lexbuf =
173 HExtlib.floc_of_loc (Ulexing.loc lexbuf)
175 let return_with_loc token begin_cnum end_cnum =
176 let flocation = HExtlib.floc_of_loc (begin_cnum,end_cnum) in
179 let return lexbuf token =
180 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
181 return_with_loc token begin_cnum end_cnum
183 let return_lexeme lexbuf name = return lexbuf (name, Ulexing.utf8_lexeme lexbuf)
185 let return_symbol lexbuf s = return lexbuf ("SYMBOL", s)
186 let return_eoi lexbuf = return lexbuf ("EOI", "")
188 let remove_quotes s = String.sub s 1 (String.length s - 2)
191 let tok_func stream =
192 (* let lexbuf = Ulexing.from_utf8_stream stream in *)
193 (** XXX Obj.magic rationale.
195 * camlp5 constraints the tok_func field of Token.glexer to have type:
196 * Stream.t char -> (Stream.t 'te * flocation_function)
197 * In order to use ulex we have (in theory) to instantiate a new lexbuf each
198 * time a char Stream.t is passed, destroying the previous lexbuf which may
199 * have consumed a character from the old stream which is lost forever :-(
201 * Instead of passing to camlp5 a char Stream.t we pass a lexbuf, casting it to
202 * char Stream.t with Obj.magic where needed.
204 let lexbuf = Obj.magic stream in
205 Token.make_stream_and_location
210 | Ulexing.Error -> error_at_end lexbuf "Unexpected character"
211 | Ulexing.InvalidCodepoint p ->
212 error_at_end lexbuf (sprintf "Invalid code point: %d" p))
215 Token.tok_func = tok_func;
216 Token.tok_using = (fun _ -> ());
217 Token.tok_removing = (fun _ -> ());
218 Token.tok_match = Token.default_match;
219 Token.tok_text = Token.lexer_text;
220 Token.tok_comm = None;
223 let expand_macro lexbuf =
225 Ulexing.utf8_sub_lexeme lexbuf 1 (Ulexing.lexeme_length lexbuf - 1)
228 ("SYMBOL", Utf8Macro.expand macro)
229 with Utf8Macro.Macro_not_found _ ->
230 (* FG: unexpanded TeX macros are terminated by a space for rendering *)
231 "SYMBOL", (Ulexing.utf8_lexeme lexbuf ^ " ")
233 let remove_quotes s = String.sub s 1 (String.length s - 2)
234 let remove_left_quote s = String.sub s 1 (String.length s - 1)
236 let rec level2_pattern_token_group counter buffer =
239 if (counter > 0) then
240 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
241 snd (Ulexing.loc lexbuf)
243 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
244 ignore (level2_pattern_token_group (counter + 1) buffer lexbuf) ;
245 level2_pattern_token_group counter buffer lexbuf
247 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
248 level2_pattern_token_group counter buffer lexbuf
250 let read_unparsed_group token_name lexbuf =
251 let buffer = Buffer.create 16 in
252 let begin_cnum, _ = Ulexing.loc lexbuf in
253 let end_cnum = level2_pattern_token_group 0 buffer lexbuf in
254 return_with_loc (token_name, Buffer.contents buffer) begin_cnum end_cnum
256 let handle_keywords lexbuf k name =
257 let s = Ulexing.utf8_lexeme lexbuf in
259 return lexbuf ("", s)
261 return lexbuf (name, s)
264 let rec level2_meta_token =
266 | utf8_blank+ -> level2_meta_token lexbuf
267 | hreftag -> return lexbuf ("ATAG","")
268 | hrefclose -> return lexbuf ("ATAGEND","")
270 | closetag -> level2_meta_token lexbuf
