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 nlet_rec = "nlet" utf8_blank+ "rec"
56 let regexp nlet_corec = "nlet" utf8_blank+ "corec"
57 let regexp ident_decoration = '\'' | '?' | '`'
58 let regexp ident_cont = ident_letter | xml_digit | '_'
59 let regexp ident = ident_letter ident_cont* ident_decoration*
61 let regexp tex_token = '\\' ident
63 let regexp delim_begin = "\\["
64 let regexp delim_end = "\\]"
66 let regexp qkeyword = "'" ident "'"
68 let regexp implicit = '?'
69 let regexp placeholder = '%'
70 let regexp meta = implicit number
72 let regexp csymbol = '\'' ident
74 let regexp begin_group = "@{" | "${"
75 let regexp end_group = '}'
76 let regexp wildcard = "$_"
77 let regexp ast_ident = "@" ident
78 let regexp ast_csymbol = "@" csymbol
79 let regexp meta_ident = "$" ident
80 let regexp meta_anonymous = "$_"
81 let regexp qstring = '"' [^ '"']* '"'
83 let regexp begincomment = "(**" utf8_blank
84 let regexp beginnote = "(*"
85 let regexp endcomment = "*)"
86 (* let regexp comment_char = [^'*'] | '*'[^')']
87 let regexp note = "|+" ([^'*'] | "**") comment_char* "+|" *)
92 "over"; "atop"; "frac";
93 "sqrt"; "root"; "mstyle" ; "mpadded"; "maction"
98 [ "hbox"; "hvbox"; "hovbox"; "vbox";
100 "list0"; "list1"; "sep";
102 "term"; "ident"; "number";
105 let level2_meta_keywords =
106 [ "if"; "then"; "elCicNotationParser.se";
107 "fold"; "left"; "right"; "rec";
110 "anonymous"; "ident"; "number"; "term"; "fresh"
113 (* (string, unit) Hashtbl.t, to exploit multiple bindings *)
114 let initial_level2_ast_keywords () = Hashtbl.create 23;;
116 let level2_ast_keywords = ref (initial_level2_ast_keywords ())
118 let initialize_keywords () =
119 List.iter (fun k -> Hashtbl.add !level2_ast_keywords k ())
120 [ "CProp"; "Prop"; "Type"; "Set"; "let"; "match";
121 "with"; "in"; "and"; "to"; "as"; "on"; "return"; "done" ]
124 let _ = initialize_keywords ();;
126 let add_level2_ast_keyword k = Hashtbl.add !level2_ast_keywords k ()
127 let remove_level2_ast_keyword k = Hashtbl.remove !level2_ast_keywords k
129 (* (string, int) Hashtbl.t, with multiple bindings.
130 * int is the unicode codepoint *)
131 let ligatures = Hashtbl.create 23
135 (fun (ligature, symbol) -> Hashtbl.add ligatures ligature symbol)
136 [ ("->", <:unicode<to>>); ("=>", <:unicode<Rightarrow>>);
137 (":=", <:unicode<def>>);
140 let regexp uri_step = [ 'a' - 'z' 'A' - 'Z' '0' - '9' '_' '-' ''' ]+
143 ("cic:/" | "theory:/") (* schema *)
144 (* ident ('/' ident)* |+ path +| *)
145 uri_step ('/' uri_step)* (* path *)
146 ('.' ident)+ (* ext *)
147 ("#xpointer(" number ('/' number)+ ")")? (* xpointer *)
149 let error lexbuf msg =
150 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
151 raise (Error (begin_cnum, end_cnum, msg))
152 let error_at_end lexbuf msg =
153 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
154 raise (Error (begin_cnum, end_cnum, msg))
156 let return_with_loc token begin_cnum end_cnum =
157 let flocation = HExtlib.floc_of_loc (begin_cnum,end_cnum) in
160 let return lexbuf token =
161 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
162 return_with_loc token begin_cnum end_cnum
164 let return_lexeme lexbuf name = return lexbuf (name, Ulexing.utf8_lexeme lexbuf)
166 let return_symbol lexbuf s = return lexbuf ("SYMBOL", s)
167 let return_eoi lexbuf = return lexbuf ("EOI", "")
169 let remove_quotes s = String.sub s 1 (String.length s - 2)
172 let tok_func stream =
173 (* let lexbuf = Ulexing.from_utf8_stream stream in *)
174 (** XXX Obj.magic rationale.
