10 prerr_endline (sprintf "Lexer error : %s" callerID);
12 raise Parsing.Parse_error
14 let pos lexbuf = (lexeme_start lexbuf, lexeme_end lexbuf)
16 let advance_line_pos pos =
17 { pos with pos_lnum = pos.pos_lnum + 1; pos_bol = pos.pos_cnum; }
19 let advance_line lexbuf =
20 lexbuf.lex_curr_p <- advance_line_pos lexbuf.lex_curr_p
39 "precision",PRECISION;
47 "character",CHARACTER;
78 "intersect",INTERSECT;
80 "temporary",TEMPORARY;
87 "references",REFERENCES;
91 "timestamp",TIMESTAMP;
97 "constraint",CONSTRAINT;
124 "interval", INTERVAL;
125 "microsecond", MICROSECOND;
134 "second_microsecond", SECOND_MICROSECOND;
135 "minute_microsecond", MINUTE_MICROSECOND;
136 "minute_second", MINUTE_SECOND;
137 "hour_microsecond", HOUR_MICROSECOND;
138 "hour_second", HOUR_SECOND;
139 "hour_minute", HOUR_MINUTE;
140 "day_microsecond", DAY_MICROSECOND;
141 "day_second", DAY_SECOND;
142 "day_minute", DAY_MINUTE;
143 "day_hour", DAY_HOUR;
144 "year_month", YEAR_MONTH;
147 "duplicate", DUPLICATE;
148 "function", FUNCTION;
149 "procedure", PROCEDURE;
153 "language", LANGUAGE;
154 "substring", SUBSTRING;
157 let all token l = k := !k @ List.map (fun x -> x,token) l in
158 all DATETIME_FUNC ["current_date";"current_timestamp";"current_time";"localtime";"localtimestamp";"now";];
159 all DATETIME_FUNC ["getdate"]; (* mssql? *)
160 all CONFLICT_ALGO ["ignore"; "abort"; "fail"; "rollback"];
161 all JOIN_TYPE1 ["left";"right";"full"];
162 all JOIN_TYPE2 ["inner";"outer"];
163 all LIKE_OP ["glob";"regexp";"match"];
164 all AUTOINCREMENT ["autoincrement";"auto_increment"];
165 (* standard built-in types
166 CHARACTER, CHARACTER VARYING, CHARACTER LARGE OBJECT,
167 BINARY, BINARY VARYING, BINARY LARGE OBJECT,
168 NUMERIC, DECIMAL, INTEGER, SMALLINT, BIGINT,
169 FLOAT, REAL, DOUBLE PRECISION,
171 DATE, TIME, TIMESTAMP, INTERVAL
173 all T_INTEGER ["integer";"int";"smallint";"bigint";"tinyint";"mediumint";"middleint";"serial";"identity"];
174 all T_INTEGER ["numeric";"decimal";"dec";"fixed"];
175 all T_INTEGER ["number"]; (* oracle *)
176 all T_BOOLEAN ["bool";"boolean"];
177 all T_FLOAT ["float";"real";"double";"float4";"float8";"int1";"int2";"int3";"int4";"int8"];
178 all T_BLOB ["blob";"varbinary";"tinyblob";"mediumblob";"longblob"];
179 all T_TEXT ["text";"char";"varchar";"tinytext";"mediumtext";"longtext"];
180 all T_TEXT ["varchar2"]; (* oracle *)
181 all T_DATETIME ["datetime"];
182 all T_UUID ["uuid"]; (* http://www.postgresql.org/docs/9.4/static/datatype-uuid.html *)
186 Q: Why not convert all input to lowercase before lexing?
187 A: Sometimes SQL is case-sensitive, also string contents should be preserved
190 module Keywords = Map.Make(String)
194 let k = String.lowercase k in
195 if Keywords.mem k map then
196 failwith (sprintf "Lexeme %s is already associated with keyword." k)
200 List.fold_left add Keywords.empty keywords
202 (* FIXME case sensitivity??! *)
205 let str = String.lowercase str in
206 try Keywords.find str keywords with Not_found -> IDENT str
208 let ident str = IDENT (String.lowercase str)
210 let as_literal ch s =
211 let s = String.replace_chars (fun x -> String.make (if x = ch then 2 else 1) x) s in
212 sprintf "%c%s%c" ch s ch
215 let digit = ['0'-'9']
216 let alpha = ['a'-'z' 'A'-'Z']
217 let ident = (alpha) (alpha | digit | '_' )*
218 let wsp = [' ' '\r' '\t']
219 let cmnt = "--" | "//" | "#"
221 (* extract separate statements *)
222 rule ruleStatement = parse
223 | ['\n' ' ' '\r' '\t']+ as tok { `Space tok }
224 | cmnt wsp* "[sqlgg]" wsp+ (ident+ as n) wsp* "=" wsp* ([^'\n']* as v) '\n' { `Prop (n,v) }
225 | cmnt wsp* "@" (ident+ as name) [^'\n']* '\n' { `Prop ("name",name) }
226 | '"' { let s = ruleInQuotes "" lexbuf in `Token (as_literal '"' s) }
227 | "'" { let s = ruleInSingleQuotes "" lexbuf in `Token (as_literal '\'' s) }
228 | "$" (ident? as tag) "$" { let s = ruleInDollarQuotes tag "" lexbuf in `Token (sprintf "$%s$%s$%s$" tag s tag) }
