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 floatwithunit =
35 [ '0' - '9' ] + ["."] [ '0' - '9' ] + ([ 'a' - 'z' ] + | "" )
36 let regexp color = "#" [ '0' - '9' 'a' - 'f' 'A' - 'F' ] [ '0' - '9' 'a' - 'f'
37 'A' - 'F' ] [ '0' - '9' 'a' - 'f' 'A' - 'F' ] [ '0' - '9' 'a' - 'f' 'A' - 'F' ]
38 [ '0' - '9' 'a' - 'f' 'A' - 'F' ] [ '0' - '9' 'a' - 'f' 'A' - 'F' ]
40 (* ZACK: breaks unicode's binder followed by an ascii letter without blank *)
41 (* let regexp ident_letter = xml_letter *)
43 let regexp ident_letter = [ 'a' - 'z' 'A' - 'Z' ]
45 (* must be in sync with "is_ligature_char" below *)
46 let regexp ligature_char = [ "'`~!?@*()[]<>-+=|:;.,/\"" ]
47 let regexp ligature = ligature_char ligature_char+
49 let is_ligature_char =
50 (* must be in sync with "regexp ligature_char" above *)
51 let chars = "'`~!?@*()[]<>-+=|:;.,/\"" in
54 ignore (String.index chars char);
56 with Not_found -> false))
58 let regexp we_proved = "we" utf8_blank+ "proved"
59 let regexp we_have = "we" utf8_blank+ "have"
60 let regexp let_rec = "let" utf8_blank+ "rec"
61 let regexp let_corec = "let" utf8_blank+ "corec"
62 let regexp ident_decoration = '\'' | '?' | '`'
63 let regexp ident_cont = ident_letter | xml_digit | '_'
64 let regexp ident = ident_letter ident_cont* ident_decoration*
66 let regexp tex_token = '\\' ident
68 let regexp delim_begin = "\\["
69 let regexp delim_end = "\\]"
71 let regexp qkeyword = "'" ident "'"
73 let regexp implicit = '?'
74 let regexp placeholder = '%'
75 let regexp meta = implicit number
77 let regexp csymbol = '\'' ident
79 let regexp begin_group = "@{" | "${"
80 let regexp end_group = '}'
81 let regexp wildcard = "$_"
82 let regexp ast_ident = "@" ident
83 let regexp ast_csymbol = "@" csymbol
84 let regexp meta_ident = "$" ident
85 let regexp meta_anonymous = "$_"
86 let regexp qstring = '"' [^ '"']* '"'
88 let regexp begincomment = "(**" utf8_blank
89 let regexp beginnote = "(*"
90 let regexp endcomment = "*)"
91 (* let regexp comment_char = [^'*'] | '*'[^')']
92 let regexp note = "|+" ([^'*'] | "**") comment_char* "+|" *)
97 "over"; "atop"; "frac";
101 let level1_keywords =
102 [ "hbox"; "hvbox"; "hovbox"; "vbox";
104 "list0"; "list1"; "sep";
106 "term"; "ident"; "number"; "mstyle"
109 let level2_meta_keywords =
110 [ "if"; "then"; "else";
111 "fold"; "left"; "right"; "rec";
114 "anonymous"; "ident"; "number"; "term"; "fresh"
117 (* (string, unit) Hashtbl.t, to exploit multiple bindings *)
118 let level2_ast_keywords = Hashtbl.create 23
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" ]
124 let add_level2_ast_keyword k = Hashtbl.add level2_ast_keywords k ()
125 let remove_level2_ast_keyword k = Hashtbl.remove level2_ast_keywords k
127 (* (string, int) Hashtbl.t, with multiple bindings.