176 * camlp5 constraints the tok_func field of Token.glexer to have type:
177 * Stream.t char -> (Stream.t 'te * flocation_function)
178 * In order to use ulex we have (in theory) to instantiate a new lexbuf each
179 * time a char Stream.t is passed, destroying the previous lexbuf which may
180 * have consumed a character from the old stream which is lost forever :-(
182 * Instead of passing to camlp5 a char Stream.t we pass a lexbuf, casting it to
183 * char Stream.t with Obj.magic where needed.
185 let lexbuf = Obj.magic stream in
186 Token.make_stream_and_location
191 | Ulexing.Error -> error_at_end lexbuf "Unexpected character"
192 | Ulexing.InvalidCodepoint p ->
193 error_at_end lexbuf (sprintf "Invalid code point: %d" p))
196 Token.tok_func = tok_func;
197 Token.tok_using = (fun _ -> ());
198 Token.tok_removing = (fun _ -> ());
199 Token.tok_match = Token.default_match;
200 Token.tok_text = Token.lexer_text;
201 Token.tok_comm = None;
204 let expand_macro lexbuf =
206 Ulexing.utf8_sub_lexeme lexbuf 1 (Ulexing.lexeme_length lexbuf - 1)
209 ("SYMBOL", Utf8Macro.expand macro)
210 with Utf8Macro.Macro_not_found _ -> "SYMBOL", Ulexing.utf8_lexeme lexbuf
212 let remove_quotes s = String.sub s 1 (String.length s - 2)
213 let remove_left_quote s = String.sub s 1 (String.length s - 1)
215 let rec level2_pattern_token_group counter buffer =
218 if (counter > 0) then
219 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
220 snd (Ulexing.loc lexbuf)
222 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
223 ignore (level2_pattern_token_group (counter + 1) buffer lexbuf) ;
224 level2_pattern_token_group counter buffer lexbuf
226 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
227 level2_pattern_token_group counter buffer lexbuf
229 let read_unparsed_group token_name lexbuf =
230 let buffer = Buffer.create 16 in
231 let begin_cnum, _ = Ulexing.loc lexbuf in
232 let end_cnum = level2_pattern_token_group 0 buffer lexbuf in
233 return_with_loc (token_name, Buffer.contents buffer) begin_cnum end_cnum
235 let rec level2_meta_token =
237 | utf8_blank+ -> level2_meta_token lexbuf
239 let s = Ulexing.utf8_lexeme lexbuf in
241 if List.mem s level2_meta_keywords then
242 return lexbuf ("", s)
244 return lexbuf ("IDENT", s)
246 | "@{" -> read_unparsed_group "UNPARSED_AST" lexbuf
248 return lexbuf ("UNPARSED_AST",
249 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
251 return lexbuf ("UNPARSED_AST",
252 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
253 | eof -> return_eoi lexbuf
255 let rec comment_token acc depth =
258 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
259 comment_token acc (depth + 1) lexbuf
261 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
264 else comment_token acc (depth - 1) lexbuf
266 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
267 comment_token acc depth lexbuf
269 (** @param k continuation to be invoked when no ligature has been found *)
270 let rec ligatures_token k =
273 let lexeme = Ulexing.utf8_lexeme lexbuf in
274 (match List.rev (Hashtbl.find_all ligatures lexeme) with
275 | [] -> (* ligature not found, rollback and try default lexer *)
276 Ulexing.rollback lexbuf;
278 | default_lig :: _ -> (* ligatures found, use the default one *)
279 return_symbol lexbuf default_lig)
280 | eof -> return_eoi lexbuf
281 | _ -> (* not a ligature, rollback and try default lexer *)
282 Ulexing.rollback lexbuf;
285 and level2_ast_token =
287 | let_rec -> return lexbuf ("LETREC","")
288 | let_corec -> return lexbuf ("LETCOREC","")
289 | nlet_rec -> return lexbuf ("NLETREC","")
290 | nlet_corec -> return lexbuf ("NLETCOREC","")
291 | we_proved -> return lexbuf ("WEPROVED","")
