Line data Source code
1 : /*
2 : * SPDX-License-Identifier: MPL-2.0
3 : *
4 : * This Source Code Form is subject to the terms of the Mozilla Public
5 : * License, v. 2.0. If a copy of the MPL was not distributed with this
6 : * file, You can obtain one at http://mozilla.org/MPL/2.0/.
7 : *
8 : * Copyright 2024, 2025 MonetDB Foundation;
9 : * Copyright August 2008 - 2023 MonetDB B.V.;
10 : * Copyright 1997 - July 2008 CWI.
11 : */
12 :
13 : #include "monetdb_config.h"
14 : #include <wctype.h>
15 : #include "sql_mem.h"
16 : #include "sql_scan.h"
17 : #include "sql_types.h"
18 : #include "sql_symbol.h"
19 : #include "sql_mvc.h"
20 : #include "sql_parser.tab.h"
21 : #include "sql_semantic.h"
22 : #include "sql_parser.h" /* for sql_error() */
23 :
24 : #include "stream.h"
25 : #include "mapi_prompt.h"
26 : #include <unistd.h>
27 : #include <string.h>
28 : #include <ctype.h>
29 : #include "sql_keyword.h"
30 :
31 : static char *
32 17 : uescape_xform(char *restrict s, const char *restrict esc)
33 : {
34 17 : size_t i, j;
35 :
36 60 : for (i = j = 0; s[i]; i++) {
37 43 : if (s[i] == *esc) {
38 43 : if (s[i + 1] == *esc) {
39 0 : s[j++] = *esc;
40 0 : i++;
41 : } else {
42 43 : int c = 0;
43 43 : int n;
44 43 : if (s[i + 1] == '+') {
45 26 : n = 6;
46 26 : i++;
47 : } else {
48 : n = 4;
49 : }
50 224 : do {
51 224 : i++;
52 224 : c <<= 4;
53 224 : if ('0' <= s[i] && s[i] <= '9')
54 170 : c |= s[i] - '0';
55 54 : else if ('a' <= s[i] && s[i] <= 'f')
56 24 : c |= s[i] - 'a' + 10;
57 30 : else if ('A' <= s[i] && s[i] <= 'F')
58 30 : c |= s[i] - 'A' + 10;
59 : else
60 : return NULL;
61 224 : } while (--n > 0);
62 43 : if (c == 0 || c > 0x10FFFF || (c & 0xFFF800) == 0xD800)
63 : return NULL;
64 43 : if (c < 0x80) {
65 14 : s[j++] = c;
66 : } else {
67 29 : if (c < 0x800) {
68 5 : s[j++] = 0xC0 | (c >> 6);
69 : } else {
70 24 : if (c < 0x10000) {
71 12 : s[j++] = 0xE0 | (c >> 12);
72 : } else {
73 12 : s[j++] = 0xF0 | (c >> 18);
74 12 : s[j++] = 0x80 | ((c >> 12) & 0x3F);
75 : }
76 24 : s[j++] = 0x80 | ((c >> 6) & 0x3F);
77 : }
78 29 : s[j++] = 0x80 | (c & 0x3F);
79 : }
80 : }
81 : } else {
82 0 : s[j++] = s[i];
83 : }
84 : }
85 17 : s[j] = 0;
86 17 : return s;
87 : }
88 :
89 : /**
90 : * Removes all comments before the query. In query comments are kept.
91 : */
92 : char *
93 438941 : query_cleaned(allocator *sa, const char *query)
94 : {
95 438941 : char *q, *r, *c = NULL;
96 438941 : int lines = 0;
97 438941 : int quote = 0; /* inside quotes ('..', "..", {..}) */
98 438941 : bool bs = false; /* seen a backslash in a quoted string */
99 438941 : bool incomment1 = false; /* inside traditional C style comment */
100 438941 : bool incomment2 = false; /* inside comment starting with -- */
101 438941 : bool inline_comment = false;
102 :
103 438941 : r = SA_NEW_ARRAY(sa, char, strlen(query) + 1);
104 438938 : if(!r)
105 : return NULL;
106 :
107 69985888 : for (q = r; *query; query++) {
108 69546950 : if (incomment1) {
109 16396 : if (*query == '/' && query[-1] == '*') {
110 237 : incomment1 = false;
111 237 : if (c == r && lines > 0) {
112 229 : q = r; // reset to beginning
113 229 : lines = 0;
114 229 : continue;
115 : }
116 : }
117 16167 : if (*query == '\n') lines++;
118 16167 : *q++ = *query;
119 69530554 : } else if (incomment2) {
120 831472 : if (*query == '\n') {
121 2916 : incomment2 = false;
122 2916 : inline_comment = false;
123 : /* add newline only if comment doesn't
124 : * occupy whole line */
125 2916 : if (q > r && q[-1] != '\n'){
126 1000 : *q++ = '\n';
127 1000 : lines++;
128 : }
129 828556 : } else if (inline_comment){
130 24007 : *q++ = *query; // preserve in line query comments
131 : }
132 68699082 : } else if (quote) {
133 22397764 : if (bs) {
134 : bs = false;
135 22394431 : } else if (*query == '\\') {
136 : bs = true;
137 22391098 : } else if (*query == quote) {
138 690268 : quote = 0;
139 : }
140 22397764 : *q++ = *query;
141 46301318 : } else if (*query == '"' || *query == '\'') {
142 689800 : quote = *query;
143 689800 : *q++ = *query;
144 45611518 : } else if (*query == '{') {
145 513 : quote = '}';
146 513 : *q++ = *query;
147 45611005 : } else if (*query == '-' && query[1] == '-') {
148 2916 : if (q > r && q[-1] != '\n') {
149 1000 : inline_comment = true;
150 1000 : *q++ = *query; // preserve in line query comments
151 : }
152 : incomment2 = true;
153 45608089 : } else if (*query == '/' && query[1] == '*') {
154 237 : incomment1 = true;
155 237 : c = q;
156 237 : *q++ = *query;
157 45607852 : } else if (*query == '\n') {
158 : /* collapse newlines */
159 907305 : if (q > r && q[-1] != '\n') {
160 865324 : *q++ = '\n';
161 865324 : lines++;
162 : }
163 44700547 : } else if (*query == ' ' || *query == '\t') {
164 : /* collapse white space */
165 7192357 : if (q > r && q[-1] != ' ')
166 5705268 : *q++ = ' ';
167 : } else {
168 37508190 : *q++ = *query;
169 : }
170 : }
171 438938 : *q = 0;
172 438938 : return r;
173 : }
174 :
175 : int
176 357 : scanner_init_keywords(void)
177 : {
178 357 : int failed = 0;
179 :
180 357 : failed += keywords_insert("false", BOOL_FALSE);
181 357 : failed += keywords_insert("true", BOOL_TRUE);
182 357 : failed += keywords_insert("bool", sqlBOOL);
183 :
184 357 : failed += keywords_insert("ALTER", ALTER);
185 357 : failed += keywords_insert("ADD", ADD);
186 357 : failed += keywords_insert("AND", AND);
187 :
188 357 : failed += keywords_insert("RANK", RANK);
189 357 : failed += keywords_insert("DENSE_RANK", RANK);
190 357 : failed += keywords_insert("PERCENT_RANK", RANK);
191 357 : failed += keywords_insert("CUME_DIST", RANK);
192 357 : failed += keywords_insert("ROW_NUMBER", RANK);
193 357 : failed += keywords_insert("NTILE", RANK);
194 357 : failed += keywords_insert("LAG", RANK);
195 357 : failed += keywords_insert("LEAD", RANK);
196 357 : failed += keywords_insert("FETCH", FETCH);
197 357 : failed += keywords_insert("FIRST_VALUE", RANK);
198 357 : failed += keywords_insert("LAST_VALUE", RANK);
199 357 : failed += keywords_insert("NTH_VALUE", RANK);
200 :
201 357 : failed += keywords_insert("BEST", BEST);
202 357 : failed += keywords_insert("EFFORT", EFFORT);
203 :
204 357 : failed += keywords_insert("AS", AS);
205 357 : failed += keywords_insert("ASC", ASC);
206 357 : failed += keywords_insert("AUTHORIZATION", AUTHORIZATION);
207 357 : failed += keywords_insert("BETWEEN", BETWEEN);
208 357 : failed += keywords_insert("SYMMETRIC", SYMMETRIC);
209 357 : failed += keywords_insert("ASYMMETRIC", ASYMMETRIC);
210 357 : failed += keywords_insert("BY", BY);
211 357 : failed += keywords_insert("CAST", CAST);
212 357 : failed += keywords_insert("CONVERT", CONVERT);
213 357 : failed += keywords_insert("CHARACTER", CHARACTER);
214 357 : failed += keywords_insert("CHAR", CHARACTER);
215 357 : failed += keywords_insert("VARYING", VARYING);
216 357 : failed += keywords_insert("VARCHAR", VARCHAR);
217 357 : failed += keywords_insert("BINARY", BINARY);
218 357 : failed += keywords_insert("LARGE", LARGE);
219 357 : failed += keywords_insert("OBJECT", OBJECT);
220 357 : failed += keywords_insert("CLOB", CLOB);
221 357 : failed += keywords_insert("BLOB", sqlBLOB);
222 357 : failed += keywords_insert("TEXT", sqlTEXT);
223 357 : failed += keywords_insert("TINYTEXT", sqlTEXT);
224 357 : failed += keywords_insert("STRING", CLOB); /* ? */
225 357 : failed += keywords_insert("CHECK", CHECK);
226 357 : failed += keywords_insert("CLIENT", CLIENT);
227 357 : failed += keywords_insert("SERVER", SERVER);
228 357 : failed += keywords_insert("COMMENT", COMMENT);
229 357 : failed += keywords_insert("CONSTRAINT", CONSTRAINT);
230 357 : failed += keywords_insert("CREATE", CREATE);
231 357 : failed += keywords_insert("CROSS", CROSS);
232 357 : failed += keywords_insert("COPY", COPY);
233 357 : failed += keywords_insert("RECORDS", RECORDS);
234 357 : failed += keywords_insert("DELIMITERS", DELIMITERS);
235 357 : failed += keywords_insert("STDIN", STDIN);
236 357 : failed += keywords_insert("STDOUT", STDOUT);
237 :
238 357 : failed += keywords_insert("TINYINT", TINYINT);
239 357 : failed += keywords_insert("SMALLINT", SMALLINT);
240 357 : failed += keywords_insert("INTEGER", sqlINTEGER);
241 357 : failed += keywords_insert("INT", sqlINTEGER);
242 357 : failed += keywords_insert("MEDIUMINT", sqlINTEGER);
243 357 : failed += keywords_insert("BIGINT", BIGINT);
244 : #ifdef HAVE_HGE
245 357 : failed += keywords_insert("HUGEINT", HUGEINT);
246 : #endif
247 357 : failed += keywords_insert("DEC", sqlDECIMAL);
248 357 : failed += keywords_insert("DECIMAL", sqlDECIMAL);
249 357 : failed += keywords_insert("NUMERIC", sqlDECIMAL);
250 357 : failed += keywords_insert("DECLARE", DECLARE);
251 357 : failed += keywords_insert("DEFAULT", DEFAULT);
252 357 : failed += keywords_insert("DESC", DESC);