128 * int is the unicode codepoint *)
129 let ligatures = Hashtbl.create 23
132 (fun (ligature, symbol) -> Hashtbl.add ligatures ligature symbol)
133 [ ("->", <:unicode<to>>); ("=>", <:unicode<Rightarrow>>);
134 ("<=", <:unicode<leq>>); (">=", <:unicode<geq>>);
135 ("<>", <:unicode<neq>>); (":=", <:unicode<def>>);
136 ("==", <:unicode<equiv>>);
139 let regexp uri_step = [ 'a' - 'z' 'A' - 'Z' '0' - '9' '_' '-' ''' ]+
142 ("cic:/" | "theory:/") (* schema *)
143 (* ident ('/' ident)* |+ path +| *)
144 uri_step ('/' uri_step)* (* path *)
145 ('.' ident)+ (* ext *)
146 ("#xpointer(" number ('/' number)+ ")")? (* xpointer *)
148 let error lexbuf msg =
149 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
150 raise (Error (begin_cnum, end_cnum, msg))
151 let error_at_end lexbuf msg =
152 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
153 raise (Error (begin_cnum, end_cnum, msg))
155 let return_with_loc token begin_cnum end_cnum =
156 let flocation = HExtlib.floc_of_loc (begin_cnum,end_cnum) in
159 let return lexbuf token =
160 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
161 return_with_loc token begin_cnum end_cnum
163 let return_lexeme lexbuf name = return lexbuf (name, Ulexing.utf8_lexeme lexbuf)
165 let return_symbol lexbuf s = return lexbuf ("SYMBOL", s)
166 let return_eoi lexbuf = return lexbuf ("EOI", "")
168 let remove_quotes s = String.sub s 1 (String.length s - 2)
171 let tok_func stream =
172 (* let lexbuf = Ulexing.from_utf8_stream stream in *)
173 (** XXX Obj.magic rationale.
175 * camlp5 constraints the tok_func field of Token.glexer to have type:
176 * Stream.t char -> (Stream.t 'te * flocation_function)
177 * In order to use ulex we have (in theory) to instantiate a new lexbuf each
178 * time a char Stream.t is passed, destroying the previous lexbuf which may
179 * have consumed a character from the old stream which is lost forever :-(
181 * Instead of passing to camlp5 a char Stream.t we pass a lexbuf, casting it to
182 * char Stream.t with Obj.magic where needed.
184 let lexbuf = Obj.magic stream in
185 Token.make_stream_and_location
190 | Ulexing.Error -> error_at_end lexbuf "Unexpected character"
191 | Ulexing.InvalidCodepoint p ->
192 error_at_end lexbuf (sprintf "Invalid code point: %d" p))
195 Token.tok_func = tok_func;
196 Token.tok_using = (fun _ -> ());
197 Token.tok_removing = (fun _ -> ());
198 Token.tok_match = Token.default_match;
199 Token.tok_text = Token.lexer_text;
200 Token.tok_comm = None;
203 let expand_macro lexbuf =
205 Ulexing.utf8_sub_lexeme lexbuf 1 (Ulexing.lexeme_length lexbuf - 1)
208 ("SYMBOL", Utf8Macro.expand macro)
209 with Utf8Macro.Macro_not_found _ -> "SYMBOL", Ulexing.utf8_lexeme lexbuf
211 let remove_quotes s = String.sub s 1 (String.length s - 2)
212 let remove_left_quote s = String.sub s 1 (String.length s - 1)
214 let rec level2_pattern_token_group counter buffer =
217 if (counter > 0) then
218 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
219 snd (Ulexing.loc lexbuf)
221 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
222 ignore (level2_pattern_token_group (counter + 1) buffer lexbuf) ;
223 level2_pattern_token_group counter buffer lexbuf
225 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
226 level2_pattern_token_group counter buffer lexbuf
228 let read_unparsed_group token_name lexbuf =
229 let buffer = Buffer.create 16 in
230 let begin_cnum, _ = Ulexing.loc lexbuf in
231 let end_cnum = level2_pattern_token_group 0 buffer lexbuf in
232 return_with_loc (token_name, Buffer.contents buffer) begin_cnum end_cnum
234 let rec level2_meta_token =
236 | utf8_blank+ -> level2_meta_token lexbuf
238 let s = Ulexing.utf8_lexeme lexbuf in
240 if List.mem s level2_meta_keywords then
241 return lexbuf ("", s)
243 return lexbuf ("IDENT", s)
245 | "@{" -> read_unparsed_group "UNPARSED_AST" lexbuf
247 return lexbuf ("UNPARSED_AST",
248 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
250 return lexbuf ("UNPARSED_AST",
251 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
