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 *)
35 (* ZACK: breaks unicode's binder followed by an ascii letter without blank *)
36 (* let regexp ident_letter = xml_letter *)
38 let regexp ident_letter = [ 'a' - 'z' 'A' - 'Z' ]
40 (* must be in sync with "is_ligature_char" below *)
41 let regexp ligature_char = [ "'`~!?@*()[]<>-+=|:;.,/\"" ]
42 let regexp ligature = ligature_char ligature_char+
44 let is_ligature_char =
45 (* must be in sync with "regexp ligature_char" above *)
46 let chars = "'`~!?@*()[]<>-+=|:;.,/\"" in
49 ignore (String.index chars char);
51 with Not_found -> false))
53 let regexp we_proved = "we" utf8_blank+ "proved"
54 let regexp we_have = "we" utf8_blank+ "have"
55 let regexp let_rec = "let" utf8_blank+ "rec"
56 let regexp let_corec = "let" 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";
97 [ "hbox"; "hvbox"; "hovbox"; "vbox";
99 "list0"; "list1"; "sep";
101 "term"; "ident"; "number"
104 let level2_meta_keywords =
105 [ "if"; "then"; "else";
106 "fold"; "left"; "right"; "rec";
109 "anonymous"; "ident"; "number"; "term"; "fresh"
112 (* (string, unit) Hashtbl.t, to exploit multiple bindings *)
113 let level2_ast_keywords = Hashtbl.create 23
115 List.iter (fun k -> Hashtbl.add level2_ast_keywords k ())
116 [ "CProp"; "Prop"; "Type"; "Set"; "let"; "match";
117 "with"; "in"; "and"; "to"; "as"; "on"; "return"; "done" ]
119 let add_level2_ast_keyword k = Hashtbl.add level2_ast_keywords k ()
120 let remove_level2_ast_keyword k = Hashtbl.remove level2_ast_keywords k
122 (* (string, int) Hashtbl.t, with multiple bindings.
123 * int is the unicode codepoint *)
124 let ligatures = Hashtbl.create 23
127 (fun (ligature, symbol) -> Hashtbl.add ligatures ligature symbol)
128 [ ("->", <:unicode<to>>); ("=>", <:unicode<Rightarrow>>);
129 ("<=", <:unicode<leq>>); (">=", <:unicode<geq>>);
130 ("<>", <:unicode<neq>>); (":=", <:unicode<def>>);
131 ("==", <:unicode<equiv>>);
134 let regexp uri_step = [ 'a' - 'z' 'A' - 'Z' '0' - '9' '_' '-' ''' ]+
137 ("cic:/" | "theory:/") (* schema *)
138 (* ident ('/' ident)* |+ path +| *)
139 uri_step ('/' uri_step)* (* path *)
140 ('.' ident)+ (* ext *)
141 ("#xpointer(" number ('/' number)+ ")")? (* xpointer *)
143 let error lexbuf msg =
144 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
145 raise (Error (begin_cnum, end_cnum, msg))
146 let error_at_end lexbuf msg =
147 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
148 raise (Error (begin_cnum, end_cnum, msg))
150 let return_with_loc token begin_cnum end_cnum =
151 let flocation = HExtlib.floc_of_loc (begin_cnum,end_cnum) in
154 let return lexbuf token =
155 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
156 return_with_loc token begin_cnum end_cnum
158 let return_lexeme lexbuf name = return lexbuf (name, Ulexing.utf8_lexeme lexbuf)
160 let return_symbol lexbuf s = return lexbuf ("SYMBOL", s)
161 let return_eoi lexbuf = return lexbuf ("EOI", "")
163 let remove_quotes s = String.sub s 1 (String.length s - 2)
166 let tok_func stream =
167 (* let lexbuf = Ulexing.from_utf8_stream stream in *)
168 (** XXX Obj.magic rationale.
170 * camlp5 constraints the tok_func field of Token.glexer to have type:
171 * Stream.t char -> (Stream.t 'te * flocation_function)
172 * In order to use ulex we have (in theory) to instantiate a new lexbuf each
173 * time a char Stream.t is passed, destroying the previous lexbuf which may
174 * have consumed a character from the old stream which is lost forever :-(
176 * Instead of passing to camlp5 a char Stream.t we pass a lexbuf, casting it to
177 * char Stream.t with Obj.magic where needed.
