1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057
|
//
// mysql_parserAppDelegate.m
// mysql.parser
//
// Created by Mike on 03.04.12.
// Copyright 2012, 2016, Oracle Corporation. All rights reserved.
//
#import "mysql_parserAppDelegate.h"
#include <sstream>
#include <vector>
#include <stdlib.h>
static std::string last_error;
extern "C" {
void onMySQLParseError(struct ANTLR3_BASE_RECOGNIZER_struct * recognizer, pANTLR3_UINT8 * tokenNames)
{
std::ostringstream error;
pANTLR3_PARSER parser;
pANTLR3_TREE_PARSER tparser;
pANTLR3_LEXER lexer;
pANTLR3_INT_STREAM is;
pANTLR3_STRING ttext;
pANTLR3_STRING ftext;
pANTLR3_EXCEPTION ex;
pANTLR3_COMMON_TOKEN theToken;
pANTLR3_BASE_TREE theBaseTree;
pANTLR3_COMMON_TREE theCommonTree;
// Retrieve some info for easy reading.
//
ex = recognizer->state->exception;
ttext = NULL;
// See if there is a 'filename' we can use
//
if (ex->streamName == NULL)
{
if (((pANTLR3_COMMON_TOKEN)(ex->token))->type == ANTLR3_TOKEN_EOF)
error << "-end of input-(";
else
error << "-unknown source-(";
}
else
{
ftext = ex->streamName->to8(ex->streamName);
error << ftext->chars << "(";
}
// Next comes the line number.
error << recognizer->state->exception->line << ") ";
// How we determine the next piece is dependent on which thing raised the error.
switch (recognizer->type)
{
case ANTLR3_TYPE_LEXER:
{
lexer = (pANTLR3_LEXER)(recognizer->super);
parser = NULL;
tparser = NULL;
error << " : lexer error " << recognizer->state->exception->type << " : " << (pANTLR3_UINT8)(ex->message);
error << ", at offset " << recognizer->state->exception->charPositionInLine + 1;
{
ANTLR3_INT32 width;
width = ANTLR3_UINT32_CAST(( (pANTLR3_UINT8)(lexer->input->data) + (lexer->input->size(lexer->input) )) - (pANTLR3_UINT8)(ex->index));
if (width >= 1)
{
if (isprint(ex->c))
error << " near '" << (char)ex->c << "' :\n";
else
error << " near char(" << (ANTLR3_UINT8)(ex->c) << ") :\n";
}
else
{
error << "(end of input).\n";
width = ANTLR3_UINT32_CAST(((pANTLR3_UINT8)(lexer->input->data)+(lexer->input->size(lexer->input))) - (pANTLR3_UINT8)(lexer->rec->state->tokenStartCharIndex));
}
}
break;
}
case ANTLR3_TYPE_PARSER:
lexer = NULL;
parser = (pANTLR3_PARSER) (recognizer->super);
tparser = NULL;
is = parser->tstream->istream;
theToken = (pANTLR3_COMMON_TOKEN)(recognizer->state->exception->token);
ttext = theToken->toString(theToken);
error << " : parser error " << recognizer->state->exception->type << " : " << (pANTLR3_UINT8)(ex->message);
error << ", at offset " << recognizer->state->exception->charPositionInLine;
if (theToken != NULL)
{
if (theToken->type == ANTLR3_TOKEN_EOF)
error << ", at <EOF>";
else
error << "\n near " << (ttext == NULL ? (pANTLR3_UINT8)"<no text for the token>" : ttext->chars) << "\n ";
}
break;
case ANTLR3_TYPE_TREE_PARSER:
lexer = NULL;
parser = NULL;
tparser = (pANTLR3_TREE_PARSER) (recognizer->super);
is = tparser->ctnstream->tnstream->istream;
theBaseTree = (pANTLR3_BASE_TREE)(recognizer->state->exception->token);
ttext = theBaseTree->toStringTree(theBaseTree);
if (theBaseTree != NULL)
{
theCommonTree = (pANTLR3_COMMON_TREE) theBaseTree->super;
if (theCommonTree != NULL)
theToken = (pANTLR3_COMMON_TOKEN)theBaseTree->getToken(theBaseTree);
error << " : tree parser error " << recognizer->state->exception->type << " : " << (pANTLR3_UINT8)(recognizer->state->exception->message);
error << ", at offset " << theBaseTree->getCharPositionInLine(theBaseTree) << ", near " << ttext->chars;
}
break;
default:
error << "Internal error: unknown recognizer type appeared in error reporting.\n";
return;
break;
}
// Although this function should generally be provided by the implementation, this one
// should be as helpful as possible for grammar developers and serve as an example
// of what you can do with each exception type. In general, when you make up your
// 'real' handler, you should debug the routine with all possible errors you expect
// which will then let you be as specific as possible about all circumstances.
