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 let regexp number = xml_digit+
33 let regexp utf8_blank = " " | "\r\n" | "\n" | "\t" | [160] (* this is a nbsp *)
34 let regexp percentage =
35 ('-' | "") [ '0' - '9' ] + '%'
36 let regexp floatwithunit =
37 ('-' | "") [ '0' - '9' ] + ["."] [ '0' - '9' ] + ([ 'a' - 'z' ] + | "" )
38 let regexp color = "#" [ '0' - '9' 'a' - 'f' 'A' - 'F' ] [ '0' - '9' 'a' - 'f'
39 'A' - 'F' ] [ '0' - '9' 'a' - 'f' 'A' - 'F' ] [ '0' - '9' 'a' - 'f' 'A' - 'F' ]
40 [ '0' - '9' 'a' - 'f' 'A' - 'F' ] [ '0' - '9' 'a' - 'f' 'A' - 'F' ]
42 (* ZACK: breaks unicode's binder followed by an ascii letter without blank *)
43 (* let regexp ident_letter = xml_letter *)
45 let regexp ident_letter = [ 'a' - 'z' 'A' - 'Z' ]
47 (* must be in sync with "is_ligature_char" below *)
48 let regexp ligature_char = [ "'`~!?@*()[]<>-+=|:;.,/\"" ]
49 let regexp ligature = ligature_char ligature_char+
51 let regexp we_proved = "we" utf8_blank+ "proved"
52 let regexp we_have = "we" utf8_blank+ "have"
53 let regexp let_rec = "let" utf8_blank+ "rec"
54 let regexp let_corec = "let" utf8_blank+ "corec"
55 let regexp ident_decoration = '\'' | '?' | '`'
56 let regexp ident_cont = ident_letter | xml_digit | '_'
57 let regexp ident_start = ident_letter
58 let regexp ident = ident_letter ident_cont* ident_decoration*
59 let regexp variable_ident = '_' '_' number
60 let regexp pident = '_' ident
62 let regexp tex_token = '\\' ident
64 let regexp delim_begin = "\\["
65 let regexp delim_end = "\\]"
67 let regexp qkeyword = "'" ( ident | pident ) "'"
69 let regexp implicit = '?'
70 let regexp placeholder = '%'
71 let regexp meta = implicit number
73 let regexp csymbol = '\'' ident
75 let regexp begin_group = "@{" | "${"
76 let regexp end_group = '}'
77 let regexp wildcard = "$_"
78 let regexp ast_ident = "@" ident
79 let regexp ast_csymbol = "@" csymbol
80 let regexp meta_ident = "$" ident
81 let regexp meta_anonymous = "$_"
82 let regexp qstring = '"' [^ '"']* '"'
84 let regexp begincomment = "(**" utf8_blank
85 let regexp beginnote = "(*"
86 let regexp endcomment = "*)"
87 (* let regexp comment_char = [^'*'] | '*'[^')']
88 let regexp note = "|+" ([^'*'] | "**") comment_char* "+|" *)
93 "over"; "atop"; "frac";
94 "sqrt"; "root"; "mstyle" ; "mpadded"; "maction"
99 [ "hbox"; "hvbox"; "hovbox"; "vbox";
101 "list0"; "list1"; "sep";
103 "term"; "ident"; "number";
106 let level2_meta_keywords =
107 [ "if"; "then"; "elCicNotationParser.se";
108 "fold"; "left"; "right"; "rec";
111 "anonymous"; "ident"; "number"; "term"; "fresh"
114 (* (string, unit) Hashtbl.t, to exploit multiple bindings *)
115 let initial_level2_ast_keywords () = Hashtbl.create 23;;
117 let level2_ast_keywords = ref (initial_level2_ast_keywords ())
119 let initialize_keywords () =
120 List.iter (fun k -> Hashtbl.add !level2_ast_keywords k ())
121 [ "CProp"; "Prop"; "Type"; "Set"; "let"; "match";
122 "with"; "in"; "and"; "to"; "as"; "on"; "return"; "done" ]
125 let _ = initialize_keywords ();;
127 let add_level2_ast_keyword k = Hashtbl.add !level2_ast_keywords k ()
128 let remove_level2_ast_keyword k = Hashtbl.remove !level2_ast_keywords k
130 (* (string, int) Hashtbl.t, with multiple bindings.