253 357 : failed += keywords_insert("DISTINCT", DISTINCT);
254 357 : failed += keywords_insert("DOUBLE", sqlDOUBLE);
255 357 : failed += keywords_insert("REAL", sqlREAL);
256 357 : failed += keywords_insert("DROP", DROP);
257 357 : failed += keywords_insert("ESCAPE", ESCAPE);
258 357 : failed += keywords_insert("EXISTS", EXISTS);
259 357 : failed += keywords_insert("UESCAPE", UESCAPE);
260 357 : failed += keywords_insert("EXTRACT", EXTRACT);
261 357 : failed += keywords_insert("FLOAT", sqlFLOAT);
262 357 : failed += keywords_insert("FOR", FOR);
263 357 : failed += keywords_insert("FOREIGN", FOREIGN);
264 357 : failed += keywords_insert("FROM", FROM);
265 357 : failed += keywords_insert("FWF", FWF);
266 :
267 357 : failed += keywords_insert("BIG", BIG);
268 357 : failed += keywords_insert("LITTLE", LITTLE);
269 357 : failed += keywords_insert("NATIVE", NATIVE);
270 357 : failed += keywords_insert("ENDIAN", ENDIAN);
271 :
272 357 : failed += keywords_insert("REFERENCES", REFERENCES);
273 :
274 357 : failed += keywords_insert("MATCH", MATCH);
275 357 : failed += keywords_insert("FULL", FULL);
276 357 : failed += keywords_insert("PARTIAL", PARTIAL);
277 357 : failed += keywords_insert("SIMPLE", SIMPLE);
278 :
279 357 : failed += keywords_insert("INSERT", INSERT);
280 357 : failed += keywords_insert("UPDATE", UPDATE);
281 357 : failed += keywords_insert("DELETE", sqlDELETE);
282 357 : failed += keywords_insert("TRUNCATE", TRUNCATE);
283 357 : failed += keywords_insert("MATCHED", MATCHED);
284 :
285 357 : failed += keywords_insert("ACTION", ACTION);
286 357 : failed += keywords_insert("CASCADE", CASCADE);
287 357 : failed += keywords_insert("RESTRICT", RESTRICT);
288 357 : failed += keywords_insert("FIRST", FIRST);
289 357 : failed += keywords_insert("GLOBAL", GLOBAL);
290 357 : failed += keywords_insert("GROUP", sqlGROUP);
291 357 : failed += keywords_insert("GROUPING", GROUPING);
292 357 : failed += keywords_insert("ROLLUP", ROLLUP);
293 357 : failed += keywords_insert("CUBE", CUBE);
294 357 : failed += keywords_insert("HAVING", HAVING);
295 357 : failed += keywords_insert("ILIKE", ILIKE);
296 357 : failed += keywords_insert("IMPRINTS", IMPRINTS);
297 357 : failed += keywords_insert("IN", sqlIN);
298 357 : failed += keywords_insert("INNER", INNER);
299 357 : failed += keywords_insert("INTO", INTO);
300 357 : failed += keywords_insert("IS", IS);
301 357 : failed += keywords_insert("JOIN", JOIN);
302 357 : failed += keywords_insert("KEY", KEY);
303 357 : failed += keywords_insert("LATERAL", LATERAL);
304 357 : failed += keywords_insert("LEFT", LEFT);
305 357 : failed += keywords_insert("LIKE", LIKE);
306 357 : failed += keywords_insert("LIMIT", LIMIT);
307 357 : failed += keywords_insert("SAMPLE", SAMPLE);
308 357 : failed += keywords_insert("SEED", SEED);
309 357 : failed += keywords_insert("LAST", LAST);
310 357 : failed += keywords_insert("LOCAL", LOCAL);
311 357 : failed += keywords_insert("NATURAL", NATURAL);
312 357 : failed += keywords_insert("NOT", NOT);
313 357 : failed += keywords_insert("NULL", sqlNULL);
314 357 : failed += keywords_insert("NULLS", NULLS);
315 357 : failed += keywords_insert("OFFSET", OFFSET);
316 357 : failed += keywords_insert("ON", ON);
317 357 : failed += keywords_insert("OPTIONS", OPTIONS);
318 357 : failed += keywords_insert("OPTION", OPTION);
319 357 : failed += keywords_insert("OR", OR);
320 357 : failed += keywords_insert("ORDER", ORDER);
321 357 : failed += keywords_insert("ORDERED", ORDERED);
322 357 : failed += keywords_insert("OUTER", OUTER);
323 357 : failed += keywords_insert("OVER", OVER);
324 357 : failed += keywords_insert("PARTITION", PARTITION);
325 357 : failed += keywords_insert("PATH", PATH);
326 357 : failed += keywords_insert("PRECISION", PRECISION);
327 357 : failed += keywords_insert("PRIMARY", PRIMARY);
328 :
329 357 : failed += keywords_insert("USER", USER);
330 357 : failed += keywords_insert("RENAME", RENAME);
331 357 : failed += keywords_insert("UNENCRYPTED", UNENCRYPTED);
332 357 : failed += keywords_insert("ENCRYPTED", ENCRYPTED);
333 357 : failed += keywords_insert("PASSWORD", PASSWORD);
334 357 : failed += keywords_insert("GRANT", GRANT);
335 357 : failed += keywords_insert("REVOKE", REVOKE);
336 357 : failed += keywords_insert("ROLE", ROLE);
337 357 : failed += keywords_insert("ADMIN", ADMIN);
338 357 : failed += keywords_insert("PRIVILEGES", PRIVILEGES);
339 357 : failed += keywords_insert("PUBLIC", PUBLIC);
340 357 : failed += keywords_insert("CURRENT_USER", CURRENT_USER);
341 357 : failed += keywords_insert("CURRENT_ROLE", CURRENT_ROLE);
342 357 : failed += keywords_insert("SESSION_USER", SESSION_USER);
343 357 : failed += keywords_insert("CURRENT_SCHEMA", CURRENT_SCHEMA);
344 357 : failed += keywords_insert("SESSION", sqlSESSION);
345 357 : failed += keywords_insert("MAX_MEMORY", MAX_MEMORY);
346 357 : failed += keywords_insert("MAX_WORKERS", MAX_WORKERS);
347 357 : failed += keywords_insert("OPTIMIZER", OPTIMIZER);
348 :
349 357 : failed += keywords_insert("RIGHT", RIGHT);
350 357 : failed += keywords_insert("SCHEMA", SCHEMA);
351 357 : failed += keywords_insert("SELECT", SELECT);
352 357 : failed += keywords_insert("SET", SET);
353 357 : failed += keywords_insert("SETS", SETS);
354 357 : failed += keywords_insert("AUTO_COMMIT", AUTO_COMMIT);
355 :
356 357 : failed += keywords_insert("ALL", ALL);
357 357 : failed += keywords_insert("ANY", ANY);
358 357 : failed += keywords_insert("SOME", SOME);
359 357 : failed += keywords_insert("EVERY", ANY);
360 : /*
361 : failed += keywords_insert("SQLCODE", SQLCODE );
362 : */
363 357 : failed += keywords_insert("COLUMN", COLUMN);
364 357 : failed += keywords_insert("TABLE", TABLE);
365 357 : failed += keywords_insert("TEMPORARY", TEMPORARY);
366 357 : failed += keywords_insert("TEMP", TEMP);
367 357 : failed += keywords_insert("REMOTE", REMOTE);
368 357 : failed += keywords_insert("MERGE", MERGE);
369 357 : failed += keywords_insert("REPLICA", REPLICA);
370 357 : failed += keywords_insert("UNLOGGED", UNLOGGED);
371 357 : failed += keywords_insert("TO", TO);
372 357 : failed += keywords_insert("UNION", UNION);
373 357 : failed += keywords_insert("EXCEPT", EXCEPT);
374 357 : failed += keywords_insert("INTERSECT", INTERSECT);
375 357 : failed += keywords_insert("CORRESPONDING", CORRESPONDING);
376 357 : failed += keywords_insert("UNIQUE", UNIQUE);
377 357 : failed += keywords_insert("USING", USING);
378 357 : failed += keywords_insert("VALUES", VALUES);
379 357 : failed += keywords_insert("VIEW", VIEW);
380 357 : failed += keywords_insert("WHERE", WHERE);
381 357 : failed += keywords_insert("WITH", WITH);
382 357 : failed += keywords_insert("WITHIN", WITHIN);
383 357 : failed += keywords_insert("WITHOUT", WITHOUT);
384 357 : failed += keywords_insert("DATA", DATA);
385 :
386 357 : failed += keywords_insert("DATE", sqlDATE);
387 357 : failed += keywords_insert("TIME", TIME);
388 357 : failed += keywords_insert("TIMESTAMP", TIMESTAMP);
389 357 : failed += keywords_insert("INTERVAL", INTERVAL);
390 357 : failed += keywords_insert("CURRENT_DATE", CURRENT_DATE);
391 357 : failed += keywords_insert("CURRENT_TIME", CURRENT_TIME);
392 357 : failed += keywords_insert("CURRENT_TIMESTAMP", CURRENT_TIMESTAMP);
393 357 : failed += keywords_insert("CURRENT_TIMEZONE", CURRENT_TIMEZONE);
394 357 : failed += keywords_insert("NOW", CURRENT_TIMESTAMP);
395 357 : failed += keywords_insert("LOCALTIME", LOCALTIME);
396 357 : failed += keywords_insert("LOCALTIMESTAMP", LOCALTIMESTAMP);
397 357 : failed += keywords_insert("ZONE", ZONE);
398 :
399 357 : failed += keywords_insert("CENTURY", CENTURY);
400 357 : failed += keywords_insert("DECADE", DECADE);
401 357 : failed += keywords_insert("YEAR", YEAR);
402 357 : failed += keywords_insert("QUARTER", QUARTER);
403 357 : failed += keywords_insert("MONTH", MONTH);
404 357 : failed += keywords_insert("WEEK", WEEK);
405 357 : failed += keywords_insert("DOW", DOW);
406 357 : failed += keywords_insert("DOY", DOY);
407 357 : failed += keywords_insert("DAY", DAY);
408 357 : failed += keywords_insert("HOUR", HOUR);
409 357 : failed += keywords_insert("MINUTE", MINUTE);
410 357 : failed += keywords_insert("SECOND", SECOND);
411 357 : failed += keywords_insert("EPOCH", EPOCH);
412 :
413 357 : failed += keywords_insert("POSITION", POSITION);
414 357 : failed += keywords_insert("SUBSTRING", SUBSTRING);
415 357 : failed += keywords_insert("SPLIT_PART", SPLIT_PART);
416 357 : failed += keywords_insert("TRIM", TRIM);
417 357 : failed += keywords_insert("LEADING", LEADING);
418 357 : failed += keywords_insert("TRAILING", TRAILING);
419 357 : failed += keywords_insert("BOTH", BOTH);
420 :
421 357 : failed += keywords_insert("CASE", CASE);