//
// Note that in the general case, errors thrown by tree parsers indicate a problem
// with the output of the parser or with the tree grammar itself. The job of the parser
// is to produce a perfect (in traversal terms) syntactically correct tree, so errors
// at that stage should really be semantic errors that your own code determines and handles
// in whatever way is appropriate.
switch (ex->type)
{
case ANTLR3_UNWANTED_TOKEN_EXCEPTION:
// Indicates that the recognizer was fed a token which seesm to be
// spurious input. We can detect this when the token that follows
// this unwanted token would normally be part of the syntactically
// correct stream. Then we can see that the token we are looking at
// is just something that should not be there and throw this exception.
//
if (tokenNames == NULL)
error << " : Extraneous input...";
else
{
if (ex->expecting == ANTLR3_TOKEN_EOF)
error << " : Extraneous input - expected <EOF>\n";
else
error << " : Extraneous input - expected " << tokenNames[ex->expecting] << "...\n";
}
break;
case ANTLR3_MISSING_TOKEN_EXCEPTION:
// Indicates that the recognizer detected that the token we just
// hit would be valid syntactically if preceeded by a particular
// token. Perhaps a missing ';' at line end or a missing ',' in an
// expression list, and such like.
//
if (tokenNames == NULL)
{
error << " : Missing token (" << ex->expecting << ")...\n";
}
else
{
if (ex->expecting == ANTLR3_TOKEN_EOF)
error << " : Missing <EOF>\n";
else
error << " : Missing " << tokenNames[ex->expecting] << " \n";
}
break;
case ANTLR3_RECOGNITION_EXCEPTION:
// Indicates that the recognizer received a token
// in the input that was not predicted. This is the basic exception type
// from which all others are derived. So we assume it was a syntax error.
// You may get this if there are not more tokens and more are needed
// to complete a parse for instance.
error << " : syntax error...\n";
break;
case ANTLR3_MISMATCHED_TOKEN_EXCEPTION:
// We were expecting to see one thing and got another. This is the
// most common error if we coudl not detect a missing or unwanted token.
// Here you can spend your efforts to
// derive more useful error messages based on the expected
// token set and the last token and so on. The error following
// bitmaps do a good job of reducing the set that we were looking
// for down to something small. Knowing what you are parsing may be
// able to allow you to be even more specific about an error.
if (tokenNames == NULL)
error << " : syntax error...\n";
else
{
if (ex->expecting == ANTLR3_TOKEN_EOF)
error << " : expected <EOF>\n";
else
error << " : expected " << tokenNames[ex->expecting] << " ...\n";
}
break;
case ANTLR3_NO_VIABLE_ALT_EXCEPTION:
error << " unexpected input\n";
break;
case ANTLR3_MISMATCHED_SET_EXCEPTION:
{
ANTLR3_UINT32 count;
ANTLR3_UINT32 bit;
ANTLR3_UINT32 size;
ANTLR3_UINT32 numbits;
pANTLR3_BITSET errBits;
// This means we were able to deal with one of a set of
// possible tokens at this point, but we did not see any
// member of that set.
error << " : unexpected input...\n expected one of : ";
// What tokens could we have accepted at this point in the parse?
count = 0;
errBits = antlr3BitsetLoad(ex->expectingSet);
numbits = errBits->numBits(errBits);
size = errBits->size(errBits);
if (size > 0)
{
// However many tokens we could have dealt with here, it is usually
// not useful to print ALL of the set here. I arbitrarily chose 8
// here, but you should do whatever makes sense for you of course.
// No token number 0, so look for bit 1 and on.
for (bit = 1; bit < numbits && count < 8 && count < size; bit++)
{
// TODO: This doesn;t look right - should be asking if the bit is set!!
if (tokenNames[bit])
{
error << (count > 0 ? ", " : "") << tokenNames[bit];
count++;
}
}
error << "\n";
}
else
{
error << "Actually dude, we didn't seem to be expecting anything here, or at least\n";
error << "I could not work out what I was expecting, like so many of us these days!\n";
}
}
break;
case ANTLR3_EARLY_EXIT_EXCEPTION:
// We entered a loop requiring a number of token sequences
// but found a token that ended that sequence earlier than
// we should have done.
error << " : missing elements...\n";
break;
default:
// We don't handle any other exceptions here, but you can
// if you wish. If we get an exception that hits this point
// then we are just going to report what we know about the
// token.
error << " : syntax not recognized...\n";
break;
}
last_error += error.str();
}
// Here we just fake some charsets, as we don't have the entire parser context.
std::string charsets[] = {"utf8", "ucs2", "big5", "latin2", "ujis", "binary", "cp1250", "latin1"};
/**
* Checks the given identifier whether it is an actually defined charset, which directs
* the lexer to either classify it apropriately.
