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/
28 exception Error of int * int * string
30 let regexp number = xml_digit+
32 (* ZACK: breaks unicode's binder followed by an ascii letter without blank *)
33 (* let regexp ident_letter = xml_letter *)
35 let regexp ident_letter = [ 'a' - 'z' 'A' - 'Z' ]
37 (* must be in sync with "is_ligature_char" below *)
38 let regexp ligature_char = [ "'`~!?@*()[]<>-+=|:;.,/\"" ]
39 let regexp ligature = ligature_char ligature_char+
41 let is_ligature_char =
42 (* must be in sync with "regexp ligature_char" above *)
43 let chars = "'`~!?@*()[]<>-+=|:;.,/\"" in
46 ignore (String.index chars char);
48 with Not_found -> false))
50 let regexp ident_decoration = '\'' | '?' | '`'
51 let regexp ident_cont = ident_letter | xml_digit | '_'
52 let regexp ident = ident_letter ident_cont* ident_decoration*
54 let regexp tex_token = '\\' ident
56 let regexp delim_begin = "\\["
57 let regexp delim_end = "\\]"
59 let regexp qkeyword = "'" ident "'"
61 let regexp implicit = '?'
62 let regexp placeholder = '%'
63 let regexp meta = implicit number
65 let regexp csymbol = '\'' ident
67 let regexp begin_group = "@{" | "${"
68 let regexp end_group = '}'
69 let regexp wildcard = "$_"
70 let regexp ast_ident = "@" ident
71 let regexp ast_csymbol = "@" csymbol
72 let regexp meta_ident = "$" ident
73 let regexp meta_anonymous = "$_"
74 let regexp qstring = '"' [^ '"']* '"'
76 let regexp begincomment = "(**" xml_blank
77 let regexp beginnote = "(*"
78 let regexp endcomment = "*)"
79 (* let regexp comment_char = [^'*'] | '*'[^')']
80 let regexp note = "|+" ([^'*'] | "**") comment_char* "+|" *)
85 "over"; "atop"; "frac";
90 [ "hbox"; "hvbox"; "hovbox"; "vbox";
92 "list0"; "list1"; "sep";
94 "term"; "ident"; "number"
97 let level2_meta_keywords =
98 [ "if"; "then"; "else";
99 "fold"; "left"; "right"; "rec";
102 "anonymous"; "ident"; "number"; "term"; "fresh"
105 (* (string, unit) Hashtbl.t, to exploit multiple bindings *)
106 let level2_ast_keywords = Hashtbl.create 23
108 List.iter (fun k -> Hashtbl.add level2_ast_keywords k ())
109 [ "CProp"; "Prop"; "Type"; "Set"; "let"; "rec"; "corec"; "match";
110 "with"; "in"; "and"; "to"; "as"; "on"; "return" ]
112 let add_level2_ast_keyword k = Hashtbl.add level2_ast_keywords k ()
113 let remove_level2_ast_keyword k = Hashtbl.remove level2_ast_keywords k
115 (* (string, int) Hashtbl.t, with multiple bindings.
116 * int is the unicode codepoint *)
117 let ligatures = Hashtbl.create 23
120 (fun (ligature, symbol) -> Hashtbl.add ligatures ligature symbol)
121 [ ("->", <:unicode<to>>); ("=>", <:unicode<Rightarrow>>);
122 ("<=", <:unicode<leq>>); (">=", <:unicode<geq>>);
123 ("<>", <:unicode<neq>>); (":=", <:unicode<def>>);
126 let regexp uri_step = [ 'a' - 'z' 'A' - 'Z' '0' - '9' '_' '-' ]+
129 ("cic:/" | "theory:/") (* schema *)
130 (* ident ('/' ident)* |+ path +| *)
131 uri_step ('/' uri_step)* (* path *)
132 ('.' ident)+ (* ext *)
133 ("#xpointer(" number ('/' number)+ ")")? (* xpointer *)
135 let error lexbuf msg =
136 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
137 raise (Error (begin_cnum, end_cnum, msg))
138 let error_at_end lexbuf msg =
139 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
140 raise (Error (begin_cnum, end_cnum, msg))
142 let return_with_loc token begin_cnum end_cnum =
143 (* TODO handle line/column numbers *)
144 let flocation_begin =
145 { Lexing.pos_fname = "";
146 Lexing.pos_lnum = -1; Lexing.pos_bol = -1;
147 Lexing.pos_cnum = begin_cnum }
149 let flocation_end = { flocation_begin with Lexing.pos_cnum = end_cnum } in
150 (token, (flocation_begin, flocation_end))
152 let return lexbuf token =
153 let begin_cnum, end_cnum = Ulexing.loc lexbuf in
154 return_with_loc token begin_cnum end_cnum
156 let return_lexeme lexbuf name = return lexbuf (name, Ulexing.utf8_lexeme lexbuf)
158 let return_symbol lexbuf s = return lexbuf ("SYMBOL", s)
159 let return_eoi lexbuf = return lexbuf ("EOI", "")
161 let remove_quotes s = String.sub s 1 (String.length s - 2)
164 let tok_func stream =
165 (* let lexbuf = Ulexing.from_utf8_stream stream in *)
166 (** XXX Obj.magic rationale.