422 357 : failed += keywords_insert("WHEN", WHEN);
423 357 : failed += keywords_insert("THEN", THEN);
424 357 : failed += keywords_insert("ELSE", ELSE);
425 357 : failed += keywords_insert("END", END);
426 357 : failed += keywords_insert("NULLIF", NULLIF);
427 357 : failed += keywords_insert("COALESCE", COALESCE);
428 357 : failed += keywords_insert("ELSEIF", ELSEIF);
429 357 : failed += keywords_insert("IF", IF);
430 357 : failed += keywords_insert("WHILE", WHILE);
431 357 : failed += keywords_insert("DO", DO);
432 :
433 357 : failed += keywords_insert("COMMIT", COMMIT);
434 357 : failed += keywords_insert("ROLLBACK", ROLLBACK);
435 357 : failed += keywords_insert("SAVEPOINT", SAVEPOINT);
436 357 : failed += keywords_insert("RELEASE", RELEASE);
437 357 : failed += keywords_insert("WORK", WORK);
438 357 : failed += keywords_insert("CHAIN", CHAIN);
439 357 : failed += keywords_insert("PRESERVE", PRESERVE);
440 357 : failed += keywords_insert("ROWS", ROWS);
441 357 : failed += keywords_insert("NO", NO);
442 357 : failed += keywords_insert("START", START);
443 357 : failed += keywords_insert("TRANSACTION", TRANSACTION);
444 357 : failed += keywords_insert("READ", READ);
445 357 : failed += keywords_insert("WRITE", WRITE);
446 357 : failed += keywords_insert("ONLY", ONLY);
447 357 : failed += keywords_insert("ISOLATION", ISOLATION);
448 357 : failed += keywords_insert("LEVEL", LEVEL);
449 357 : failed += keywords_insert("UNCOMMITTED", UNCOMMITTED);
450 357 : failed += keywords_insert("COMMITTED", COMMITTED);
451 357 : failed += keywords_insert("REPEATABLE", sqlREPEATABLE);
452 357 : failed += keywords_insert("SNAPSHOT", SNAPSHOT);
453 357 : failed += keywords_insert("SERIALIZABLE", SERIALIZABLE);
454 357 : failed += keywords_insert("DIAGNOSTICS", DIAGNOSTICS);
455 357 : failed += keywords_insert("SIZE", sqlSIZE);
456 357 : failed += keywords_insert("STORAGE", STORAGE);
457 :
458 357 : failed += keywords_insert("TYPE", TYPE);
459 357 : failed += keywords_insert("PROCEDURE", PROCEDURE);
460 357 : failed += keywords_insert("FUNCTION", FUNCTION);
461 357 : failed += keywords_insert("LOADER", sqlLOADER);
462 357 : failed += keywords_insert("REPLACE", REPLACE);
463 :
464 : //failed += keywords_insert("FIELD", FIELD);
465 357 : failed += keywords_insert("FILTER", FILTER);
466 357 : failed += keywords_insert("AGGREGATE", AGGREGATE);
467 357 : failed += keywords_insert("RETURNS", RETURNS);
468 357 : failed += keywords_insert("EXTERNAL", EXTERNAL);
469 357 : failed += keywords_insert("NAME", sqlNAME);
470 357 : failed += keywords_insert("RETURN", RETURN);
471 357 : failed += keywords_insert("CALL", CALL);
472 357 : failed += keywords_insert("LANGUAGE", LANGUAGE);
473 :
474 357 : failed += keywords_insert("ANALYZE", ANALYZE);
475 357 : failed += keywords_insert("EXPLAIN", SQL_EXPLAIN);
476 357 : failed += keywords_insert("PLAN", SQL_PLAN);
477 357 : failed += keywords_insert("TRACE", SQL_TRACE);
478 357 : failed += keywords_insert("PREPARE", PREPARE);
479 357 : failed += keywords_insert("PREP", PREP);
480 357 : failed += keywords_insert("EXECUTE", EXECUTE);
481 357 : failed += keywords_insert("EXEC", EXEC);
482 357 : failed += keywords_insert("DEALLOCATE", DEALLOCATE);
483 :
484 357 : failed += keywords_insert("INDEX", INDEX);
485 :
486 357 : failed += keywords_insert("SEQUENCE", SEQUENCE);
487 357 : failed += keywords_insert("RESTART", RESTART);
488 357 : failed += keywords_insert("INCREMENT", INCREMENT);
489 357 : failed += keywords_insert("MAXVALUE", MAXVALUE);
490 357 : failed += keywords_insert("MINVALUE", MINVALUE);
491 357 : failed += keywords_insert("CYCLE", CYCLE);
492 357 : failed += keywords_insert("CACHE", CACHE);
493 357 : failed += keywords_insert("NEXT", NEXT);
494 357 : failed += keywords_insert("VALUE", VALUE);
495 357 : failed += keywords_insert("GENERATED", GENERATED);
496 357 : failed += keywords_insert("ALWAYS", ALWAYS);
497 357 : failed += keywords_insert("IDENTITY", IDENTITY);
498 357 : failed += keywords_insert("SERIAL", SERIAL);
499 357 : failed += keywords_insert("BIGSERIAL", BIGSERIAL);
500 357 : failed += keywords_insert("AUTO_INCREMENT", AUTO_INCREMENT);
501 357 : failed += keywords_insert("CONTINUE", CONTINUE);
502 :
503 357 : failed += keywords_insert("TRIGGER", TRIGGER);
504 357 : failed += keywords_insert("ATOMIC", ATOMIC);
505 357 : failed += keywords_insert("BEGIN", BEGIN);
506 357 : failed += keywords_insert("OF", OF);
507 357 : failed += keywords_insert("BEFORE", BEFORE);
508 357 : failed += keywords_insert("AFTER", AFTER);
509 357 : failed += keywords_insert("ROW", ROW);
510 357 : failed += keywords_insert("STATEMENT", STATEMENT);
511 357 : failed += keywords_insert("NEW", sqlNEW);
512 357 : failed += keywords_insert("OLD", OLD);
513 357 : failed += keywords_insert("EACH", EACH);
514 357 : failed += keywords_insert("REFERENCING", REFERENCING);
515 :
516 357 : failed += keywords_insert("RANGE", RANGE);
517 357 : failed += keywords_insert("UNBOUNDED", UNBOUNDED);
518 357 : failed += keywords_insert("PRECEDING", PRECEDING);
519 357 : failed += keywords_insert("FOLLOWING", FOLLOWING);
520 357 : failed += keywords_insert("CURRENT", CURRENT);
521 357 : failed += keywords_insert("EXCLUDE", EXCLUDE);
522 357 : failed += keywords_insert("OTHERS", OTHERS);
523 357 : failed += keywords_insert("TIES", TIES);
524 357 : failed += keywords_insert("GROUPS", GROUPS);
525 357 : failed += keywords_insert("WINDOW", WINDOW);
526 357 : failed += keywords_insert("QUALIFY", QUALIFY);
527 :
528 : /* special SQL/XML keywords */
529 357 : failed += keywords_insert("XMLCOMMENT", XMLCOMMENT);
530 357 : failed += keywords_insert("XMLCONCAT", XMLCONCAT);
531 357 : failed += keywords_insert("XMLDOCUMENT", XMLDOCUMENT);
532 357 : failed += keywords_insert("XMLELEMENT", XMLELEMENT);
533 357 : failed += keywords_insert("XMLATTRIBUTES", XMLATTRIBUTES);
534 357 : failed += keywords_insert("XMLFOREST", XMLFOREST);
535 357 : failed += keywords_insert("XMLPARSE", XMLPARSE);
536 357 : failed += keywords_insert("STRIP", STRIP);
537 357 : failed += keywords_insert("WHITESPACE", WHITESPACE);
538 357 : failed += keywords_insert("XMLPI", XMLPI);
539 357 : failed += keywords_insert("XMLQUERY", XMLQUERY);
540 357 : failed += keywords_insert("PASSING", PASSING);
541 357 : failed += keywords_insert("XMLTEXT", XMLTEXT);
542 357 : failed += keywords_insert("NIL", NIL);
543 357 : failed += keywords_insert("REF", REF);
544 357 : failed += keywords_insert("ABSENT", ABSENT);
545 357 : failed += keywords_insert("DOCUMENT", DOCUMENT);
546 357 : failed += keywords_insert("ELEMENT", ELEMENT);
547 357 : failed += keywords_insert("CONTENT", CONTENT);
548 357 : failed += keywords_insert("XMLNAMESPACES", XMLNAMESPACES);
549 357 : failed += keywords_insert("NAMESPACE", NAMESPACE);
550 357 : failed += keywords_insert("XMLVALIDATE", XMLVALIDATE);
551 357 : failed += keywords_insert("RETURNING", RETURNING);
552 357 : failed += keywords_insert("RECURSIVE", RECURSIVE);
553 357 : failed += keywords_insert("LOCATION", LOCATION);
554 357 : failed += keywords_insert("ID", ID);
555 357 : failed += keywords_insert("ACCORDING", ACCORDING);
556 357 : failed += keywords_insert("XMLSCHEMA", XMLSCHEMA);
557 357 : failed += keywords_insert("URI", URI);
558 357 : failed += keywords_insert("XMLAGG", XMLAGG);
559 :
560 : /* keywords for opengis */
561 357 : failed += keywords_insert("GEOMETRY", GEOMETRY);
562 :
563 357 : failed += keywords_insert("POINT", GEOMETRYSUBTYPE);
564 357 : failed += keywords_insert("LINESTRING", GEOMETRYSUBTYPE);
565 357 : failed += keywords_insert("POLYGON", GEOMETRYSUBTYPE);
566 357 : failed += keywords_insert("MULTIPOINT", GEOMETRYSUBTYPE);
567 357 : failed += keywords_insert("MULTILINESTRING", GEOMETRYSUBTYPE);
568 357 : failed += keywords_insert("MULTIPOLYGON", GEOMETRYSUBTYPE);
569 357 : failed += keywords_insert("GEOMETRYCOLLECTION", GEOMETRYSUBTYPE);
570 :
571 357 : failed += keywords_insert("POINTZ", GEOMETRYSUBTYPE);
572 357 : failed += keywords_insert("LINESTRINGZ", GEOMETRYSUBTYPE);
573 357 : failed += keywords_insert("POLYGONZ", GEOMETRYSUBTYPE);
574 357 : failed += keywords_insert("MULTIPOINTZ", GEOMETRYSUBTYPE);
575 357 : failed += keywords_insert("MULTILINESTRINGZ", GEOMETRYSUBTYPE);
576 357 : failed += keywords_insert("MULTIPOLYGONZ", GEOMETRYSUBTYPE);
577 357 : failed += keywords_insert("GEOMETRYCOLLECTIONZ", GEOMETRYSUBTYPE);
578 :
579 357 : failed += keywords_insert("POINTM", GEOMETRYSUBTYPE);
580 357 : failed += keywords_insert("LINESTRINGM", GEOMETRYSUBTYPE);
581 357 : failed += keywords_insert("POLYGONM", GEOMETRYSUBTYPE);
582 357 : failed += keywords_insert("MULTIPOINTM", GEOMETRYSUBTYPE);
583 357 : failed += keywords_insert("MULTILINESTRINGM", GEOMETRYSUBTYPE);