252 | eof -> return_eoi lexbuf
254 let rec comment_token acc depth =
257 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
258 comment_token acc (depth + 1) lexbuf
260 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
263 else comment_token acc (depth - 1) lexbuf
265 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
266 comment_token acc depth lexbuf
268 (** @param k continuation to be invoked when no ligature has been found *)
269 let rec ligatures_token k =
272 let lexeme = Ulexing.utf8_lexeme lexbuf in
273 (match List.rev (Hashtbl.find_all ligatures lexeme) with
274 | [] -> (* ligature not found, rollback and try default lexer *)
275 Ulexing.rollback lexbuf;
277 | default_lig :: _ -> (* ligatures found, use the default one *)
278 return_symbol lexbuf default_lig)
279 | eof -> return_eoi lexbuf
280 | _ -> (* not a ligature, rollback and try default lexer *)
281 Ulexing.rollback lexbuf;
284 and level2_ast_token =
286 | let_rec -> return lexbuf ("LETREC","")
287 | let_corec -> return lexbuf ("LETCOREC","")
288 | we_proved -> return lexbuf ("WEPROVED","")
289 | we_have -> return lexbuf ("WEHAVE","")
290 | utf8_blank+ -> ligatures_token level2_ast_token lexbuf
292 let s = Ulexing.utf8_lexeme lexbuf in
293 return lexbuf ("META", String.sub s 1 (String.length s - 1))
294 | implicit -> return lexbuf ("IMPLICIT", "")
295 | placeholder -> return lexbuf ("PLACEHOLDER", "")
297 let lexeme = Ulexing.utf8_lexeme lexbuf in
298 if Hashtbl.mem level2_ast_keywords lexeme then
299 return lexbuf ("", lexeme)
301 return lexbuf ("IDENT", lexeme)
302 | number -> return lexbuf ("NUMBER", Ulexing.utf8_lexeme lexbuf)
303 | tex_token -> return lexbuf (expand_macro lexbuf)
304 | uri -> return lexbuf ("URI", Ulexing.utf8_lexeme lexbuf)
306 return lexbuf ("QSTRING", remove_quotes (Ulexing.utf8_lexeme lexbuf))
308 return lexbuf ("CSYMBOL", remove_left_quote (Ulexing.utf8_lexeme lexbuf))
309 | "${" -> read_unparsed_group "UNPARSED_META" lexbuf
310 | "@{" -> read_unparsed_group "UNPARSED_AST" lexbuf
311 | '(' -> return lexbuf ("LPAREN", "")
312 | ')' -> return lexbuf ("RPAREN", "")
314 return lexbuf ("UNPARSED_META",
315 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
316 | meta_anonymous -> return lexbuf ("UNPARSED_META", "anonymous")
318 let _comment = comment_token (Ulexing.utf8_lexeme lexbuf) 0 lexbuf in
320 Ulexing.utf8_sub_lexeme lexbuf 2 (Ulexing.lexeme_length lexbuf - 4)
322 return lexbuf ("NOTE", comment) *)
323 ligatures_token level2_ast_token lexbuf
324 | begincomment -> return lexbuf ("BEGINCOMMENT","")
325 | endcomment -> return lexbuf ("ENDCOMMENT","")
326 | eof -> return_eoi lexbuf
327 | _ -> return_symbol lexbuf (Ulexing.utf8_lexeme lexbuf)
329 and level1_pattern_token =
331 | utf8_blank+ -> ligatures_token level1_pattern_token lexbuf
332 | number -> return lexbuf ("NUMBER", Ulexing.utf8_lexeme lexbuf)
334 let s = Ulexing.utf8_lexeme lexbuf in
336 if List.mem s level1_keywords then
337 return lexbuf ("", s)
339 return lexbuf ("IDENT", s)
341 | color -> return lexbuf ("COLOR", Ulexing.utf8_lexeme lexbuf)
343 return lexbuf ("FLOATWITHUNIT", Ulexing.utf8_lexeme lexbuf)
344 | tex_token -> return lexbuf (expand_macro lexbuf)
346 return lexbuf ("QKEYWORD", remove_quotes (Ulexing.utf8_lexeme lexbuf))
347 | '(' -> return lexbuf ("LPAREN", "")
348 | ')' -> return lexbuf ("RPAREN", "")
349 | eof -> return_eoi lexbuf
350 | _ -> return_symbol lexbuf (Ulexing.utf8_lexeme lexbuf)
352 let level1_pattern_token = ligatures_token level1_pattern_token
353 let level2_ast_token = ligatures_token level2_ast_token
355 (* API implementation *)
357 let level1_pattern_lexer = mk_lexer level1_pattern_token
358 let level2_ast_lexer = mk_lexer level2_ast_token
359 let level2_meta_lexer = mk_lexer level2_meta_token
361 let lookup_ligatures lexeme =
364 then [ Utf8Macro.expand (String.sub lexeme 1 (String.length lexeme - 1)) ]
365 else List.rev (Hashtbl.find_all ligatures lexeme)
366 with Invalid_argument _ | Utf8Macro.Macro_not_found _ -> []