168 * camlp4 constraints the tok_func field of Token.glexer to have type:
169 * Stream.t char -> (Stream.t 'te * flocation_function)
170 * In order to use ulex we have (in theory) to instantiate a new lexbuf each
171 * time a char Stream.t is passed, destroying the previous lexbuf which may
172 * have consumed a character from the old stream which is lost forever :-(
174 * Instead of passing to camlp4 a char Stream.t we pass a lexbuf, casting it to
175 * char Stream.t with Obj.magic where needed.
177 let lexbuf = Obj.magic stream in
178 Token.make_stream_and_flocation
183 | Ulexing.Error -> error_at_end lexbuf "Unexpected character"
184 | Ulexing.InvalidCodepoint p ->
185 error_at_end lexbuf (sprintf "Invalid code point: %d" p))
188 Token.tok_func = tok_func;
189 Token.tok_using = (fun _ -> ());
190 Token.tok_removing = (fun _ -> ());
191 Token.tok_match = Token.default_match;
192 Token.tok_text = Token.lexer_text;
193 Token.tok_comm = None;
196 let expand_macro lexbuf =
198 Ulexing.utf8_sub_lexeme lexbuf 1 (Ulexing.lexeme_length lexbuf - 1)
201 ("SYMBOL", Utf8Macro.expand macro)
202 with Utf8Macro.Macro_not_found _ -> "SYMBOL", Ulexing.utf8_lexeme lexbuf
204 let remove_quotes s = String.sub s 1 (String.length s - 2)
205 let remove_left_quote s = String.sub s 1 (String.length s - 1)
207 let rec level2_pattern_token_group counter buffer =
210 if (counter > 0) then
211 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
212 snd (Ulexing.loc lexbuf)
214 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
215 ignore (level2_pattern_token_group (counter + 1) buffer lexbuf) ;
216 level2_pattern_token_group counter buffer lexbuf
218 Buffer.add_string buffer (Ulexing.utf8_lexeme lexbuf) ;
219 level2_pattern_token_group counter buffer lexbuf
221 let read_unparsed_group token_name lexbuf =
222 let buffer = Buffer.create 16 in
223 let begin_cnum, _ = Ulexing.loc lexbuf in
224 let end_cnum = level2_pattern_token_group 0 buffer lexbuf in
225 return_with_loc (token_name, Buffer.contents buffer) begin_cnum end_cnum
227 let rec level2_meta_token =
229 | xml_blank+ -> level2_meta_token lexbuf
231 let s = Ulexing.utf8_lexeme lexbuf in
233 if List.mem s level2_meta_keywords then
234 return lexbuf ("", s)
236 return lexbuf ("IDENT", s)
238 | "@{" -> read_unparsed_group "UNPARSED_AST" lexbuf
240 return lexbuf ("UNPARSED_AST",
241 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
243 return lexbuf ("UNPARSED_AST",
244 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
245 | eof -> return_eoi lexbuf
247 let rec comment_token acc depth =
250 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
251 comment_token acc (depth + 1) lexbuf
253 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
256 else comment_token acc (depth - 1) lexbuf
258 let acc = acc ^ Ulexing.utf8_lexeme lexbuf in
259 comment_token acc depth lexbuf
261 (** @param k continuation to be invoked when no ligature has been found *)
262 let rec ligatures_token k =
265 let lexeme = Ulexing.utf8_lexeme lexbuf in
266 (match List.rev (Hashtbl.find_all ligatures lexeme) with