584 357 : failed += keywords_insert("MULTIPOLYGONM", GEOMETRYSUBTYPE);
585 357 : failed += keywords_insert("GEOMETRYCOLLECTIONM", GEOMETRYSUBTYPE);
586 :
587 357 : failed += keywords_insert("POINTZM", GEOMETRYSUBTYPE);
588 357 : failed += keywords_insert("LINESTRINGZM", GEOMETRYSUBTYPE);
589 357 : failed += keywords_insert("POLYGONZM", GEOMETRYSUBTYPE);
590 357 : failed += keywords_insert("MULTIPOINTZM", GEOMETRYSUBTYPE);
591 357 : failed += keywords_insert("MULTILINESTRINGZM", GEOMETRYSUBTYPE);
592 357 : failed += keywords_insert("MULTIPOLYGONZM", GEOMETRYSUBTYPE);
593 357 : failed += keywords_insert("GEOMETRYCOLLECTIONZM", GEOMETRYSUBTYPE);
594 357 : failed += keywords_insert("LOGIN", LOGIN);
595 : // odbc keywords
596 357 : failed += keywords_insert("d", ODBC_DATE_ESCAPE_PREFIX);
597 357 : failed += keywords_insert("t", ODBC_TIME_ESCAPE_PREFIX);
598 357 : failed += keywords_insert("ts", ODBC_TIMESTAMP_ESCAPE_PREFIX);
599 357 : failed += keywords_insert("guid", ODBC_GUID_ESCAPE_PREFIX);
600 357 : failed += keywords_insert("fn", ODBC_FUNC_ESCAPE_PREFIX);
601 357 : failed += keywords_insert("oj", ODBC_OJ_ESCAPE_PREFIX);
602 357 : failed += keywords_insert("DAYNAME", DAYNAME);
603 357 : failed += keywords_insert("IFNULL", IFNULL);
604 357 : failed += keywords_insert("MONTHNAME", MONTHNAME);
605 357 : failed += keywords_insert("TIMESTAMPADD", TIMESTAMPADD);
606 357 : failed += keywords_insert("TIMESTAMPDIFF", TIMESTAMPDIFF);
607 357 : failed += keywords_insert("SQL_BIGINT", SQL_BIGINT);
608 357 : failed += keywords_insert("SQL_BINARY", SQL_BINARY);
609 357 : failed += keywords_insert("SQL_BIT", SQL_BIT);
610 357 : failed += keywords_insert("SQL_CHAR", SQL_CHAR);
611 357 : failed += keywords_insert("SQL_DATE", SQL_DATE);
612 357 : failed += keywords_insert("SQL_DECIMAL", SQL_DECIMAL);
613 357 : failed += keywords_insert("SQL_DOUBLE", SQL_DOUBLE);
614 357 : failed += keywords_insert("SQL_FLOAT", SQL_FLOAT);
615 357 : failed += keywords_insert("SQL_GUID", SQL_GUID);
616 357 : failed += keywords_insert("SQL_HUGEINT", SQL_HUGEINT);
617 357 : failed += keywords_insert("SQL_INTEGER", SQL_INTEGER);
618 357 : failed += keywords_insert("SQL_INTERVAL_DAY", SQL_INTERVAL_DAY);
619 357 : failed += keywords_insert("SQL_INTERVAL_DAY_TO_HOUR", SQL_INTERVAL_DAY_TO_HOUR);
620 357 : failed += keywords_insert("SQL_INTERVAL_DAY_TO_MINUTE", SQL_INTERVAL_DAY_TO_MINUTE);
621 357 : failed += keywords_insert("SQL_INTERVAL_DAY_TO_SECOND", SQL_INTERVAL_DAY_TO_SECOND);
622 357 : failed += keywords_insert("SQL_INTERVAL_HOUR", SQL_INTERVAL_HOUR);
623 357 : failed += keywords_insert("SQL_INTERVAL_HOUR_TO_MINUTE", SQL_INTERVAL_HOUR_TO_MINUTE);
624 357 : failed += keywords_insert("SQL_INTERVAL_HOUR_TO_SECOND", SQL_INTERVAL_HOUR_TO_SECOND);
625 357 : failed += keywords_insert("SQL_INTERVAL_MINUTE", SQL_INTERVAL_MINUTE);
626 357 : failed += keywords_insert("SQL_INTERVAL_MINUTE_TO_SECOND", SQL_INTERVAL_MINUTE_TO_SECOND);
627 357 : failed += keywords_insert("SQL_INTERVAL_MONTH", SQL_INTERVAL_MONTH);
628 357 : failed += keywords_insert("SQL_INTERVAL_SECOND", SQL_INTERVAL_SECOND);
629 357 : failed += keywords_insert("SQL_INTERVAL_YEAR", SQL_INTERVAL_YEAR);
630 357 : failed += keywords_insert("SQL_INTERVAL_YEAR_TO_MONTH", SQL_INTERVAL_YEAR_TO_MONTH);
631 357 : failed += keywords_insert("SQL_LONGVARBINARY", SQL_LONGVARBINARY);
632 357 : failed += keywords_insert("SQL_LONGVARCHAR", SQL_LONGVARCHAR);
633 357 : failed += keywords_insert("SQL_NUMERIC", SQL_NUMERIC);
634 357 : failed += keywords_insert("SQL_REAL", SQL_REAL);
635 357 : failed += keywords_insert("SQL_SMALLINT", SQL_SMALLINT);
636 357 : failed += keywords_insert("SQL_TIME", SQL_TIME);
637 357 : failed += keywords_insert("SQL_TIMESTAMP", SQL_TIMESTAMP);
638 357 : failed += keywords_insert("SQL_TINYINT", SQL_TINYINT);
639 357 : failed += keywords_insert("SQL_VARBINARY", SQL_VARBINARY);
640 357 : failed += keywords_insert("SQL_VARCHAR", SQL_VARCHAR);
641 357 : failed += keywords_insert("SQL_WCHAR", SQL_WCHAR);
642 357 : failed += keywords_insert("SQL_WLONGVARCHAR", SQL_WLONGVARCHAR);
643 357 : failed += keywords_insert("SQL_WVARCHAR", SQL_WVARCHAR);
644 357 : failed += keywords_insert("SQL_TSI_FRAC_SECOND", SQL_TSI_FRAC_SECOND);
645 357 : failed += keywords_insert("SQL_TSI_SECOND", SQL_TSI_SECOND);
646 357 : failed += keywords_insert("SQL_TSI_MINUTE", SQL_TSI_MINUTE);
647 357 : failed += keywords_insert("SQL_TSI_HOUR", SQL_TSI_HOUR);
648 357 : failed += keywords_insert("SQL_TSI_DAY", SQL_TSI_DAY);
649 357 : failed += keywords_insert("SQL_TSI_WEEK", SQL_TSI_WEEK);
650 357 : failed += keywords_insert("SQL_TSI_MONTH", SQL_TSI_MONTH);
651 357 : failed += keywords_insert("SQL_TSI_QUARTER", SQL_TSI_QUARTER);
652 357 : failed += keywords_insert("SQL_TSI_YEAR", SQL_TSI_YEAR);
653 :
654 357 : failed += keywords_insert("LEAST", MARGFUNC);
655 357 : failed += keywords_insert("GREATEST", MARGFUNC);
656 357 : return failed;
657 : }
658 :
659 : #define find_keyword_bs(lc, s) find_keyword(lc->rs->buf+lc->rs->pos+s)
660 :
661 : void
662 251986 : scanner_init(struct scanner *s, bstream *rs, stream *ws)
663 : {
664 503972 : *s = (struct scanner) {
665 : .rs = rs,
666 : .ws = ws,
667 : .mode = LINE_N,
668 251986 : .raw_string_mode = GDKgetenv_istrue("raw_strings"),
669 : .aborted = false,
670 : };
671 251986 : }
672 :
673 : void
674 1398086 : scanner_query_processed(struct scanner *s)
675 : {
676 1398086 : int cur;
677 :
678 1398086 : if (s->yybak) {
679 520612 : s->rs->buf[s->rs->pos + s->yycur] = s->yybak;
680 520612 : s->yybak = 0;
681 : }
682 1398086 : if (s->rs) {
683 1398086 : s->rs->pos += s->yycur;
684 : /* completely eat the query including white space after the ; */
685 2564596 : while (s->rs->pos < s->rs->len &&
686 2161110 : (cur = s->rs->buf[s->rs->pos], iswspace(cur))) {
687 1166510 : s->rs->pos++;
688 : }
689 : }
690 : /*assert(s->rs->pos <= s->rs->len);*/
691 1398086 : s->yycur = 0;
692 1398086 : s->started = 0;
693 1398086 : s->as = 0;
694 1398086 : s->schema = NULL;
695 1398086 : }
696 :
697 : static int
698 33 : scanner_error(mvc *lc, int cur)
699 : {
700 33 : switch (cur) {
701 0 : case EOF:
702 0 : (void) sql_error(lc, 1, SQLSTATE(42000) "Unexpected end of input");
703 0 : return EOF;
704 33 : default:
705 : /* on Windows at least, iswcntrl returns TRUE for
706 : * U+FEFF, but we just want consistent error
707 : * messages */
708 33 : (void) sql_error(lc, 1, SQLSTATE(42000) "Unexpected%s character (U+%04X)", iswcntrl(cur) && cur != 0xFEFF ? " control" : "", (unsigned) cur);
709 : }
710 33 : return LEX_ERROR;
711 : }
712 :
713 :
714 : /*
715 : UTF-8 encoding is as follows:
716 : U-00000000 - U-0000007F: 0xxxxxxx
717 : U-00000080 - U-000007FF: 110xxxxx 10xxxxxx
718 : U-00000800 - U-0000FFFF: 1110xxxx 10xxxxxx 10xxxxxx
719 : U-00010000 - U-001FFFFF: 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx
720 : U-00200000 - U-03FFFFFF: 111110xx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx
721 : U-04000000 - U-7FFFFFFF: 1111110x 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx
722 : */
723 : /* To be correctly coded UTF-8, the sequence should be the shortest
724 : possible encoding of the value being encoded. This means that for
725 : an encoding of length n+1 (1 <= n <= 5), at least one of the bits in
726 : utf8chkmsk[n] should be non-zero (else the encoding could be
727 : shorter).
728 : */
729 : static const int utf8chkmsk[] = {
730 : 0x0000007f,
731 : 0x00000780,
732 : 0x0000f800,
733 : 0x001f0000,
734 : 0x03e00000,
735 : 0x7c000000
736 : };
737 :
738 : static void
739 32682091 : utf8_putchar(struct scanner *lc, int ch)
740 : {
741 32682091 : if ((ch) < 0x80) {
742 32682086 : lc->yycur--;
743 5 : } else if ((ch) < 0x800) {
744 0 : lc->yycur -= 2;
745 5 : } else if ((ch) < 0x10000) {
746 5 : lc->yycur -= 3;
747 : } else {
748 0 : lc->yycur -= 4;
749 : }
750 32682091 : }
751 :
752 : static inline int
753 142242748 : scanner_read_more(struct scanner *lc, size_t n)
754 : {
755 142242748 : bstream *b = lc->rs;
756 142242748 : bool more = false;
757 :
758 :
759 142242748 : if (lc->aborted)
760 : return EOF;
761 142246341 : while (b->len < b->pos + lc->yycur + n) {
762 :
763 138734 : if (lc->mode == LINE_1 || !lc->started)
764 : return EOF;
765 :
766 : /* query is not finished ask for more */
767 6203 : if (b->eof || !isa_block_stream(b->s)) {
768 4403 : if (bstream_getoob(b)) {
769 0 : lc->aborted = true;
770 0 : return EOF;
771 : }
772 1800 : if (mnstr_write(lc->ws, PROMPT2, sizeof(PROMPT2) - 1, 1) == 1)
773 1800 : mnstr_flush(lc->ws, MNSTR_FLUSH_DATA);
774 1800 : b->eof = false;
775 1800 : more = true;
776 : }
777 : /* we need more query text */
778 3600 : if (bstream_next(b) < 0) {
779 0 : if (mnstr_errnr(b->s) == MNSTR_INTERRUPT) {
780 : // now what?