*/
ANTLR3_UINT32 check_charset(void *payload, pANTLR3_STRING text)
{
// Get the actual token text and skip the initial underscore char.
// There's an additional char at the end of the input for some reason (maybe a lexer bug),
// so we also ignore the last char.
std::string token_text((const char*)text->chars + 1, text->len - 2);
std::transform(token_text.begin(), token_text.end(), token_text.begin(), ::tolower);
for (int i = 0; i < sizeof(charsets) / sizeof(charsets[0]); i++)
if (charsets[i] == token_text)
return UNDERSCORE_CHARSET;
return IDENTIFIER;
}
}
NSString *sql1 = @"select 2 as expected, /*!01000/**/*/ 2 as result";
NSString *sql2 = @"select (select\n t1.id as a, sakila.actor.actor_id b, t2.id c, "
"(select 1 * 0.123, a from t3) from `ÄÖÜ丈` t1, sakila.actor as t2\n"
"where ((t1.id = t2.id)) and (t1.id = sakila.actor.actor_id)) as r1, 2";
NSString *sql3 = @"select 1 from\n\t{OJ d left outer join e on a = b} left outer join ee on aa = bb,\n\ta t1 force index for "
"order by (a, b),\n\t(select 2 from zz) yy straight_join (select 3) as xx on ww,\n\t(b, c),\n\t(f, g) inner join h using (aa, bb),\n"
"\t(h) straight_join `schema`.`table` on yy < zz natural right join ({OJ i left outer join j on ii = jj})";
NSString *sql4 = @"CREATE PROCEDURE do_insert(value INT)\nBEGIN\n -- declare variables to hold diagnostics area information\n"
" DECLARE code CHAR(5) DEFAULT '00000';\n DECLARE msg TEXT;\n DECLARE rows INT;\n DECLARE result TEXT;\n"
" -- declare exception handler for failed insert\n DECLARE CONTINUE HANDLER FOR SQLEXCEPTION\n BEGIN\n"
" GET DIAGNOSTICS CONDITION 1\n code = RETURNED_SQLSTATE, msg = MESSAGE_TEXT;\n END;\n\n"
" -- perform the insert\n INSERT INTO t1 (int_col) VALUES(value);\n -- check whether the insert was successful\n"
" IF code = '00000' THEN\n GET DIAGNOSTICS rows = ROW_COUNT;\n -- SET result = CONCAT('insert succeeded, row count = ',rows);\n"
" ELSE\n -- SET result = CONCAT('insert failed, error = ',code,', message = ',msg);\n END IF;\n"
" -- say what happened\n SELECT result;\nEND";
NSString *sql5 = @"grant alter (a, b) on table * to 'mike'@'%'";
NSString *sql6 = @"select A, B, A OR B, A XOR B, A AND B from t1_30237_bool where C is null order by A, B";
NSString *sql7 = @"select count(distinct a.actor_id), phone, first_name, a.last_name, country.country \n"
"from sakila.actor a, address aa, country\nwhere (a.actor_id = 0 and country_id > 0) \ngroup by actor_id";
NSString *sql8 = @"drop user current_user(), 'mike'@localhost";
NSString *sql9 = @"CREATE definer = `root`@`localhost` trigger `upd_film` AFTER UPDATE ON `film` FOR EACH ROW BEGIN\n"
" IF (old.title != new.title) or (old.description != new.description)\n"
" THEN\n"
" UPDATE film_text\n"
" SET title=new.title,\n"
" description=new.description,\n"
" film_id=new.film_id\n"
" WHERE film_id=old.film_id;\n"
" END IF;\n"
"END";
NSString *sql10 = @"CREATE TABLE total (\n"
" a INT NOT NULL AUTO_INCREMENT,\n"
" message CHAR(20), INDEX(a))\n"
" ENGINE=MERGE UNION=(t1,t2) INSERT_METHOD=LAST;\n";
NSString *sql11 = @"SELECT a FROM tick t WHERE timestamp > (((((((SELECT 1) + 1))))))";
NSString *sql12 = @"select * from (select 1 from dual)";
NSString *sql13 = @"ALTER USER u@localhost IDENTIFIED WITH sha256_password BY 'test';";
@implementation mysql_parserAppDelegate
@synthesize window;
- (void)applicationDidFinishLaunching:(NSNotification *)aNotification {
// Make the SQL edit control scroll horizontally too.