131 * int is the unicode codepoint *)
132 let ligatures = Hashtbl.create 23
136 (fun (ligature, symbol) -> Hashtbl.add ligatures ligature symbol)
137 [ ("->", <:unicode<to>>); ("=>", <:unicode<Rightarrow>>);
138 (":=", <:unicode<def>>);
141 let regexp uri_step = [ 'a' - 'z' 'A' - 'Z' '0' - '9' '_' '-' ''' ]+
144 ("cic:/" | "theory:/") (* schema *)
145 (* ident ('/' ident)* |+ path +| *)
146 uri_step ('/' uri_step)* (* path *)
147 ('.' ident)+ (* ext *)
148 ("#xpointer(" number ('/' number)+ ")")? (* xpointer *)
150 let regexp nreference =
152 uri_step ('/' uri_step)* (* path *)
155 | "def" "(" number ")"
156 | "fix" "(" number "," number "," number ")"
157 | "cfx" "(" number ")"
158 | "ind" "(" number "," number "," number ")"
159 | "con" "(" number "," number "," number ")") (* ext + reference *)
161 let error lexbuf msg =
162 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
163 raise (Error (begin_cnum, end_cnum, msg))
164 let error_at_end lexbuf msg =
165 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
166 raise (Error (begin_cnum, end_cnum, msg))
168 let return_with_loc token begin_cnum end_cnum =
169 let flocation = HExtlib.floc_of_loc (begin_cnum,end_cnum) in
172 let return lexbuf token =
173 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
174 return_with_loc token begin_cnum end_cnum
176 let return_lexeme lexbuf name = return lexbuf (name, Ulexing.utf8_lexeme lexbuf)
178 let return_symbol lexbuf s = return lexbuf ("SYMBOL", s)
179 let return_eoi lexbuf = return lexbuf ("EOI", "")
181 let remove_quotes s = String.sub s 1 (String.length s - 2)
184 let tok_func stream =
185 (* let lexbuf = Ulexing.from_utf8_stream stream in *)
186 (** XXX Obj.magic rationale.
188 * camlp5 constraints the tok_func field of Token.glexer to have type:
189 * Stream.t char -> (Stream.t 'te * flocation_function)
190 * In order to use ulex we have (in theory) to instantiate a new lexbuf each
191 * time a char Stream.t is passed, destroying the previous lexbuf which may
192 * have consumed a character from the old stream which is lost forever :-(
194 * Instead of passing to camlp5 a char Stream.t we pass a lexbuf, casting it to
195 * char Stream.t with Obj.magic where needed.