292 | we_have -> return lexbuf ("WEHAVE","")
293 | utf8_blank+ -> ligatures_token level2_ast_token lexbuf
295 let s = Ulexing.utf8_lexeme lexbuf in
296 return lexbuf ("META", String.sub s 1 (String.length s - 1))
297 | implicit -> return lexbuf ("IMPLICIT", "")
298 | placeholder -> return lexbuf ("PLACEHOLDER", "")
300 let lexeme = Ulexing.utf8_lexeme lexbuf in
301 if Hashtbl.mem !level2_ast_keywords lexeme then
302 return lexbuf ("", lexeme)
304 return lexbuf ("IDENT", lexeme)
305 | number -> return lexbuf ("NUMBER", Ulexing.utf8_lexeme lexbuf)
306 | tex_token -> return lexbuf (expand_macro lexbuf)
307 | uri -> return lexbuf ("URI", Ulexing.utf8_lexeme lexbuf)
309 return lexbuf ("QSTRING", remove_quotes (Ulexing.utf8_lexeme lexbuf))
311 return lexbuf ("CSYMBOL", remove_left_quote (Ulexing.utf8_lexeme lexbuf))
312 | "${" -> read_unparsed_group "UNPARSED_META" lexbuf
313 | "@{" -> read_unparsed_group "UNPARSED_AST" lexbuf
314 | '(' -> return lexbuf ("LPAREN", "")
315 | ')' -> return lexbuf ("RPAREN", "")
317 return lexbuf ("UNPARSED_META",
318 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
319 | meta_anonymous -> return lexbuf ("UNPARSED_META", "anonymous")
321 let _comment = comment_token (Ulexing.utf8_lexeme lexbuf) 0 lexbuf in
323 Ulexing.utf8_sub_lexeme lexbuf 2 (Ulexing.lexeme_length lexbuf - 4)
325 return lexbuf ("NOTE", comment) *)
326 ligatures_token level2_ast_token lexbuf
327 | begincomment -> return lexbuf ("BEGINCOMMENT","")
328 | endcomment -> return lexbuf ("ENDCOMMENT","")
329 | eof -> return_eoi lexbuf
330 | _ -> return_symbol lexbuf (Ulexing.utf8_lexeme lexbuf)
332 and level1_pattern_token =
334 | utf8_blank+ -> ligatures_token level1_pattern_token lexbuf
335 | number -> return lexbuf ("NUMBER", Ulexing.utf8_lexeme lexbuf)
337 let s = Ulexing.utf8_lexeme lexbuf in
339 if List.mem s level1_keywords then
340 return lexbuf ("", s)
342 return lexbuf ("IDENT", s)
344 | color -> return lexbuf ("COLOR", Ulexing.utf8_lexeme lexbuf)
346 return lexbuf ("PERCENTAGE", Ulexing.utf8_lexeme lexbuf)
348 return lexbuf ("FLOATWITHUNIT", Ulexing.utf8_lexeme lexbuf)
349 | tex_token -> return lexbuf (expand_macro lexbuf)
351 return lexbuf ("QKEYWORD", remove_quotes (Ulexing.utf8_lexeme lexbuf))
352 | '(' -> return lexbuf ("LPAREN", "")
353 | ')' -> return lexbuf ("RPAREN", "")
354 | eof -> return_eoi lexbuf
355 | _ -> return_symbol lexbuf (Ulexing.utf8_lexeme lexbuf)
357 let level1_pattern_token = ligatures_token level1_pattern_token
358 let level2_ast_token = ligatures_token level2_ast_token
360 (* API implementation *)
362 let initial_level1_pattern_lexer () = mk_lexer level1_pattern_token
363 let initial_level2_ast_lexer () = mk_lexer level2_ast_token
364 let initial_level2_meta_lexer () = mk_lexer level2_meta_token
367 let level1_pattern_lexer_ref = ref (initial_level1_pattern_lexer ())
368 let level2_ast_lexer_ref = ref (initial_level2_ast_lexer ())
369 let level2_meta_lexer_ref = ref (initial_level2_meta_lexer ())
371 let level1_pattern_lexer () = !level1_pattern_lexer_ref
372 let level2_ast_lexer () = !level2_ast_lexer_ref
373 let level2_meta_lexer () = !level2_meta_lexer_ref
375 let history = ref [];;
379 (!level2_ast_keywords,!level1_pattern_lexer_ref,
380 !level2_ast_lexer_ref,!level2_meta_lexer_ref) :: !history;
381 level2_ast_keywords := initial_level2_ast_keywords ();
382 initialize_keywords ();
383 level1_pattern_lexer_ref := initial_level1_pattern_lexer ();
384 level2_ast_lexer_ref := initial_level2_ast_lexer ();
385 level2_meta_lexer_ref := initial_level2_meta_lexer ();
391 | (kwd,pl,al,ml) :: tl ->
392 level2_ast_keywords := kwd;
393 level1_pattern_lexer_ref := pl;
394 level2_ast_lexer_ref := al;
395 level2_meta_lexer_ref := ml;