[singleQueryText.textContainer setContainerSize: NSMakeSize(FLT_MAX, FLT_MAX)];
[singleQueryText.textContainer setWidthTracksTextView: NO];
singleQueryText.string = sql13;
NSUserDefaults *defaults = NSUserDefaults.standardUserDefaults;
NSString *text = [defaults objectForKey: @"statements-file"];
if (text != nil)
pathEdit.stringValue = text;
text = [defaults objectForKey: @"single-query"];
if (text != nil)
singleQueryText.string = text;
}
- (NSString*)dumpTree: (pANTLR3_BASE_TREE)tree state: (pANTLR3_RECOGNIZER_SHARED_STATE)state indentation: (NSString*)indentation
{
NSString *result;
ANTLR3_UINT32 char_pos = tree->getCharPositionInLine(tree);
ANTLR3_UINT32 line = tree->getLine(tree);
pANTLR3_STRING token_text = tree->getText(tree);
pANTLR3_COMMON_TOKEN token = tree->getToken(tree);
NSString *utf8 = [NSString stringWithUTF8String: (const char*)token_text->chars];
if (token != NULL) {
ANTLR3_UINT32 token_type = token->getType(token);
pANTLR3_UINT8 token_name;
if (token_type != ANTLR3_TOKEN_EOF)
token_name = state->tokenNames[token_type];
if (token_type == ANTLR3_TOKEN_EOF || token_name == (pANTLR3_UINT8)ANTLR3_TOKEN_EOF)
token_name = (pANTLR3_UINT8)"ANTLR3_TOKEN_EOF";
result = [NSString stringWithFormat: @"%@(line: %i, offset: %i, length: %lli, index: %lli, %s [%i]) %@\n",
indentation, line, char_pos, token->stop - token->start + 1, token->index, token_name, token_type, utf8];
} else {
result = [NSString stringWithFormat: @"%@(line: %i, offset: %i, nil) %@\n", indentation, line, char_pos, utf8];
}
for (ANTLR3_UINT32 index = 0; index < tree->getChildCount(tree); index++)
{
pANTLR3_BASE_TREE child = (pANTLR3_BASE_TREE)tree->getChild(tree, index);
NSString *child_text = [self dumpTree: child state: state indentation: [indentation stringByAppendingString: @"\t"]];
result = [result stringByAppendingString: child_text];
}
return result;
}
- (NSString*)dumpTokens: (pANTLR3_BASE_TREE)tree state: (pANTLR3_RECOGNIZER_SHARED_STATE)state
{
NSString *result;
ANTLR3_UINT32 char_pos = tree->getCharPositionInLine(tree);
ANTLR3_UINT32 line = tree->getLine(tree);
pANTLR3_STRING token_text = tree->getText(tree);
pANTLR3_COMMON_TOKEN token = tree->getToken(tree);
if (token != NULL) {
ANTLR3_UINT32 token_type = token->getType(token);
pANTLR3_UINT8 token_name;
if (token_type != ANTLR3_TOKEN_EOF)
token_name = state->tokenNames[token_type];
if (token_type == ANTLR3_TOKEN_EOF || token_name == (pANTLR3_UINT8)ANTLR3_TOKEN_EOF)
token_name = (pANTLR3_UINT8)"ANTLR3_TOKEN_EOF";
result = [NSString stringWithFormat: @"(%s) ", token_name];
} else {
result = @"";
}
for (ANTLR3_UINT32 index = 0; index < tree->getChildCount(tree); index++)
{
pANTLR3_BASE_TREE child = (pANTLR3_BASE_TREE)tree->getChild(tree, index);
NSString *child_text = [self dumpTokens: child state: state];
result = [result stringByAppendingString: child_text];
}
return result;
}
- (unsigned)getSqlModes
{
unsigned result = 0;
if (modeIgnoreSpaceButton.state == NSOnState)
result |= SQL_MODE_IGNORE_SPACE;
if (modeAnsiQuotesButton.state == NSOnState)
result |= SQL_MODE_ANSI_QUOTES;
if (modePipesAsConcatButton.state == NSOnState)
result |= SQL_MODE_PIPES_AS_CONCAT;
if (modeHighNotPrecedenceButton.state == NSOnState)
result |= SQL_MODE_HIGH_NOT_PRECEDENCE;
if (modeNoBackslashEscapeButton.state == NSOnState)
result |= SQL_MODE_NO_BACKSLASH_ESCAPES;
return result;
}
- (unsigned)getServerVersion
{
NSString *t = versionText.stringValue;
if (t.length == 0)
return 50630;
return t.intValue;
}
- (IBAction)parseFromClipboard: (id)sender
{
[singleQueryText selectAll: sender];