197 let lexbuf = Obj.magic stream in
198 Token.make_stream_and_location
203 | Ulexing.Error -> error_at_end lexbuf "Unexpected character"
204 | Ulexing.InvalidCodepoint p ->
205 error_at_end lexbuf (sprintf "Invalid code point: %d" p))
208 Token.tok_func = tok_func;
209 Token.tok_using = (fun _ -> ());
210 Token.tok_removing = (fun _ -> ());
211 Token.tok_match = Token.default_match;
212 Token.tok_text = Token.lexer_text;
213 Token.tok_comm = None;
216 let expand_macro lexbuf =
218 Ulexing.utf8_sub_lexeme lexbuf 1 (Ulexing.lexeme_length lexbuf - 1)
221 ("SYMBOL", Utf8Macro.expand macro)
222 with Utf8Macro.Macro_not_found _ ->
223 (* FG: unexpanded TeX macros are terminated by a space for rendering *)
224 "SYMBOL", (Ulexing.utf8_lexeme lexbuf ^ " ")
226 let remove_quotes s = String.sub s 1 (String.length s - 2)
227 let remove_left_quote s = String.sub s 1 (String.length s - 1)
229 let rec level2_pattern_token_group counter buffer =
232 if (counter > 0) then
233 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
234 snd (Ulexing.loc lexbuf)
236 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
237 ignore (level2_pattern_token_group (counter + 1) buffer lexbuf) ;
238 level2_pattern_token_group counter buffer lexbuf
240 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
241 level2_pattern_token_group counter buffer lexbuf
243 let read_unparsed_group token_name lexbuf =
244 let buffer = Buffer.create 16 in
245 let begin_cnum, _ = Ulexing.loc lexbuf in
246 let end_cnum = level2_pattern_token_group 0 buffer lexbuf in
247 return_with_loc (token_name, Buffer.contents buffer) begin_cnum end_cnum
249 let handle_keywords lexbuf k name =
250 let s = Ulexing.utf8_lexeme lexbuf in
252 return lexbuf ("", s)
254 return lexbuf (name, s)
257 let rec level2_meta_token =
259 | utf8_blank+ -> level2_meta_token lexbuf
261 handle_keywords lexbuf (fun x -> List.mem x level2_meta_keywords) "IDENT"
262 | variable_ident -> return lexbuf ("IDENT", Ulexing.utf8_lexeme lexbuf)
264 handle_keywords lexbuf (fun x -> List.mem x level2_meta_keywords) "PIDENT"
265 | "@{" -> read_unparsed_group "UNPARSED_AST" lexbuf
267 return lexbuf ("UNPARSED_AST",
268 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
270 return lexbuf ("UNPARSED_AST",
271 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
272 | eof -> return_eoi lexbuf
274 let rec comment_token acc depth =
277 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
278 comment_token acc (depth + 1) lexbuf
280 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
283 else comment_token acc (depth - 1) lexbuf
285 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
286 comment_token acc depth lexbuf
288 (** @param k continuation to be invoked when no ligature has been found *)
289 let rec ligatures_token k =
292 let lexeme = Ulexing.utf8_lexeme lexbuf in
293 (match List.rev (Hashtbl.find_all ligatures lexeme) with
294 | [] -> (* ligature not found, rollback and try default lexer *)
295 Ulexing.rollback lexbuf;
297 | default_lig :: _ -> (* ligatures found, use the default one *)
298 return_symbol lexbuf default_lig)
299 | eof -> return_eoi lexbuf
300 | _ -> (* not a ligature, rollback and try default lexer *)
301 Ulexing.rollback lexbuf;
304 and level2_ast_token =
306 | let_rec -> return lexbuf ("LETREC","")
307 | let_corec -> return lexbuf ("LETCOREC","")
308 | we_proved -> return lexbuf ("WEPROVED","")
309 | we_have -> return lexbuf ("WEHAVE","")
310 | utf8_blank+ -> ligatures_token level2_ast_token lexbuf
312 let s = Ulexing.utf8_lexeme lexbuf in
313 return lexbuf ("META", String.sub s 1 (String.length s - 1))
314 | implicit -> return lexbuf ("IMPLICIT", "")