272 handle_keywords lexbuf (fun x -> List.mem x level2_meta_keywords) "IDENT"
273 | variable_ident -> return lexbuf ("IDENT", Ulexing.utf8_lexeme lexbuf)
275 handle_keywords lexbuf (fun x -> List.mem x level2_meta_keywords) "PIDENT"
276 | "@{" -> read_unparsed_group "UNPARSED_AST" lexbuf
278 return lexbuf ("UNPARSED_AST",
279 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
281 return lexbuf ("UNPARSED_AST",
282 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
283 | eof -> return_eoi lexbuf
285 let rec comment_token acc depth =
288 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
289 comment_token acc (depth + 1) lexbuf
291 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
294 else comment_token acc (depth - 1) lexbuf
296 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
297 comment_token acc depth lexbuf
299 (** @param k continuation to be invoked when no ligature has been found *)
300 let ligatures_token k =
303 let lexeme = Ulexing.utf8_lexeme lexbuf in
304 (match List.rev (Hashtbl.find_all ligatures lexeme) with
305 | [] -> (* ligature not found, rollback and try default lexer *)
306 Ulexing.rollback lexbuf;
308 | default_lig :: _ -> (* ligatures found, use the default one *)
309 return_symbol lexbuf default_lig)
310 | eof -> return_eoi lexbuf
311 | _ -> (* not a ligature, rollback and try default lexer *)
312 Ulexing.rollback lexbuf;
317 type t = Stdpp.location
318 let compare = Pervasives.compare
321 module LocalizeEnv = Map.Make (LocalizeOD)
328 let update_table loc desc href loctable =
329 if desc <> None || href <> None
331 (let s,e = HExtlib.loc_of_floc loc in
332 prerr_endline (Printf.sprintf "*** [%d,%d] \"%s\",\"%s\""
333 s e (so_pp href) (so_pp desc));
334 LocalizeEnv.add loc (href,desc) loctable)
338 let level2_ast_token loctable status =
339 let rec aux desc href =
341 | let_rec -> return lexbuf ("LETREC","")
342 | let_corec -> return lexbuf ("LETCOREC","")
343 | we_proved -> return lexbuf ("WEPROVED","")
344 | we_have -> return lexbuf ("WEHAVE","")
345 | utf8_blank+ -> ligatures_token (aux desc href) lexbuf
347 let s = Ulexing.utf8_lexeme lexbuf in
348 return lexbuf ("META", String.sub s 1 (String.length s - 1))
349 | implicit -> return lexbuf ("IMPLICIT", "")
350 | placeholder -> return lexbuf ("PLACEHOLDER", "")
351 | hreftag -> aux_in_tag None None lexbuf
352 | hrefclose -> aux None None lexbuf
354 | closetag -> ligatures_token (aux desc href) lexbuf
355 | ident -> loctable :=
356 update_table (loc_of_buf lexbuf) desc href !loctable;
357 handle_keywords lexbuf (fun x -> StringSet.mem x status) "IDENT"
359 return lexbuf ("IDENT", Ulexing.utf8_lexeme lexbuf)
360 | pident -> loctable :=
361 update_table (loc_of_buf lexbuf) desc href !loctable;
362 handle_keywords lexbuf (fun x -> StringSet.mem x status) "PIDENT"
363 | number -> loctable :=
364 update_table (loc_of_buf lexbuf) desc href !loctable;
365 return lexbuf ("NUMBER", Ulexing.utf8_lexeme lexbuf)
366 | tex_token -> loctable :=
367 update_table (loc_of_buf lexbuf) desc href !loctable;
368 return lexbuf (expand_macro lexbuf)
369 | nreference -> return lexbuf ("NREF", Ulexing.utf8_lexeme lexbuf)
370 | uri -> return lexbuf ("URI", Ulexing.utf8_lexeme lexbuf)