[singleQueryText pasteAsPlainText: sender];
[self parse: sender];
}
- (IBAction)parse: (id)sender
{
NSUserDefaults *defaults = NSUserDefaults.standardUserDefaults;
[defaults setObject: singleQueryText.string forKey: @"single-query"];
last_error = "";
[errorText setString: @""];
pANTLR3_INPUT_STREAM input;
pMySQLLexer lexer;
pANTLR3_COMMON_TOKEN_STREAM tokens;
pMySQLParser parser;
RecognitionContext context = { 0, 0, [self getServerVersion], [self getSqlModes], nil };
NSString *sql = singleQueryText.string;
std::string utf8 = [sql UTF8String];
input = antlr3StringStreamNew((pANTLR3_UINT8)utf8.c_str(), ANTLR3_ENC_UTF8, utf8.size(), (pANTLR3_UINT8)"sql-script");
input->setUcaseLA(input, ANTLR3_TRUE); // Make input case-insensitive. String literals must all be upper case in the grammar!
lexer = MySQLLexerNew(input);
lexer->pLexer->rec->state->userp = &context;
tokens = antlr3CommonTokenStreamSourceNew(ANTLR3_SIZE_HINT, TOKENSOURCE(lexer));
parser = MySQLParserNew(tokens);
parser->pParser->rec->state->userp = &context;
pANTLR3_BASE_TREE tree = parser->query(parser).tree;
NSString *combinedErrorText = @"No errors found";
ANTLR3_UINT32 error_count = parser->pParser->rec->state->errorCount;
error_count += lexer->pLexer->rec->state->errorCount;
if (error_count > 0) {
combinedErrorText = [NSString stringWithFormat: @"%i errors found\n%s", error_count, last_error.c_str()];
}
[errorText setString: combinedErrorText];
if (tree != NULL) {
NSString *token_text = [self dumpTree: tree state: parser->pParser->rec->state indentation: @""];
[output setString: token_text];
} else {
[output setString: @"no tree"];
}
// Must manually clean up.
parser->free(parser);
tokens ->free(tokens);
lexer->free(lexer);
input->close(input);
}
- (BOOL)parseQuery: (NSString *)query version: (unsigned)serverVersion modes: (unsigned)sqlModes
{
pANTLR3_INPUT_STREAM input;
pMySQLLexer lexer;
pANTLR3_COMMON_TOKEN_STREAM tokens;
pMySQLParser parser;
RecognitionContext context = { 0, 0, serverVersion, sqlModes, nil };
std::string utf8 = [query UTF8String];
input = antlr3StringStreamNew((pANTLR3_UINT8)utf8.c_str(), ANTLR3_ENC_UTF8, utf8.size(), (pANTLR3_UINT8)"sql-script");
input->setUcaseLA(input, ANTLR3_TRUE);
lexer = MySQLLexerNew(input);
lexer->pLexer->rec->state->userp = &context;
tokens = antlr3CommonTokenStreamSourceNew(ANTLR3_SIZE_HINT, TOKENSOURCE(lexer));
parser = MySQLParserNew(tokens);
parser->pParser->rec->state->userp = &context;
MySQLParser_query_return ast = parser->query(parser);
BOOL result = last_error.size() == 0;
// Must manually clean up.
parser->free(parser);
tokens ->free(tokens);
lexer->free(lexer);
input->close(input);
return result;
}
bool parse_and_compare(const std::string query, pANTLR3_BASE_TREE tree, unsigned sql_modes, unsigned server_version)
{
pANTLR3_INPUT_STREAM input;
pMySQLLexer lexer;
pANTLR3_COMMON_TOKEN_STREAM tokens;
pMySQLParser parser;
RecognitionContext context = { 0, 0, server_version, sql_modes, nil };
input = antlr3StringStreamNew((pANTLR3_UINT8)query.c_str(), ANTLR3_ENC_UTF8, query.size(), (pANTLR3_UINT8)"sql-script");
input->setUcaseLA(input, ANTLR3_TRUE);
lexer = MySQLLexerNew(input);
lexer->pLexer->rec->state->userp = &context;
tokens = antlr3CommonTokenStreamSourceNew(ANTLR3_SIZE_HINT, TOKENSOURCE(lexer));
parser = MySQLParserNew(tokens);
parser->pParser->rec->state->userp = &context;
MySQLParser_query_return ast = parser->query(parser);
BOOL result = parser->pParser->rec->state->errorCount == 0;
if (result)
{
// If there was no error then check the order of tokens.