781 0 : lc->errstr = "Query aborted";
782 0 : lc->aborted = true;
783 0 : mnstr_clearerr(b->s);
784 : }
785 0 : return EOF;
786 3600 : } else if (/* we asked for more data but didn't get any */
787 1800 : (more && b->eof && b->len < b->pos + lc->yycur + n))
788 : return EOF;
789 3593 : if (more && b->pos + lc->yycur + 2 == b->len && b->buf[b->pos + lc->yycur] == '\200' && b->buf[b->pos + lc->yycur + 1] == '\n') {
790 0 : lc->errstr = "Query aborted";
791 0 : b->len -= 2;
792 0 : b->buf[b->len] = 0;
793 0 : return EOF;
794 : }
795 : }
796 : return 1;
797 : }
798 :
799 : static inline int
800 140957503 : scanner_getc(struct scanner *lc)
801 : {
802 140957503 : bstream *b = lc->rs;
803 140957503 : unsigned char *s = NULL;
804 140957503 : int c, m, n, mask;
805 :
806 140957503 : if (scanner_read_more(lc, 1) == EOF) {
807 : //lc->errstr = SQLSTATE(42000) "end of input stream";
808 : return EOF;
809 : }
810 140824567 : lc->errstr = NULL;
811 :
812 140824567 : s = (unsigned char *) b->buf + b->pos + lc->yycur++;
813 140824567 : if (((c = *s) & 0x80) == 0) {
814 : /* 7-bit char */
815 : return c;
816 : }
817 88250 : for (n = 0, m = 0x40; c & m; n++, m >>= 1)
818 : ;
819 : /* n now is number of 10xxxxxx bytes that should follow */
820 29443 : if (n == 0 || n >= 6 || (b->pos + n) > b->len) {
821 : /* incorrect UTF-8 sequence */
822 : /* n==0: c == 10xxxxxx */
823 : /* n>=6: c == 1111111x */
824 0 : lc->errstr = SQLSTATE(42000) "invalid start of UTF-8 sequence";
825 0 : goto error;
826 : }
827 :
828 29443 : if (scanner_read_more(lc, (size_t) n) == EOF)
829 : return EOF;
830 29443 : s = (unsigned char *) b->buf + b->pos + lc->yycur;
831 :
832 29443 : mask = utf8chkmsk[n];
833 29443 : c &= ~(0xFFC0 >> n); /* remove non-x bits */
834 88249 : while (--n >= 0) {
835 58807 : c <<= 6;
836 58807 : lc->yycur++;
837 58807 : if (((m = *s++) & 0xC0) != 0x80) {
838 : /* incorrect UTF-8 sequence: byte is not 10xxxxxx */
839 : /* this includes end-of-string (m == 0) */
840 1 : lc->errstr = SQLSTATE(42000) "invalid continuation in UTF-8 sequence";
841 1 : goto error;
842 : }
843 58806 : c |= m & 0x3F;
844 : }
845 29442 : if ((c & mask) == 0) {
846 : /* incorrect UTF-8 sequence: not shortest possible */
847 0 : lc->errstr = SQLSTATE(42000) "not shortest possible UTF-8 sequence";
848 0 : goto error;
849 : }
850 :
851 : return c;
852 :
853 1 : error:
854 1 : if (b->pos + lc->yycur < b->len) /* skip bogus char */
855 0 : lc->yycur++;
856 : return EOF;
857 : }
858 :
859 : static int
860 29201293 : scanner_token(struct scanner *lc, int token)
861 : {
862 29201293 : lc->yybak = lc->rs->buf[lc->rs->pos + lc->yycur];
863 29201293 : lc->rs->buf[lc->rs->pos + lc->yycur] = 0;
864 29201293 : lc->yyval = token;
865 29201293 : return lc->yyval;
866 : }
867 :
868 : static int
869 2160781 : scanner_string(mvc *c, int quote, bool escapes)
870 : {
871 2160781 : struct scanner *lc = &c->scanner;
872 2160781 : bstream *rs = lc->rs;
873 2160781 : int cur = quote;
874 2160781 : bool escape = false;
875 2160781 : const size_t limit = quote == '"' ? 1 << 11 : 1 << 30;
876 :
877 2160781 : lc->started = 1;
878 2198929 : while (cur != EOF) {
879 2198914 : size_t pos = 0;
880 2198914 : const size_t yycur = rs->pos + lc->yycur;
881 :
882 35747924 : while (cur != EOF && (quote != '"' || cur != 0xFEFF) && pos < limit &&
883 33549010 : (((cur = rs->buf[yycur + pos++]) & 0x80) == 0) &&
884 67068566 : cur && (cur != quote || escape)) {
885 31350097 : if (escapes && cur == '\\')
886 6840 : escape = !escape;
887 : else
888 : escape = false;
889 : }
890 2198914 : if (pos == limit) {
891 0 : (void) sql_error(c, 2, SQLSTATE(42000) "string too long");
892 0 : return LEX_ERROR;
893 : }
894 : /* BOM character not allowed as an identifier */
895 2198914 : if (cur == EOF || (quote == '"' && cur == 0xFEFF))
896 1 : return scanner_error(c, cur);
897 2198913 : lc->yycur += pos;
898 : /* check for quote escaped quote: Obscure SQL Rule */
899 2198913 : if (cur == quote && rs->buf[yycur + pos] == quote) {
900 8707 : lc->yycur++;
901 8707 : continue;
902 : }
903 2190206 : assert(yycur + pos <= rs->len + 1);
904 2190206 : if (cur == quote && !escape) {
905 2160751 : return scanner_token(lc, STRING);
906 : }
907 29455 : lc->yycur--; /* go back to current (possibly invalid) char */
908 : /* long utf8, if correct isn't the quote */
909 29455 : if (!cur) {
910 30 : if (lc->rs->len >= lc->rs->pos + lc->yycur + 1) {
911 14 : (void) sql_error(c, 2, SQLSTATE(42000) "NULL byte in string");
912 14 : return LEX_ERROR;
913 : }
914 16 : cur = scanner_read_more(lc, 1);
915 : } else {
916 29425 : cur = scanner_getc(lc);
917 : }
918 : }
919 15 : (void) sql_error(c, 2, "%s", lc->errstr ? lc->errstr : SQLSTATE(42000) "Unexpected end of input");
920 15 : return EOF;
921 : }
922 :
923 : /* scan a structure {blah} into a string. We only count the matching {}
924 : * unless escaped. We do not consider embeddings in string literals yet
925 : */
926 :
927 : static int
928 234 : scanner_body(mvc *c)
929 : {
930 234 : struct scanner *lc = &c->scanner;
931 234 : bstream *rs = lc->rs;
932 234 : int cur = (int) 'x';
933 234 : int blk = 1;
934 234 : bool escape = false;
935 :
936 234 : lc->started = 1;
937 234 : assert(rs->buf[rs->pos + lc->yycur-1] == '{');
938 290 : while (cur != EOF) {
939 290 : size_t pos = rs->pos + lc->yycur;
940 :
941 32350 : while ((((cur = rs->buf[pos++]) & 0x80) == 0) && cur && (blk || escape)) {
942 32060 : if (cur != '\\')
943 : escape = false;
944 : else
945 12 : escape = !escape;
946 32060 : blk += cur =='{';
947 32060 : blk -= cur =='}';
948 : }
949 290 : lc->yycur = pos - rs->pos;
950 290 : assert(pos <= rs->len + 1);
951 290 : if (blk == 0 && !escape){
952 234 : lc->yycur--; /* go back to current (possibly invalid) char */
953 234 : return scanner_token(lc, X_BODY);
954 : }
955 56 : lc->yycur--; /* go back to current (possibly invalid) char */
956 56 : if (!cur) {
957 56 : if (lc->rs->len >= lc->rs->pos + lc->yycur + 1) {
958 0 : (void) sql_error(c, 2, SQLSTATE(42000) "NULL byte in string");
959 0 : return LEX_ERROR;
960 : }
961 56 : cur = scanner_read_more(lc, 1);
962 : } else {
963 0 : cur = scanner_getc(lc);
964 : }
965 : }
966 0 : (void) sql_error(c, 2, SQLSTATE(42000) "Unexpected end of input");
967 0 : return EOF;
968 : }
969 :
970 : static int
971 14019080 : keyword_or_ident(mvc * c, int cur)
972 : {
973 14019080 : struct scanner *lc = &c->scanner;
974 14019080 : keyword *k = NULL;
975 14019080 : size_t s;
976 :
977 14019080 : lc->started = 1;
978 14019080 : utf8_putchar(lc, cur);
979 14019085 : s = lc->yycur;
980 14019085 : lc->yyval = IDENT;
981 84194319 : while ((cur = scanner_getc(lc)) != EOF) {
982 84194216 : if (!iswalnum(cur) && cur != '_') {
983 14018982 : utf8_putchar(lc, cur);
984 14018978 : (void)scanner_token(lc, IDENT);
985 14018978 : if ((k = find_keyword_bs(lc,s)))
986 8575579 : lc->yyval = k->token;
987 14019091 : return lc->yyval;
988 : }
989 : }
990 : if (cur < 0)
991 : return cur;
992 : (void)scanner_token(lc, IDENT);
993 : if ((k = find_keyword_bs(lc,s)))
994 : lc->yyval = k->token;
995 : return lc->yyval;
996 : }
997 :
998 : static int
999 14582163 : skip_white_space(struct scanner * lc)
1000 : {
1001 18228006 : int cur;
1002 :
1003 18228006 : do {
1004 18228006 : lc->yysval = lc->yycur;
1005 18228006 : } while ((cur = scanner_getc(lc)) != EOF && iswspace(cur));
1006 14582364 : return cur;
1007 : }
1008 :
1009 : static int
1010 71222 : skip_c_comment(struct scanner * lc)
1011 : {
1012 71222 : int cur;
1013 71222 : int prev = 0;
1014 71222 : int started = lc->started;
1015 71222 : int depth = 1;
1016 :
1017 71222 : lc->started = 1;
1018 1436228 : while (depth > 0 && (cur = scanner_getc(lc)) != EOF) {
1019 1365006 : if (prev == '*' && cur == '/')
1020 71222 : depth--;
1021 1293784 : else if (prev == '/' && cur == '*') {
1022 : /* block comments can nest */
1023 0 : cur = 0; /* prevent slash-star-slash from matching */
1024 0 : depth++;
1025 : }
1026 : prev = cur;
1027 : }
1028 71222 : lc->yysval = lc->yycur;
1029 71222 : lc->started = started;
1030 : /* a comment is equivalent to a newline */
1031 71222 : return cur == EOF ? cur : '\n';
1032 : }
1033 :
1034 : static int
1035 3273 : skip_sql_comment(struct scanner * lc)
1036 : {
1037 3273 : int cur;
1038 3273 : int started = lc->started;
1039 :
1040 3273 : lc->started = 1;
1041 833837 : while ((cur = scanner_getc(lc)) != EOF && (cur != '\n'))
1042 : ;
1043 3273 : lc->yysval = lc->yycur;
1044 3273 : lc->started = started;
1045 : /* a comment is equivalent to a newline */
1046 3273 : return cur;
1047 : }
1048 :
1049 : static int tokenize(mvc * lc, int cur);
1050 :
1051 5949016 : static inline bool is_valid_decimal_digit(int cur) { return (iswdigit(cur)); }
1052 13 : static inline bool is_valid_binary_digit(int cur) { return (iswdigit(cur) && cur < '2'); }