267 | [] -> (* ligature not found, rollback and try default lexer *)
268 Ulexing.rollback lexbuf;
270 | default_lig :: _ -> (* ligatures found, use the default one *)
271 return_symbol lexbuf default_lig)
272 | eof -> return_eoi lexbuf
273 | _ -> (* not a ligature, rollback and try default lexer *)
274 Ulexing.rollback lexbuf;
277 and level2_ast_token =
279 | xml_blank+ -> ligatures_token level2_ast_token lexbuf
280 | meta -> return lexbuf ("META", Ulexing.utf8_lexeme lexbuf)
281 | implicit -> return lexbuf ("IMPLICIT", "")
282 | placeholder -> return lexbuf ("PLACEHOLDER", "")
284 let lexeme = Ulexing.utf8_lexeme lexbuf in
285 if Hashtbl.mem level2_ast_keywords lexeme then
286 return lexbuf ("", lexeme)
288 return lexbuf ("IDENT", lexeme)
289 | number -> return lexbuf ("NUMBER", Ulexing.utf8_lexeme lexbuf)
290 | tex_token -> return lexbuf (expand_macro lexbuf)
291 | uri -> return lexbuf ("URI", Ulexing.utf8_lexeme lexbuf)
293 return lexbuf ("QSTRING", remove_quotes (Ulexing.utf8_lexeme lexbuf))
295 return lexbuf ("CSYMBOL", remove_left_quote (Ulexing.utf8_lexeme lexbuf))
296 | "${" -> read_unparsed_group "UNPARSED_META" lexbuf
297 | "@{" -> read_unparsed_group "UNPARSED_AST" lexbuf
298 | '(' -> return lexbuf ("LPAREN", "")
299 | ')' -> return lexbuf ("RPAREN", "")
301 return lexbuf ("UNPARSED_META",
302 remove_left_quote (Ulexing.utf8_lexeme lexbuf))
303 | meta_anonymous -> return lexbuf ("UNPARSED_META", "anonymous")
305 let _comment = comment_token (Ulexing.utf8_lexeme lexbuf) 0 lexbuf in
307 Ulexing.utf8_sub_lexeme lexbuf 2 (Ulexing.lexeme_length lexbuf - 4)
309 return lexbuf ("NOTE", comment) *)
310 ligatures_token level2_ast_token lexbuf
311 | begincomment -> return lexbuf ("BEGINCOMMENT","")
312 | endcomment -> return lexbuf ("ENDCOMMENT","")
313 | eof -> return_eoi lexbuf
314 | _ -> return_symbol lexbuf (Ulexing.utf8_lexeme lexbuf)
316 and level1_pattern_token =
318 | xml_blank+ -> ligatures_token level1_pattern_token lexbuf
319 | number -> return lexbuf ("NUMBER", Ulexing.utf8_lexeme lexbuf)
321 let s = Ulexing.utf8_lexeme lexbuf in
323 if List.mem s level1_keywords then
324 return lexbuf ("", s)
326 return lexbuf ("IDENT", s)
328 | tex_token -> return lexbuf (expand_macro lexbuf)
330 return lexbuf ("QKEYWORD", remove_quotes (Ulexing.utf8_lexeme lexbuf))
331 | '(' -> return lexbuf ("LPAREN", "")
332 | ')' -> return lexbuf ("RPAREN", "")
333 | eof -> return_eoi lexbuf
334 | _ -> return_symbol lexbuf (Ulexing.utf8_lexeme lexbuf)
336 let level1_pattern_token = ligatures_token level1_pattern_token
337 let level2_ast_token = ligatures_token level2_ast_token
339 (* API implementation *)
341 let level1_pattern_lexer = mk_lexer level1_pattern_token
342 let level2_ast_lexer = mk_lexer level2_ast_token
343 let level2_meta_lexer = mk_lexer level2_meta_token
345 let lookup_ligatures lexeme =
348 then [ Utf8Macro.expand (String.sub lexeme 1 (String.length lexeme - 1)) ]
349 else List.rev (Hashtbl.find_all ligatures lexeme)
350 with Invalid_argument _ | Utf8Macro.Macro_not_found _ -> []