}
parser->free(parser);
tokens ->free(tokens);
lexer->free(lexer);
input->close(input);
return result;
}
- (IBAction)selectFile: (id)sender
{
NSOpenPanel *panel = [NSOpenPanel openPanel];
panel.canChooseDirectories = NO;
panel.allowsMultipleSelection = NO;
panel.canChooseFiles = YES;
panel.allowedFileTypes = [NSArray arrayWithObjects: @"txt", @"sql", nil];
panel.directoryURL = [NSURL fileURLWithPath: [pathEdit stringValue]];
if ([panel runModal] == NSFileHandlingPanelOKButton) {
pathEdit.stringValue = panel.URL.path;
}
}
- (void)appendStepText: (NSString *)newText
{
stepsList.string = [stepsList.string stringByAppendingString: newText];
}
- (IBAction)startTests: (id)sender
{
if (running) {
stopTests = YES;
return;
}
NSString *fileName = pathEdit.stringValue;
NSError *error = nil;
NSString *contents = [NSString stringWithContentsOfFile: fileName encoding: NSUTF8StringEncoding error: &error];
if (error != nil) {
NSAlert *alert = [NSAlert alertWithError: error];
[alert runModal];
return;
}
NSUserDefaults *defaults = NSUserDefaults.standardUserDefaults;
[defaults setObject: fileName forKey: @"statements-file"];
stopTests = NO;
errorQueryText.string = @"";
last_error = "";
statusText.stringValue = @"Counting queries...";
stepsList.string = @"Counting queries\n";
progressLabel.stringValue = @"0 of 0";
[self countQueriesInFile: fileName];
[self appendStepText: [NSString stringWithFormat: @"Found %lu queries in file\n", (unsigned long)queryCount]];
[topView setNeedsDisplay: YES];
NSThread *thread = [[[NSThread alloc] initWithTarget: self
selector: @selector(runTestsWithData:)
object: contents] autorelease];
[thread setStackSize: 8 * 1024 * 1024];
[thread start];
}
- (void)countQueriesInFile: (NSString*)fileName
{
NSFileHandle * file = [NSFileHandle fileHandleForReadingAtPath: fileName];
NSUInteger chunkSize = 1024 * 1024;
NSData * chunk = [file readDataOfLength: chunkSize];
queryCount = 0;
NSUInteger seenDollars = 0;
while ([chunk length] > 0) {
const char *chars = (const char *)[chunk bytes];
for (int index = 0; index < [chunk length]; ++index) {
switch (chars[index]) {
case '$':
seenDollars++;
break;
case '\n':
if (seenDollars == 2) {
seenDollars = 0;
queryCount++;
}
break;
default:
seenDollars = 0;
break;
}
}
chunk = [file readDataOfLength: chunkSize];
}
}
- (void)runTestsWithData: (NSString*)data
{
try {
// 1. Part: parse example queries from file.
dispatch_async(dispatch_get_main_queue(), ^{
statusText.stringValue = @"Running tests from file...";
[self appendStepText: @"Running tests from file\n"];
});
NSUInteger count = 0;
BOOL gotError = NO;
NSArray *queries = [data componentsSeparatedByString: @"$$\n"];
unsigned serverVersion = [self getServerVersion];
unsigned sqlModes = [self getSqlModes];
for (NSString* query in queries) {
if (stopTests)
break;
if (query.length > 0) {
// The progress label is updated in the worker thread, otherwise it would not show changes in realtime.
progressLabel.stringValue = [NSString stringWithFormat: @"%lu of %lu", ++count, queryCount];
[progressLabel setNeedsDisplay];
if (![self parseQuery: query version: serverVersion modes: sqlModes])
{
dispatch_async(dispatch_get_main_queue(), ^{
statusText.stringValue = @"Error encountered, offending query below:";
errorQueryText.string = query;
});
gotError = YES;
break;
}
}
}
// 2. Part: parse queries with function names for foreign keys.