1053 10 : static inline bool is_valid_octal_digit(int cur) { return (iswdigit(cur) && cur < '8'); }
1054 3688 : static inline bool is_valid_hexadecimal_digit(int cur) { return iswxdigit(cur); }
1055 :
1056 1969812 : static inline int check_validity_number(mvc* c, int pcur, bool initial_underscore_allowed, int *token, int type) {
1057 1969812 : struct scanner *lc = &c->scanner;
1058 1969812 : bool (*is_valid_n_ary_digit)(int);
1059 :
1060 1969812 : if (pcur == '_' && !initial_underscore_allowed) /* ERROR: initial underscore not allowed */ {
1061 0 : *token = 0;
1062 0 : return '_';
1063 : }
1064 :
1065 1969812 : switch (type) {
1066 : case BINARYNUM:
1067 : is_valid_n_ary_digit = &is_valid_binary_digit;
1068 : break;
1069 3 : case OCTALNUM:
1070 3 : is_valid_n_ary_digit = &is_valid_octal_digit;
1071 3 : break;
1072 280 : case HEXADECIMALNUM:
1073 280 : is_valid_n_ary_digit = &is_valid_hexadecimal_digit;
1074 280 : break;
1075 1969527 : default:
1076 1969527 : is_valid_n_ary_digit = &is_valid_decimal_digit;
1077 1969527 : break;
1078 : }
1079 :
1080 1969812 : if ( !(pcur == '_' || is_valid_n_ary_digit(pcur)) ) /* ERROR: first digit is not valid */ {
1081 18 : *token = 0;
1082 18 : return pcur;
1083 : }
1084 :
1085 1969646 : int cur = scanner_getc(lc);
1086 1969560 : *token = type;
1087 3984959 : while (cur != EOF) {
1088 3984970 : if (cur == '_') {
1089 25 : if (pcur == '_') /* ERROR: multiple consecutive underscores */ {
1090 2 : *token = 0;
1091 2 : return '_';
1092 : }
1093 : }
1094 3984945 : else if (!is_valid_n_ary_digit(cur))
1095 : break;
1096 2015558 : pcur = cur;
1097 2015558 : cur = scanner_getc(lc);
1098 : }
1099 :
1100 1969261 : if (pcur == '_') {
1101 3 : *token = 0;
1102 3 : if (iswalnum(cur)) /* ERROR: not a valid digit */
1103 : return cur;
1104 : else /* ERROR: number ends with underscore */
1105 : return '_';
1106 : }
1107 :
1108 : return cur;
1109 : }
1110 :
1111 : static int
1112 1956430 : number(mvc * c, int cur)
1113 : {
1114 1956430 : struct scanner *lc = &c->scanner;
1115 1956430 : int token = sqlINT;
1116 :
1117 : /* a number has one of these forms (expressed in regular expressions):
1118 : * 0x[0-9A-Fa-f]+ -- (hexadecimal) INTEGER
1119 : * \.[0-9]+ -- DECIMAL
1120 : * [0-9]+\.[0-9]* -- DECIMAL
1121 : * [0-9]+@0 -- OID
1122 : * [0-9]*\.[0-9]+[eE][-+]?[0-9]+ -- REAL
1123 : * [0-9]+(\.[0-9]*)?[eE][-+]?[0-9]+ -- REAL
1124 : * [0-9]+ -- (decimal) INTEGER
1125 : */
1126 1956430 : lc->started = 1;
1127 1956430 : if (cur == '0') {
1128 317956 : switch ((cur = scanner_getc(lc))) {
1129 2 : case 'b':
1130 2 : cur = scanner_getc(lc);
1131 2 : if ((cur = check_validity_number(c, cur, true, &token, BINARYNUM)) == EOF) return cur;
1132 : break;
1133 3 : case 'o':
1134 3 : cur = scanner_getc(lc);
1135 3 : if ((cur = check_validity_number(c, cur, true, &token, OCTALNUM)) == EOF) return cur;
1136 : break;
1137 280 : case 'x':
1138 280 : cur = scanner_getc(lc);
1139 280 : if ((cur = check_validity_number(c, cur, true, &token, HEXADECIMALNUM)) == EOF) return cur;
1140 : break;
1141 317669 : default:
1142 317669 : utf8_putchar(lc, cur);
1143 317669 : cur = '0';
1144 : }
1145 : }
1146 1956425 : if (token == sqlINT) {
1147 1956125 : if ((cur = check_validity_number(c, cur, false, &token, sqlINT)) == EOF) return cur;
1148 1955595 : if (cur == '@') {
1149 0 : if (token == sqlINT) {
1150 0 : cur = scanner_getc(lc);
1151 0 : if (cur == EOF)
1152 : return cur;
1153 0 : if (cur == '0') {
1154 0 : cur = scanner_getc(lc);
1155 0 : if (cur == EOF)
1156 : return cur;
1157 0 : token = OIDNUM;
1158 : } else {
1159 : /* number + '@' not followed by 0: show '@' as erroneous */
1160 0 : utf8_putchar(lc, cur);
1161 0 : cur = '@';
1162 0 : token = 0;
1163 : }
1164 : }
1165 : } else {
1166 1955595 : if (cur == '.') {
1167 11188 : cur = scanner_getc(lc);
1168 11188 : if (iswalnum(cur)) /* early exit for numerical forms with final . e.g. 10. */
1169 11181 : if ((cur = check_validity_number(c, cur, false, &token, INTNUM)) == EOF) return cur;
1170 : }
1171 1955595 : if (token != 0)
1172 1955601 : if (cur == 'e' || cur == 'E') {
1173 2231 : cur = scanner_getc(lc);
1174 2231 : if (cur == '+' || cur == '-')
1175 2111 : cur = scanner_getc(lc);
1176 2231 : if ((cur = check_validity_number(c, cur, false, &token, APPROXNUM)) == EOF) return cur;
1177 : }
1178 : }
1179 : }
1180 :
1181 1953664 : assert(cur != EOF);
1182 :
1183 1955895 : if (iswalnum(cur)) /* ERROR: not a valid digit */
1184 6 : token = 0;
1185 :
1186 1955895 : utf8_putchar(lc, cur);
1187 :
1188 1955853 : if (token) {
1189 1955843 : return scanner_token(lc, token);
1190 : } else {
1191 10 : (void)sql_error( c, 2, SQLSTATE(42000) "Unexpected symbol %lc", (wint_t) cur);
1192 10 : return LEX_ERROR;
1193 : }
1194 : }
1195 :
1196 : static
1197 13293798 : int scanner_symbol(mvc * c, int cur)
1198 : {
1199 13293798 : struct scanner *lc = &c->scanner;
1200 13293798 : int next = 0;
1201 13293798 : int started = lc->started;
1202 :
1203 13293798 : switch (cur) {
1204 73884 : case '/':
1205 73884 : lc->started = 1;
1206 73884 : next = scanner_getc(lc);
1207 73884 : if (next < 0)
1208 : return EOF;
1209 73884 : if (next == '*') {
1210 71222 : lc->started = started;
1211 71222 : cur = skip_c_comment(lc);
1212 71222 : if (cur < 0)
1213 : return EOF;
1214 71222 : return tokenize(c, cur);
1215 : } else {
1216 2662 : utf8_putchar(lc, next);
1217 2662 : return scanner_token(lc, cur);
1218 : }
1219 0 : case '0':
1220 : case '1':
1221 : case '2':
1222 : case '3':
1223 : case '4':
1224 : case '5':
1225 : case '6':
1226 : case '7':
1227 : case '8':
1228 : case '9':
1229 0 : return number(c, cur);
1230 8 : case '#':
1231 8 : if ((cur = skip_sql_comment(lc)) == EOF)
1232 : return cur;
1233 8 : return tokenize(c, cur);
1234 823405 : case '\'':
1235 823405 : if (lc->raw_string_mode || lc->next_string_is_raw)
1236 50 : return scanner_string(c, cur, false);
1237 823355 : return scanner_string(c, cur, true);
1238 1330151 : case '"':
1239 1330151 : return scanner_string(c, cur, false);
1240 500 : case '{':
1241 : // if previous tokens like LANGUAGE IDENT
1242 : // TODO checking on IDENT only may not be enough
1243 500 : if (lc->yylast == IDENT)
1244 234 : return scanner_body(c);
1245 266 : lc->started = 1;
1246 266 : return scanner_token(lc, cur);
1247 266 : case '}':
1248 266 : lc->started = 1;
1249 266 : return scanner_token(lc, cur);
1250 30599 : case '-':
1251 30599 : lc->started = 1;
1252 30599 : next = scanner_getc(lc);
1253 30599 : if (next < 0)
1254 : return EOF;
1255 30598 : if (next == '-') {
1256 3265 : lc->started = started;
1257 3265 : if ((cur = skip_sql_comment(lc)) == EOF)
1258 : return cur;
1259 3265 : return tokenize(c, cur);
1260 : }
1261 27333 : lc->started = 1;
1262 27333 : utf8_putchar(lc, next);
1263 27333 : return scanner_token(lc, cur);
1264 12 : case '~': /* binary not */
1265 12 : lc->started = 1;
1266 12 : next = scanner_getc(lc);
1267 12 : if (next < 0)
1268 : return EOF;
1269 12 : if (next == '=')
1270 5 : return scanner_token(lc, GEOM_MBR_EQUAL);
1271 7 : utf8_putchar(lc, next);
1272 7 : return scanner_token(lc, cur);
1273 7419115 : case '^': /* binary xor */
1274 : case '*':
1275 : case '?':
1276 : case ':':
1277 : case '%':
1278 : case '+':
1279 : case '(':
1280 : case ')':
1281 : case ',':
1282 : case '=':
1283 : case '[':
1284 : case ']':
1285 7419115 : lc->started = 1;
1286 7419115 : return scanner_token(lc, cur);
1287 6397 : case '&':
1288 6397 : lc->started = 1;
1289 6397 : cur = scanner_getc(lc);
1290 6397 : if (cur < 0)
1291 : return EOF;
1292 6397 : if (cur < 0)
1293 : return EOF;
1294 6397 : if(cur == '<') {
1295 3 : next = scanner_getc(lc);
1296 3 : if (next < 0)
1297 : return EOF;
1298 3 : if(next == '|') {
1299 0 : return scanner_token(lc, GEOM_OVERLAP_OR_BELOW);
1300 : } else {
1301 3 : utf8_putchar(lc, next); //put the char back
1302 3 : return scanner_token(lc, GEOM_OVERLAP_OR_LEFT);
1303 : }
1304 6394 : } else if(cur == '>')
1305 3 : return scanner_token(lc, GEOM_OVERLAP_OR_RIGHT);
1306 6391 : else if(cur == '&')
1307 3 : return scanner_token(lc, GEOM_OVERLAP);
1308 : else {/* binary and */
1309 6388 : utf8_putchar(lc, cur); //put the char back
1310 6388 : return scanner_token(lc, '&');
1311 : }
1312 19 : case '@':
1313 19 : lc->started = 1;
1314 19 : return scanner_token(lc, AT);
1315 1043618 : case ';':
1316 1043618 : lc->started = 0;
1317 1043618 : return scanner_token(lc, SCOLON);
1318 35 : case '!':
1319 35 : lc->started = 1;
1320 35 : cur = scanner_getc(lc);
1321 35 : if (cur < 0)
1322 : return EOF;
1323 35 : else if (cur == '=') {
1324 29 : lc->rs->buf[lc->rs->pos + lc->yycur - 2] = '<';
1325 29 : lc->rs->buf[lc->rs->pos + lc->yycur - 1] = '>';
1326 29 : return scanner_token( lc, COMPARISON);
1327 : } else {
1328 6 : utf8_putchar(lc, cur); //put the char back
1329 : }
1330 6 : return scanner_token(lc, '!');
1331 52805 : case '<':