229 | cmnt as s { `Comment (s ^ ruleComment "" lexbuf) }
230 | "/*" { `Comment ("/*" ^ ruleCommentMulti "" lexbuf ^ "*/") }
232 | [^ ';'] as c { `Char c }
235 (* extract tail of the input *)
238 | _* as str { ruleTail (acc ^ str) lexbuf }
241 | wsp { ruleMain lexbuf }
242 (* update line number *)
243 | '\n' { advance_line lexbuf; ruleMain lexbuf}
250 | cmnt { ignore (ruleComment "" lexbuf); ruleMain lexbuf }
251 | "/*" { ignore (ruleCommentMulti "" lexbuf); ruleMain lexbuf }
261 | "/" | "%" { NUM_DIV_OP }
262 | "<<" | ">>" | "|" | "&" { NUM_BIT_OP }
263 | ">" | ">=" | "<=" | "<" { NUM_CMP_OP }
264 | "<>" | "!=" | "==" { NUM_EQ_OP }
265 | "<=>" { NOT_DISTINCT_OP }
267 | "?" { PARAM (None,pos lexbuf) }
268 | [':' '@'] (ident as str) { PARAM (Some str,pos lexbuf) }
270 | '"' { ident (ruleInQuotes "" lexbuf) }
271 | "'" { TEXT (ruleInSingleQuotes "" lexbuf) }
272 (* http://www.postgresql.org/docs/current/interactive/sql-syntax-lexical.html#SQL-SYNTAX-DOLLAR-QUOTING *)
273 | "$" (ident? as tag) "$" { TEXT (ruleInDollarQuotes tag "" lexbuf) }
274 | "`" { ident (ruleInBackQuotes "" lexbuf) }
275 | "[" { ident (ruleInBrackets "" lexbuf) }
276 | ['x' 'X'] "'" { BLOB (ruleInSingleQuotes "" lexbuf) }
278 | ident as str { get_ident str }
279 | digit+ as str { INTEGER (int_of_string str) }
280 | digit+ '.' digit+ as str { FLOAT (float_of_string str) }
282 | _ { error lexbuf "ruleMain" }
284 (* FIXME factor out all that ruleIn* rules *)
285 ruleInQuotes acc = parse
287 | eof { error lexbuf "no terminating quote" }
288 | '\n' { advance_line lexbuf; error lexbuf "EOL before terminating quote" }
289 | "\"\"" { ruleInQuotes (acc^"\"") lexbuf }
290 | [^'"' '\n']+ as s { ruleInQuotes (acc^s) lexbuf }
291 | _ { error lexbuf "ruleInQuotes" }
293 ruleInBrackets acc = parse
295 | eof { error lexbuf "no terminating bracket" }
296 | '\n' { advance_line lexbuf; error lexbuf "EOL before terminating bracket" }
297 (* | "\"\"" { ruleInQuotes (acc ^ "\"") lexbuf } *)
298 | [^']' '\n']+ { ruleInBrackets (acc ^ lexeme lexbuf) lexbuf }
299 | _ { error lexbuf "ruleInBrackets" }
301 ruleInSingleQuotes acc = parse
303 | eof { error lexbuf "no terminating single quote" }
304 | '\n' { advance_line lexbuf; error lexbuf "EOL before terminating single quote" }
305 | "''" { ruleInSingleQuotes (acc ^ "'") lexbuf }
306 | [^'\'' '\n']+ { ruleInSingleQuotes (acc ^ lexeme lexbuf) lexbuf }
307 | _ { error lexbuf "ruleInSingleQuotes" }
309 ruleInBackQuotes acc = parse
311 | eof { error lexbuf "no terminating back quote" }
312 | '\n' { advance_line lexbuf; error lexbuf "EOL before terminating back quote" }
313 | "``" { ruleInBackQuotes (acc ^ "`") lexbuf }
314 | [^'`' '\n']+ { ruleInBackQuotes (acc ^ lexeme lexbuf) lexbuf }
315 | _ { error lexbuf "ruleInBackQuotes" }
317 ruleInDollarQuotes tag acc = parse
318 | "$" (ident? as tag_) "$" { if tag_ = tag then acc else ruleInDollarQuotes tag (acc ^ sprintf "$%s$" tag_) lexbuf }
319 | eof { error lexbuf "no terminating dollar quote" }
320 | '\n' { advance_line lexbuf; ruleInDollarQuotes tag (acc ^ "\n") lexbuf }
321 (* match one char at a time to make sure delimiter matches longer *)
322 | [^'\n'] { ruleInDollarQuotes tag (acc ^ lexeme lexbuf) lexbuf }
323 | _ { error lexbuf "ruleInDollarQuotes" }
325 ruleComment acc = parse
326 | '\n' { advance_line lexbuf; acc }
328 | [^'\n']+ { let s = lexeme lexbuf in ruleComment (acc ^ s) lexbuf; }
329 | _ { error lexbuf "ruleComment"; }
331 ruleCommentMulti acc = parse
332 | '\n' { advance_line lexbuf; ruleCommentMulti (acc ^ "\n") lexbuf }
335 | [^'\n' '*']+ { let s = lexeme lexbuf in ruleCommentMulti (acc ^ s) lexbuf }
336 | _ { error lexbuf "ruleCommentMulti" }
340 let parse_rule lexbuf =
341 let module P = Parser_state in
342 let token = ruleMain lexbuf in
346 (* eprintf "ignored: %s\n" (lexeme lexbuf); *)
347 if (token = EOF) then token else IGNORED