if (!gotError && !stopTests)
{
dispatch_async(dispatch_get_main_queue(), ^{
[self appendStepText: [NSString stringWithFormat: @"Parsed %lu queries in file successfully.\n", (unsigned long)queryCount]];
[self appendStepText: @"Running function name tests\n"];
statusText.stringValue = @"Running function name tests...";
});
gotError = ![self runFunctionNamesTest];
}
dispatch_async(dispatch_get_main_queue(), ^{
if (stopTests)
statusText.stringValue = @"Execution stopped by user.";
else
if (!gotError)
statusText.stringValue = @"All queries parsed fine. Great!";
});
running = NO;
} catch (std::exception& e) {
running = NO;
statusText.stringValue = [NSString stringWithFormat: @"Parser raised exception: %s", e.what()];
}
}
/**
* Generates queries with many MySQL functions (all?) used in foreign key creation (parser bug #21114).
* Note: this test is currently only partially useful as most of the used functions are not reserved words
* (and hence not contained in the grammar). This might change if the parsing process does an
* explicit differentiation between general identifiers and function names.
* Returns YES if all went ok.
*/
- (BOOL)runFunctionNamesTest
{
static const char* functions[] = {
"acos",
"adddate",
"addtime"
"aes_decrypt",
"aes_encrypt",
"area",
"asbinary",
"asin",
"astext",
"aswkb",
"aswkt",
"atan",
"atan2",
"benchmark",
"bin",
"bit_count",
"bit_length",
"ceil",
"ceiling",
"centroid",
"character_length",
"char_length",
"coercibility",
"compress",
"concat",
"concat_ws",
"connection_id",
"conv",
"convert_tz",
"cos",
"cot",
"crc32",
"crosses",
"datediff",
"date_format",
"dayname",
"dayofmonth",
"dayofweek",
"dayofyear",
"decode",
"degrees",
"des_decrypt",
"des_encrypt",
"dimension",
"disjoint",
"elt",
"encode",
"encrypt",
"endpoint",
"envelope",
"equals",
"exp",
"export_set",
"exteriorring",
"extractvalue",
"find_in_set",
"floor",
"found_rows",
"from_days",
"from_unixtime",
"geomcollfromtext",
"geomcollfromwkb",
"geometrycollectionfromtext",
"geometrycollectionfromwkb",
"geometryfromtext",
"geometryfromwkb",
"geometryn",
"geometrytype",
"geomfromtext",
"geomfromwkb",
"get_lock",
"glength",
"greatest",
"hex",
"ifnull",
"inet_aton",
"inet_ntoa",
"instr",
"interiorringn",
"intersects",
"isclosed",
"isempty",
"isnull",
"issimple",
"is_free_lock",
"is_used_lock",
"last_day",
"last_insert_id",
"lcase",
"least",
"length",
"linefromtext",
"linefromwkb",
"linestringfromtext",
"linestringfromwkb",
"ln",
"load_file",
"locate",
"log",
"log10",
"log2",
"lower",
"lpad",
"ltrim",
"makedate",
"maketime",
"make_set",
"master_pos_wait",
"mbrcontains",
"mbrdisjoint",
"mbrequal",
"mbrintersects",
"mbroverlaps",
"mbrtouches",
"mbrwithin",
"md5",
"mlinefromtext",
"mlinefromwkb",
"monthname",
"mpointfromtext",
"mpointfromwkb",
"mpolyfromtext",
"mpolyfromwkb",
"multilinestringfromtext",
"multilinestringfromwkb",
"multipointfromtext",
"multipointfromwkb",
"multipolygonfromtext",
"multipolygonfromwkb",
"name_const",
"nullif",
"numgeometries",
"numinteriorrings",
"numpoints",
"oct",
"octet_length",
"ord",
"overlaps",
"period_add",
"period_diff",
"pi",
"pointfromtext",
"pointfromwkb",
"pointn",
"polyfromtext",
"polyfromwkb",
"polygonfromtext",
"polygonfromwkb",
"pow",
"power",
"quote",
"radians",
"rand",
"release_lock",
"reverse",
"round",
"row_count",
"rpad",
"rtrim",
"sec_to_time",
"session_user",
"sha",
"sha1",
"sign",
"sin",
"sleep",
"soundex",
"space",
"sqrt",
"srid",
"startpoint",
"strcmp",
"str_to_date",
"subdate",
"substring_index",
"subtime",
"system_user",
"tan",
"timediff",
"time_format",