1332 52805 : lc->started = 1;
1333 52805 : cur = scanner_getc(lc);
1334 52805 : if (cur < 0)
1335 : return EOF;
1336 52805 : if (cur == '=') {
1337 3140 : return scanner_token( lc, COMPARISON);
1338 49665 : } else if (cur == '>') {
1339 35940 : return scanner_token( lc, COMPARISON);
1340 13725 : } else if (cur == '<') {
1341 44 : next = scanner_getc(lc);
1342 44 : if (next < 0)
1343 : return EOF;
1344 44 : if (next == '=') {
1345 4 : return scanner_token( lc, LEFT_SHIFT_ASSIGN);
1346 40 : } else if (next == '|') {
1347 1 : return scanner_token(lc, GEOM_BELOW);
1348 : } else {
1349 39 : utf8_putchar(lc, next); //put the char back
1350 39 : return scanner_token( lc, LEFT_SHIFT);
1351 : }
1352 13681 : } else if(cur == '-') {
1353 19 : next = scanner_getc(lc);
1354 19 : if (next < 0)
1355 : return EOF;
1356 19 : if(next == '>') {
1357 7 : return scanner_token(lc, GEOM_DIST);
1358 : } else {
1359 : //put the characters back and fall in the next possible case
1360 12 : utf8_putchar(lc, next);
1361 12 : utf8_putchar(lc, cur);
1362 12 : return scanner_token( lc, COMPARISON);
1363 : }
1364 : } else {
1365 13662 : utf8_putchar(lc, cur);
1366 13662 : return scanner_token( lc, COMPARISON);
1367 : }
1368 47895 : case '>':
1369 47895 : lc->started = 1;
1370 47895 : cur = scanner_getc(lc);
1371 47895 : if (cur < 0)
1372 : return EOF;
1373 47895 : if (cur == '>') {
1374 2713 : cur = scanner_getc(lc);
1375 2713 : if (cur < 0)
1376 : return EOF;
1377 2713 : if (cur == '=')
1378 3 : return scanner_token( lc, RIGHT_SHIFT_ASSIGN);
1379 2710 : utf8_putchar(lc, cur);
1380 2710 : return scanner_token( lc, RIGHT_SHIFT);
1381 45182 : } else if (cur != '=') {
1382 42912 : utf8_putchar(lc, cur);
1383 42912 : return scanner_token( lc, COMPARISON);
1384 : } else {
1385 2270 : return scanner_token( lc, COMPARISON);
1386 : }
1387 2276140 : case '.':
1388 2276140 : lc->started = 1;
1389 2276140 : cur = scanner_getc(lc);
1390 2276140 : if (cur < 0)
1391 : return EOF;
1392 2276139 : if (!iswdigit(cur)) {
1393 2276125 : utf8_putchar(lc, cur);
1394 2276125 : return scanner_token( lc, '.');
1395 : } else {
1396 14 : utf8_putchar(lc, cur);
1397 14 : cur = '.';
1398 14 : return number(c, cur);
1399 : }
1400 188939 : case '|': /* binary or or string concat */
1401 188939 : lc->started = 1;
1402 188939 : cur = scanner_getc(lc);
1403 188939 : if (cur < 0)
1404 : return EOF;
1405 188939 : if (cur == '|') {
1406 188914 : return scanner_token(lc, CONCATSTRING);
1407 25 : } else if (cur == '&') {
1408 0 : next = scanner_getc(lc);
1409 0 : if (next < 0)
1410 : return EOF;
1411 0 : if(next == '>') {
1412 0 : return scanner_token(lc, GEOM_OVERLAP_OR_ABOVE);
1413 : } else {
1414 0 : utf8_putchar(lc, next); //put the char back
1415 0 : utf8_putchar(lc, cur); //put the char back
1416 0 : return scanner_token(lc, '|');
1417 : }
1418 25 : } else if (cur == '>') {
1419 1 : next = scanner_getc(lc);
1420 1 : if (next < 0)
1421 : return EOF;
1422 1 : if(next == '>') {
1423 1 : return scanner_token(lc, GEOM_ABOVE);
1424 : } else {
1425 0 : utf8_putchar(lc, next); //put the char back
1426 0 : utf8_putchar(lc, cur); //put the char back
1427 0 : return scanner_token(lc, '|');
1428 : }
1429 : } else {
1430 24 : utf8_putchar(lc, cur);
1431 24 : return scanner_token(lc, '|');
1432 : }
1433 : }
1434 10 : (void)sql_error( c, 3, SQLSTATE(42000) "Unexpected symbol (%lc)", (wint_t) cur);
1435 10 : return LEX_ERROR;
1436 : }
1437 :
1438 : static int
1439 29293182 : tokenize(mvc * c, int cur)
1440 : {
1441 29293182 : struct scanner *lc = &c->scanner;
1442 58415254 : while (1) {
1443 43854218 : if (cur == 0xFEFF) {
1444 : /* on Linux at least, iswpunct returns TRUE
1445 : * for U+FEFF, but we don't want that, we just
1446 : * want to go to the scanner_error case
1447 : * below */
1448 : ;
1449 43854313 : } else if (iswspace(cur)) {
1450 14578015 : if ((cur = skip_white_space(lc)) == EOF)
1451 : return cur;
1452 14561036 : continue; /* try again */
1453 29276298 : } else if (iswdigit(cur)) {
1454 1956441 : return number(c, cur);
1455 27319857 : } else if (iswalpha(cur) || cur == '_') {
1456 13989246 : switch (cur) {
1457 666895 : case 'e': /* string with escapes */
1458 : case 'E':
1459 666895 : if (scanner_read_more(lc, 1) != EOF &&
1460 666895 : lc->rs->buf[lc->rs->pos + lc->yycur] == '\'') {
1461 3915 : return scanner_string(c, scanner_getc(lc), true);
1462 : }
1463 : break;
1464 426736 : case 'x': /* blob */
1465 : case 'X':
1466 : case 'r': /* raw string */
1467 : case 'R':
1468 426736 : if (scanner_read_more(lc, 1) != EOF &&
1469 426736 : lc->rs->buf[lc->rs->pos + lc->yycur] == '\'') {
1470 3293 : return scanner_string(c, scanner_getc(lc), false);
1471 : }
1472 : break;
1473 162136 : case 'u': /* unicode string */
1474 : case 'U':
1475 162136 : if (scanner_read_more(lc, 1) != EOF &&
1476 162153 : lc->rs->buf[lc->rs->pos + lc->yycur] == '&' &&
1477 17 : scanner_read_more(lc, 2) != EOF &&
1478 17 : (lc->rs->buf[lc->rs->pos + lc->yycur + 1] == '\'' ||
1479 : lc->rs->buf[lc->rs->pos + lc->yycur + 1] == '"')) {
1480 17 : cur = scanner_getc(lc); /* '&' */
1481 17 : return scanner_string(c, scanner_getc(lc), false);
1482 : }
1483 : break;
1484 : default:
1485 : break;
1486 : }
1487 14019076 : return keyword_or_ident(c, cur);
1488 13293556 : } else if (iswpunct(cur)) {
1489 13293429 : return scanner_symbol(c, cur);
1490 : }
1491 32 : if (cur == EOF) {
1492 0 : if (lc->mode == LINE_1 || !lc->started )
1493 : return cur;
1494 0 : return scanner_error(c, cur);
1495 : }
1496 : /* none of the above: error */
1497 32 : return scanner_error(c, cur);
1498 : }
1499 : }
1500 :
1501 : /* SQL 'quoted' idents consist of a set of any character of
1502 : * the source language character set other than a 'quote'
1503 : *
1504 : * MonetDB has 3 restrictions:
1505 : * 1 we disallow '%' as the first character.
1506 : * 2 the length is limited to 1024 characters
1507 : * 3 the identifier 'TID%' is not allowed
1508 : */
1509 : static bool
1510 1330140 : valid_ident(const char *restrict s, char *restrict dst)
1511 : {
1512 1330140 : int p = 0;
1513 :
1514 1330140 : if (*s == '%')
1515 : return false;
1516 :
1517 9844622 : while (*s) {
1518 8514482 : if ((dst[p++] = *s++) == '"' && *s == '"')
1519 68 : s++;
1520 8514482 : if (p >= 1024)
1521 : return false;
1522 : }
1523 1330140 : dst[p] = '\0';
1524 1330140 : if (strcmp(dst, TID + 1) == 0) /* an index named 'TID%' could interfere with '%TID%' */
1525 : return false;
1526 : return true;
1527 : }
1528 :
1529 : static inline int
1530 29688961 : sql_get_next_token(YYSTYPE *yylval, void *parm)
1531 : {
1532 29688961 : mvc *c = (mvc*)parm;
1533 29688961 : struct scanner *lc = &c->scanner;
1534 29688961 : int token = 0, cur = 0;
1535 :
1536 29688961 : if (lc->rs->buf == NULL) /* malloc failure */
1537 : return EOF;
1538 :
1539 29688961 : if (lc->yynext) {
1540 358073 : int next = lc->yynext;
1541 :
1542 358073 : lc->yynext = 0;
1543 358073 : return(next);
1544 : }
1545 :
1546 29330888 : if (lc->yybak) {
1547 28252524 : lc->rs->buf[lc->rs->pos + lc->yycur] = lc->yybak;
1548 28252524 : lc->yybak = 0;
1549 : }
1550 :
1551 29330888 : lc->yysval = lc->yycur;
1552 29330888 : lc->yylast = lc->yyval;
1553 29330888 : cur = scanner_getc(lc);
1554 29330414 : if (cur < 0)
1555 : return EOF;
1556 29219233 : token = tokenize(c, cur);
1557 :
1558 29220356 : yylval->sval = (lc->rs->buf + lc->rs->pos + lc->yysval);
1559 :
1560 29220356 : if (token == KW_TYPE)
1561 49974 : token = aTYPE;
1562 :
1563 29220356 : if (token == IDENT || token == COMPARISON ||
1564 23678880 : token == RANK || token == aTYPE || token == MARGFUNC) {
1565 5602089 : yylval->sval = sa_strndup(c->sa, yylval->sval, lc->yycur-lc->yysval);
1566 5602096 : lc->next_string_is_raw = false;
1567 23618267 : } else if (token == STRING) {
1568 2160751 : char quote = *yylval->sval;
1569 2160751 : char *str = sa_alloc( c->sa, (lc->yycur-lc->yysval-2)*2 + 1 );
1570 2160751 : char *dst;
1571 :
1572 2160751 : assert(quote == '"' || quote == '\'' || quote == 'E' || quote == 'e' || quote == 'U' || quote == 'u' || quote == 'X' || quote == 'x' || quote == 'R' || quote == 'r');
1573 :
1574 2160751 : lc->rs->buf[lc->rs->pos + lc->yycur - 1] = 0;
1575 2160751 : switch (quote) {
1576 1330140 : case '"':
1577 1330140 : if (valid_ident(yylval->sval+1,str)) {
1578 : token = IDENT;
1579 : } else {
1580 0 : sql_error(c, 1, SQLSTATE(42000) "Invalid identifier '%s'", yylval->sval+1);
1581 0 : return LEX_ERROR;
1582 : }
1583 : break;
1584 3914 : case 'e':
1585 : case 'E':
1586 3914 : assert(yylval->sval[1] == '\'');
1587 3914 : if (GDKstrFromStr((unsigned char *) str,
1588 : (unsigned char *) yylval->sval + 2,
1589 3914 : lc->yycur-lc->yysval - 2, '\'') < 0) {
1590 1 : char *err = GDKerrbuf;
1591 1 : if (strncmp(err, GDKERROR, strlen(GDKERROR)) == 0)
1592 1 : err += strlen(GDKERROR);
1593 0 : else if (*err == '!')