372 return lexbuf ("QSTRING", remove_quotes (Ulexing.utf8_lexeme lexbuf))
374 return lexbuf ("CSYMBOL", remove_left_quote (Ulexing.utf8_lexeme lexbuf))
375 | "${" -> read_unparsed_group "UNPARSED_META" lexbuf
376 | "@{" -> read_unparsed_group "UNPARSED_AST" lexbuf
377 | '(' -> return lexbuf ("LPAREN", "")
378 | ')' -> return lexbuf ("RPAREN", "")
380 return lexbuf ("UNPARSED_META",
381 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
382 | meta_anonymous -> return lexbuf ("UNPARSED_META", "anonymous")
384 let _comment = comment_token (Ulexing.utf8_lexeme lexbuf) 0 lexbuf in
386 Ulexing.utf8_sub_lexeme lexbuf 2 (Ulexing.lexeme_length lexbuf - 4)
388 return lexbuf ("NOTE", comment) *)
389 ligatures_token (aux desc href) lexbuf
390 | begincomment -> return lexbuf ("BEGINCOMMENT","")
391 | endcomment -> return lexbuf ("ENDCOMMENT","")
392 | eof -> return_eoi lexbuf
394 update_table (loc_of_buf lexbuf) desc href !loctable;
395 return_symbol lexbuf (Ulexing.utf8_lexeme lexbuf)
397 and aux_in_tag desc href = lexer
398 | utf8_blank+ -> ligatures_token (aux_in_tag desc href) lexbuf
401 (Some (Ulexing.utf8_sub_lexeme lexbuf 6 (Ulexing.lexeme_length lexbuf - 7)))
405 (Some (Ulexing.utf8_sub_lexeme lexbuf 7 (Ulexing.lexeme_length lexbuf - 8)))
407 | ">" -> aux desc href lexbuf
408 | _ -> aux None None lexbuf
411 let rec level1_pattern_token =
413 | utf8_blank+ -> ligatures_token level1_pattern_token lexbuf
414 | number -> return lexbuf ("NUMBER", Ulexing.utf8_lexeme lexbuf)
415 | hreftag -> return lexbuf ("ATAG", "")
416 | hrefclose -> return lexbuf ("ATAGEND","")
418 | closetag -> ligatures_token level1_pattern_token lexbuf
419 | ident ->handle_keywords lexbuf (fun x -> List.mem x level1_keywords) "IDENT"
420 | variable_ident -> return lexbuf ("IDENT", Ulexing.utf8_lexeme lexbuf)
421 | pident->handle_keywords lexbuf (fun x->List.mem x level1_keywords) "PIDENT"
422 | color -> return lexbuf ("COLOR", Ulexing.utf8_lexeme lexbuf)
424 return lexbuf ("PERCENTAGE", Ulexing.utf8_lexeme lexbuf)
426 return lexbuf ("FLOATWITHUNIT", Ulexing.utf8_lexeme lexbuf)
427 | tex_token -> return lexbuf (expand_macro lexbuf)
429 return lexbuf ("QKEYWORD", remove_quotes (Ulexing.utf8_lexeme lexbuf))
430 | '(' -> return lexbuf ("LPAREN", "")
431 | ')' -> return lexbuf ("RPAREN", "")
432 | eof -> return_eoi lexbuf
433 | _ -> return_symbol lexbuf (Ulexing.utf8_lexeme lexbuf)
435 let level1_pattern_token = ligatures_token level1_pattern_token
436 let level2_ast_token loctable status =
437 ligatures_token (level2_ast_token loctable status)
439 (* API implementation *)
441 level1_pattern_lexer : (string * string) Token.glexer;
442 level2_ast_lexer : (string * string) Token.glexer;
443 level2_meta_lexer : (string * string) Token.glexer
446 let mk_lexers loctable keywords =
447 let initial_keywords =
448 [ "CProp"; "Prop"; "Type"; "Set"; "let"; "match";
449 "with"; "in"; "and"; "to"; "as"; "on"; "return"; "done" ]
452 List.fold_right StringSet.add (initial_keywords @ keywords) StringSet.empty
455 level1_pattern_lexer = mk_lexer level1_pattern_token;
456 level2_ast_lexer = mk_lexer (level2_ast_token loctable status);
457 level2_meta_lexer = mk_lexer level2_meta_token