"time_to_sec",
"touches",
"to_days",
"ucase",
"uncompress",
"uncompressed_length",
"unhex",
"unix_timestamp",
"updatexml",
"upper",
"uuid",
"version",
"weekday",
"weekofyear",
"within",
"x",
"y",
"yearweek"};
const char *query1 = "CREATE TABLE %s(\n"
"col1 int not null,\n"
"col2 int not null,\n"
"col3 varchar(10),\n"
"CONSTRAINT pk PRIMARY KEY (col1, col2)\n"
") ENGINE InnoDB";
const char *query2 = "CREATE TABLE bug21114_child(\n"
"pk int not null,\n"
"fk_col1 int not null,\n"
"fk_col2 int not null,\n"
"fk_col3 int not null,\n"
"fk_col4 int not null,\n"
"CONSTRAINT fk_fct FOREIGN KEY (fk_col1, fk_col2)\n"
"REFERENCES %s(col1, col2),\n"
"CONSTRAINT fk_fct_space FOREIGN KEY (fk_col3, fk_col4)\n"
"REFERENCES %s (col1, col2)\n"
") ENGINE InnoDB";
NSInteger count = sizeof(functions) / sizeof(functions[0]);
unsigned serverVersion = [self getServerVersion];
unsigned sqlModes = [self getSqlModes];
for (size_t i = 0; i < count; i++)
{
if (stopTests)
return NO;
NSString *query = [NSString stringWithFormat: [NSString stringWithUTF8String: query1], functions[i]];
if (query.length > 0) {
progressLabel.stringValue = [NSString stringWithFormat: @"%lu of %li", i + 1, count];
[progressLabel setNeedsDisplay];
if (![self parseQuery: query version: serverVersion modes: sqlModes])
{
statusText.stringValue = @"Error encountered, offending query below:";
errorQueryText.string = query;
return NO;
}
}
query = [NSString stringWithFormat: [NSString stringWithUTF8String: query2], functions[i], functions[i]];
if (query.length > 0) {
[progressLabel setNeedsDisplay];
if (![self parseQuery: query version: serverVersion modes: sqlModes])
{
dispatch_async(dispatch_get_main_queue(), ^{
statusText.stringValue = @"Error encountered, offending query below:";
errorQueryText.string = query;
});
return NO;
}
}
}
return YES;
}
#pragma mark - Action handling
- (IBAction)startConversion: (id)sender
{
NSString *tokenListString = @"";
NSArray *queries = [queryList.string componentsSeparatedByCharactersInSet: [NSCharacterSet characterSetWithCharactersInString: @"$$"]];
for (NSString *entry in queries) {
NSString *query = [entry stringByTrimmingCharactersInSet: [NSCharacterSet whitespaceAndNewlineCharacterSet]];
if (query.length == 0)
continue;
pANTLR3_INPUT_STREAM input;
pMySQLLexer lexer;
pANTLR3_COMMON_TOKEN_STREAM tokenStream;
pMySQLParser parser;
RecognitionContext context = { 0, 0, [self getServerVersion], [self getSqlModes], nil};
std::string utf8 = [query UTF8String];
input = antlr3StringStreamNew((pANTLR3_UINT8)utf8.c_str(), ANTLR3_ENC_UTF8, utf8.size(), (pANTLR3_UINT8)"sql-script");
input->setUcaseLA(input, ANTLR3_TRUE);
lexer = MySQLLexerNew(input);
lexer->pLexer->rec->state->userp = &context;
tokenStream = antlr3CommonTokenStreamSourceNew(ANTLR3_SIZE_HINT, TOKENSOURCE(lexer));
parser = MySQLParserNew(tokenStream);
parser->pParser->rec->state->userp = &context;
MySQLParser_query_return ast = parser->query(parser);
if (parser->pParser->rec->state->errorCount > 0) {
conversionErrorText.stringValue = @"Errors occurred.";
}
NSString *tokens;
if (ast.tree != NULL) {
tokens = [self dumpTokens: ast.tree state: parser->pParser->rec->state];
} else {
tokens = @"(no tree)";
}
tokenListString = [tokenListString stringByAppendingString:
[NSString stringWithFormat: @"list_of %@,\n", tokens]
];
// Must manually clean up.
parser->free(parser);
tokenStream ->free(tokenStream);
lexer->free(lexer);
input->close(input);
}
tokenList.string = tokenListString;
conversionErrorText.stringValue = @"Everything went fine.";
}
@end
|