179 let lexbuf = Obj.magic stream in
180 Token.make_stream_and_location
185 | Ulexing.Error -> error_at_end lexbuf "Unexpected character"
186 | Ulexing.InvalidCodepoint p ->
187 error_at_end lexbuf (sprintf "Invalid code point: %d" p))
190 Token.tok_func = tok_func;
191 Token.tok_using = (fun _ -> ());
192 Token.tok_removing = (fun _ -> ());
193 Token.tok_match = Token.default_match;
194 Token.tok_text = Token.lexer_text;
195 Token.tok_comm = None;
198 let expand_macro lexbuf =
200 Ulexing.utf8_sub_lexeme lexbuf 1 (Ulexing.lexeme_length lexbuf - 1)
203 ("SYMBOL", Utf8Macro.expand macro)
204 with Utf8Macro.Macro_not_found _ -> "SYMBOL", Ulexing.utf8_lexeme lexbuf
206 let remove_quotes s = String.sub s 1 (String.length s - 2)
207 let remove_left_quote s = String.sub s 1 (String.length s - 1)
209 let rec level2_pattern_token_group counter buffer =
212 if (counter > 0) then
213 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
214 snd (Ulexing.loc lexbuf)
216 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
217 ignore (level2_pattern_token_group (counter + 1) buffer lexbuf) ;
218 level2_pattern_token_group counter buffer lexbuf
220 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
221 level2_pattern_token_group counter buffer lexbuf
223 let read_unparsed_group token_name lexbuf =
224 let buffer = Buffer.create 16 in
225 let begin_cnum, _ = Ulexing.loc lexbuf in
226 let end_cnum = level2_pattern_token_group 0 buffer lexbuf in
227 return_with_loc (token_name, Buffer.contents buffer) begin_cnum end_cnum
229 let rec level2_meta_token =
231 | utf8_blank+ -> level2_meta_token lexbuf
233 let s = Ulexing.utf8_lexeme lexbuf in
235 if List.mem s level2_meta_keywords then
236 return lexbuf ("", s)
238 return lexbuf ("IDENT", s)
240 | "@{" -> read_unparsed_group "UNPARSED_AST" lexbuf
242 return lexbuf ("UNPARSED_AST",
243 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
245 return lexbuf ("UNPARSED_AST",
246 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
247 | eof -> return_eoi lexbuf
249 let rec comment_token acc depth =
252 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
253 comment_token acc (depth + 1) lexbuf
255 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
258 else comment_token acc (depth - 1) lexbuf
260 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
261 comment_token acc depth lexbuf
263 (** @param k continuation to be invoked when no ligature has been found *)
264 let rec ligatures_token k =
267 let lexeme = Ulexing.utf8_lexeme lexbuf in
268 (match List.rev (Hashtbl.find_all ligatures lexeme) with
269 | [] -> (* ligature not found, rollback and try default lexer *)
270 Ulexing.rollback lexbuf;
272 | default_lig :: _ -> (* ligatures found, use the default one *)
273 return_symbol lexbuf default_lig)
274 | eof -> return_eoi lexbuf
275 | _ -> (* not a ligature, rollback and try default lexer *)
276 Ulexing.rollback lexbuf;
279 and level2_ast_token =
281 | let_rec -> return lexbuf ("LETREC","")
282 | let_corec -> return lexbuf ("LETCOREC","")
283 | we_proved -> return lexbuf ("WEPROVED","")
284 | we_have -> return lexbuf ("WEHAVE","")
285 | utf8_blank+ -> ligatures_token level2_ast_token lexbuf
287 let s = Ulexing.utf8_lexeme lexbuf in
288 return lexbuf ("META", String.sub s 1 (String.length s - 1))
289 | implicit -> return lexbuf ("IMPLICIT", "")
290 | placeholder -> return lexbuf ("PLACEHOLDER", "")
292 let lexeme = Ulexing.utf8_lexeme lexbuf in
293 if Hashtbl.mem level2_ast_keywords lexeme then
294 return lexbuf ("", lexeme)
296 return lexbuf ("IDENT", lexeme)
297 | number -> return lexbuf ("NUMBER", Ulexing.utf8_lexeme lexbuf)
298 | tex_token -> return lexbuf (expand_macro lexbuf)
299 | uri -> return lexbuf ("URI", Ulexing.utf8_lexeme lexbuf)
301 return lexbuf ("QSTRING", remove_quotes (Ulexing.utf8_lexeme lexbuf))
303 return lexbuf ("CSYMBOL", remove_left_quote (Ulexing.utf8_lexeme lexbuf))
304 | "${" -> read_unparsed_group "UNPARSED_META" lexbuf
305 | "@{" -> read_unparsed_group "UNPARSED_AST" lexbuf
306 | '(' -> return lexbuf ("LPAREN", "")
307 | ')' -> return lexbuf ("RPAREN", "")
309 return lexbuf ("UNPARSED_META",
310 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
311 | meta_anonymous -> return lexbuf ("UNPARSED_META", "anonymous")
313 let _comment = comment_token (Ulexing.utf8_lexeme lexbuf) 0 lexbuf in
315 Ulexing.utf8_sub_lexeme lexbuf 2 (Ulexing.lexeme_length lexbuf - 4)
317 return lexbuf ("NOTE", comment) *)
318 ligatures_token level2_ast_token lexbuf
319 | begincomment -> return lexbuf ("BEGINCOMMENT","")
320 | endcomment -> return lexbuf ("ENDCOMMENT","")
321 | eof -> return_eoi lexbuf
322 | _ -> return_symbol lexbuf (Ulexing.utf8_lexeme lexbuf)
324 and level1_pattern_token =
326 | utf8_blank+ -> ligatures_token level1_pattern_token lexbuf
327 | number -> return lexbuf ("NUMBER", Ulexing.utf8_lexeme lexbuf)
329 let s = Ulexing.utf8_lexeme lexbuf in
331 if List.mem s level1_keywords then
332 return lexbuf ("", s)
334 return lexbuf ("IDENT", s)
336 | tex_token -> return lexbuf (expand_macro lexbuf)
338 return lexbuf ("QKEYWORD", remove_quotes (Ulexing.utf8_lexeme lexbuf))
339 | '(' -> return lexbuf ("LPAREN", "")
340 | ')' -> return lexbuf ("RPAREN", "")
341 | eof -> return_eoi lexbuf
342 | _ -> return_symbol lexbuf (Ulexing.utf8_lexeme lexbuf)
344 let level1_pattern_token = ligatures_token level1_pattern_token
345 let level2_ast_token = ligatures_token level2_ast_token
347 (* API implementation *)
349 let level1_pattern_lexer = mk_lexer level1_pattern_token
350 let level2_ast_lexer = mk_lexer level2_ast_token
351 let level2_meta_lexer = mk_lexer level2_meta_token
353 let lookup_ligatures lexeme =
356 then [ Utf8Macro.expand (String.sub lexeme 1 (String.length lexeme - 1)) ]
357 else List.rev (Hashtbl.find_all ligatures lexeme)
358 with Invalid_argument _ | Utf8Macro.Macro_not_found _ -> []