315 | placeholder -> return lexbuf ("PLACEHOLDER", "")
316 | ident -> handle_keywords lexbuf (Hashtbl.mem !level2_ast_keywords) "IDENT"
317 | variable_ident -> return lexbuf ("IDENT", Ulexing.utf8_lexeme lexbuf)
318 | pident -> handle_keywords lexbuf (Hashtbl.mem !level2_ast_keywords) "PIDENT"
319 | number -> return lexbuf ("NUMBER", Ulexing.utf8_lexeme lexbuf)
320 | tex_token -> return lexbuf (expand_macro lexbuf)
321 | nreference -> return lexbuf ("NREF", Ulexing.utf8_lexeme lexbuf)
322 | uri -> return lexbuf ("URI", Ulexing.utf8_lexeme lexbuf)
324 return lexbuf ("QSTRING", remove_quotes (Ulexing.utf8_lexeme lexbuf))
326 return lexbuf ("CSYMBOL", remove_left_quote (Ulexing.utf8_lexeme lexbuf))
327 | "${" -> read_unparsed_group "UNPARSED_META" lexbuf
328 | "@{" -> read_unparsed_group "UNPARSED_AST" lexbuf
329 | '(' -> return lexbuf ("LPAREN", "")
330 | ')' -> return lexbuf ("RPAREN", "")
332 return lexbuf ("UNPARSED_META",
333 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
334 | meta_anonymous -> return lexbuf ("UNPARSED_META", "anonymous")
336 let _comment = comment_token (Ulexing.utf8_lexeme lexbuf) 0 lexbuf in
338 Ulexing.utf8_sub_lexeme lexbuf 2 (Ulexing.lexeme_length lexbuf - 4)
340 return lexbuf ("NOTE", comment) *)
341 ligatures_token level2_ast_token lexbuf
342 | begincomment -> return lexbuf ("BEGINCOMMENT","")
343 | endcomment -> return lexbuf ("ENDCOMMENT","")
344 | eof -> return_eoi lexbuf
345 | _ -> return_symbol lexbuf (Ulexing.utf8_lexeme lexbuf)
347 and level1_pattern_token =
349 | utf8_blank+ -> ligatures_token level1_pattern_token lexbuf
350 | number -> return lexbuf ("NUMBER", Ulexing.utf8_lexeme lexbuf)
351 | ident ->handle_keywords lexbuf (fun x -> List.mem x level1_keywords) "IDENT"
352 | variable_ident -> return lexbuf ("IDENT", Ulexing.utf8_lexeme lexbuf)
353 | pident->handle_keywords lexbuf (fun x->List.mem x level1_keywords) "PIDENT"
354 | color -> return lexbuf ("COLOR", Ulexing.utf8_lexeme lexbuf)
356 return lexbuf ("PERCENTAGE", Ulexing.utf8_lexeme lexbuf)
358 return lexbuf ("FLOATWITHUNIT", Ulexing.utf8_lexeme lexbuf)
359 | tex_token -> return lexbuf (expand_macro lexbuf)
361 return lexbuf ("QKEYWORD", remove_quotes (Ulexing.utf8_lexeme lexbuf))
362 | '(' -> return lexbuf ("LPAREN", "")
363 | ')' -> return lexbuf ("RPAREN", "")
364 | eof -> return_eoi lexbuf
365 | _ -> return_symbol lexbuf (Ulexing.utf8_lexeme lexbuf)
367 let level1_pattern_token = ligatures_token level1_pattern_token
368 let level2_ast_token = ligatures_token level2_ast_token
370 (* API implementation *)
372 let initial_level1_pattern_lexer () = mk_lexer level1_pattern_token
373 let initial_level2_ast_lexer () = mk_lexer level2_ast_token
374 let initial_level2_meta_lexer () = mk_lexer level2_meta_token
377 let level1_pattern_lexer_ref = ref (initial_level1_pattern_lexer ())
378 let level2_ast_lexer_ref = ref (initial_level2_ast_lexer ())
379 let level2_meta_lexer_ref = ref (initial_level2_meta_lexer ())
381 let level1_pattern_lexer () = !level1_pattern_lexer_ref
382 let level2_ast_lexer () = !level2_ast_lexer_ref
383 let level2_meta_lexer () = !level2_meta_lexer_ref
385 let history = ref [];;
389 (!level2_ast_keywords,!level1_pattern_lexer_ref,
390 !level2_ast_lexer_ref,!level2_meta_lexer_ref) :: !history;
391 level2_ast_keywords := initial_level2_ast_keywords ();
392 initialize_keywords ();
393 level1_pattern_lexer_ref := initial_level1_pattern_lexer ();
394 level2_ast_lexer_ref := initial_level2_ast_lexer ();
395 level2_meta_lexer_ref := initial_level2_meta_lexer ();
401 | (kwd,pl,al,ml) :: tl ->
402 level2_ast_keywords := kwd;
403 level1_pattern_lexer_ref := pl;
404 level2_ast_lexer_ref := al;
405 level2_meta_lexer_ref := ml;