1594 0 : err++;
1595 1 : sql_error(c, 1, SQLSTATE(42000) "%s", err);
1596 1 : return LEX_ERROR;
1597 : }
1598 : quote = '\'';
1599 : break;
1600 17 : case 'u':
1601 : case 'U':
1602 17 : assert(yylval->sval[1] == '&');
1603 17 : assert(yylval->sval[2] == '\'' || yylval->sval[2] == '"');
1604 17 : strcpy(str, yylval->sval + 3);
1605 17 : token = yylval->sval[2] == '\'' ? -STRING : -IDENT; /* Passing unicode string/ident as - numbers, handled
1606 : later in scanner */
1607 17 : quote = yylval->sval[2];
1608 17 : lc->next_string_is_raw = true;
1609 17 : break;
1610 1 : case 'x':
1611 : case 'X':
1612 1 : assert(yylval->sval[1] == '\'');
1613 1 : dst = str;
1614 5 : for (char *src = yylval->sval + 2; *src; dst++)
1615 4 : if ((*dst = *src++) == '\'' && *src == '\'')
1616 0 : src++;
1617 1 : *dst = 0;
1618 1 : quote = '\'';
1619 1 : token = XSTRING;
1620 1 : lc->next_string_is_raw = true;
1621 1 : break;
1622 3285 : case 'r':
1623 : case 'R':
1624 3285 : assert(yylval->sval[1] == '\'');
1625 3285 : dst = str;
1626 450421 : for (char *src = yylval->sval + 2; *src; dst++)
1627 447136 : if ((*dst = *src++) == '\'' && *src == '\'')
1628 2780 : src++;
1629 3285 : quote = '\'';
1630 3285 : *dst = 0;
1631 3285 : break;
1632 823394 : default:
1633 823394 : if (lc->raw_string_mode || lc->next_string_is_raw) {
1634 50 : dst = str;
1635 479 : for (char *src = yylval->sval + 1; *src; dst++)
1636 429 : if ((*dst = *src++) == '\'' && *src == '\'')
1637 3 : src++;
1638 50 : *dst = 0;
1639 : } else {
1640 823344 : if (GDKstrFromStr((unsigned char *)str,
1641 823344 : (unsigned char *)yylval->sval + 1,
1642 823344 : lc->yycur - lc->yysval - 1,
1643 : '\'') < 0) {
1644 1 : sql_error(c, 1, SQLSTATE(42000) "%s", GDKerrbuf);
1645 1 : return LEX_ERROR;
1646 : }
1647 : }
1648 : break;
1649 : }
1650 2160749 : yylval->sval = str;
1651 :
1652 : /* reset original */
1653 2160749 : lc->rs->buf[lc->rs->pos+lc->yycur- 1] = quote;
1654 : } else {
1655 21457516 : lc->next_string_is_raw = false;
1656 : }
1657 :
1658 : return(token);
1659 : }
1660 :
1661 : static int scanner( YYSTYPE *yylval, void *m, bool log);
1662 :
1663 : static int
1664 29184037 : scanner(YYSTYPE * yylval, void *parm, bool log)
1665 : {
1666 29184037 : int token;
1667 29184037 : mvc *c = (mvc *) parm;
1668 29184037 : struct scanner *lc = &c->scanner;
1669 29184037 : size_t pos;
1670 29184037 : int last = lc->yyval;
1671 :
1672 : /* store position for when view's query ends */
1673 29184037 : pos = lc->rs->pos + lc->yycur;
1674 :
1675 29184037 : token = sql_get_next_token(yylval, parm);
1676 : /* TODO make hash out of the possible complex tokens and add with current tokens hash */
1677 :
1678 29184970 : if (token == -IDENT || token == -STRING) {
1679 17 : char *sval = yylval->sval;
1680 17 : int next = sql_get_next_token(yylval, parm);
1681 :
1682 17 : if (token == -STRING && next == STRING) {
1683 2 : sval = sa_strconcat(c->sa, sval, yylval->sval);
1684 6 : while((next = sql_get_next_token(yylval, parm)) == STRING)
1685 4 : sval = sa_strconcat(c->sa, sval, yylval->sval);
1686 : }
1687 :
1688 17 : char *uescape = "\\";
1689 17 : if (next == UESCAPE) {
1690 14 : int nxt = sql_get_next_token(yylval, parm);
1691 14 : if (nxt == STRING) {
1692 14 : next = 0;
1693 14 : uescape = yylval->sval;
1694 14 : if (strlen(uescape) != 1 || strchr("\"'0123456789abcdefABCDEF+ \t\n\r\f", *uescape) != NULL) {
1695 0 : sqlformaterror(c, SQLSTATE(22019) "%s", "UESCAPE must be one character");
1696 0 : return LEX_ERROR;
1697 : }
1698 : } else {
1699 0 : sqlformaterror(c, SQLSTATE(22019) "%s", "UESCAPE character missing");
1700 0 : return LEX_ERROR;
1701 : }
1702 : }
1703 17 : yylval->sval = uescape_xform(sval, uescape);
1704 17 : if (yylval->sval == NULL && token == -STRING) {
1705 0 : sqlformaterror(c, SQLSTATE(22019) "%s", "Bad Unicode string");
1706 0 : return LEX_ERROR;
1707 : }
1708 :
1709 17 : if (next)
1710 3 : lc->yynext = next;
1711 17 : return (token == -IDENT)?IDENT:STRING;
1712 29184953 : } else if (token == WITH) { /* check for TIME WITH ... */
1713 15572 : int next = sql_get_next_token(yylval, parm);
1714 15572 : if (next == TIME)
1715 5416 : token = WITH_LA;
1716 15572 : lc->yynext = next;
1717 29169381 : } else if (token == INTO) { /* check for INTO followed by STRING / (BIG/LITTLE/NATIVE) for copy into file vs copy select into var */
1718 139811 : int next = sql_get_next_token(yylval, parm);
1719 139810 : if (next == STRING || next == BIG || next == LITTLE || next == NATIVE ||
1720 : next == BINARY || next == STDOUT)
1721 77 : token = INTO_LA;
1722 139810 : lc->yynext = next;
1723 29029570 : } else if (last == ODBC_FUNC_ESCAPE_PREFIX && token == TIMESTAMPADD) {
1724 : token = ODBC_TIMESTAMPADD;
1725 29029552 : } else if (last == ODBC_FUNC_ESCAPE_PREFIX && token == TIMESTAMPDIFF) {
1726 : token = ODBC_TIMESTAMPDIFF;
1727 29029464 : } else if (last == INTERVAL && (token == '-' || token == '+')) { /* backward compatibility: INTERVAL +- 'string' -> interval '+-string'*/
1728 12 : int next = sql_get_next_token(yylval, parm);
1729 12 : if (next == STRING) {
1730 12 : if (token != '+') {
1731 8 : char *sval = yylval->sval;
1732 8 : if (sval[0] == '+')
1733 0 : sval[0] = '-';
1734 8 : else if (sval[0] == '-')
1735 1 : yylval->sval++;
1736 : else
1737 7 : yylval->sval = sa_strconcat(c->sa, token=='-'?"-":"+", sval);
1738 : }
1739 : token = next;
1740 : next = 0;
1741 : }
1742 12 : lc->yynext = next;
1743 29029452 : } else if (token == OUTER) { /* check for OUTER UNION */
1744 18426 : int next = sql_get_next_token(yylval, parm);
1745 18426 : if (next == UNION)
1746 : token = OUTER_UNION;
1747 : else
1748 18417 : lc->yynext = next;
1749 29011026 : } else if (token == TO) { /* check for end_field (of interval spec) TO (MONTH etc) */
1750 121026 : int next = sql_get_next_token(yylval, parm);
1751 121026 : if (next == YEAR || next == MONTH || next == DAY || next == HOUR || next == MINUTE || next == SECOND)
1752 273 : token = TO_LA;
1753 121026 : lc->yynext = next;
1754 28890000 : } else if (token == NOT) {
1755 77712 : int next = sql_get_next_token(yylval, parm);
1756 :
1757 77712 : if (next == NOT) {
1758 2 : return scanner(yylval, parm, false);
1759 : } else if (next == EXISTS) {
1760 : token = NOT_EXISTS;
1761 : } else if (next == BETWEEN) {
1762 : token = NOT_BETWEEN;
1763 : } else if (next == sqlIN) {
1764 : token = NOT_IN;
1765 : } else if (next == LIKE) {
1766 : token = NOT_LIKE;
1767 : } else if (next == ILIKE) {
1768 : token = NOT_ILIKE;
1769 : } else {
1770 63245 : lc->yynext = next;
1771 : }
1772 28812288 : } else if (token == SCOLON) {
1773 : /* ignore semi-colon(s) following a semi-colon */
1774 1043613 : if (lc->yylast == SCOLON) {
1775 132017 : size_t prev = lc->yycur;
1776 132018 : while ((token = sql_get_next_token(yylval, parm)) == SCOLON)
1777 1 : prev = lc->yycur;
1778 :
1779 : /* skip the skipped stuff also in the buffer */
1780 132017 : lc->rs->pos += prev;
1781 132017 : lc->yycur -= prev;
1782 : }
1783 : }
1784 :
1785 29184950 : if (lc->log && log)
1786 0 : mnstr_write(lc->log, lc->rs->buf+pos, lc->rs->pos + lc->yycur - pos, 1);
1787 :
1788 29184950 : lc->started += (token != EOF);
1789 29184950 : return token;
1790 : }
1791 :
1792 : /* also see sql_parser.y */
1793 : extern int sqllex(YYSTYPE * yylval, void *parm);
1794 :
1795 : int
1796 29184472 : sqllex(YYSTYPE * yylval, void *parm)
1797 : {
1798 29184472 : return scanner(yylval, parm, true);
1799 : }
|
