wparser_def.c 50.8 KB
Newer Older
1 2 3
/*-------------------------------------------------------------------------
 *
 * wparser_def.c
4
 *		Default text search parser
5 6 7 8 9
 *
 * Portions Copyright (c) 1996-2007, PostgreSQL Global Development Group
 *
 *
 * IDENTIFICATION
10
 *	  $PostgreSQL: pgsql/src/backend/tsearch/wparser_def.c,v 1.8 2007/11/09 22:37:35 tgl Exp $
11 12 13 14 15 16
 *
 *-------------------------------------------------------------------------
 */

#include "postgres.h"

17
#include "commands/defrem.h"
18 19 20 21 22 23 24
#include "tsearch/ts_locale.h"
#include "tsearch/ts_public.h"
#include "tsearch/ts_type.h"
#include "tsearch/ts_utils.h"
#include "utils/builtins.h"


25 26 27 28
/* Define me to enable tracing of parser behavior */
/* #define WPARSER_TRACE */


29
/* Output token categories */
30

31 32 33 34 35 36 37
#define ASCIIWORD		1
#define WORD_T			2
#define NUMWORD			3
#define EMAIL			4
#define URL_T			5
#define HOST			6
#define SCIENTIFIC		7
38
#define VERSIONNUMBER	8
39 40 41 42 43
#define NUMPARTHWORD	9
#define PARTHWORD		10
#define ASCIIPARTHWORD	11
#define SPACE			12
#define TAG_T			13
44
#define PROTOCOL		14
45 46 47
#define NUMHWORD		15
#define ASCIIHWORD		16
#define HWORD			17
48
#define URLPATH			18
49 50 51 52 53 54 55 56 57
#define FILEPATH		19
#define DECIMAL			20
#define SIGNEDINT		21
#define UNSIGNEDINT 	22
#define HTMLENTITY		23

#define LASTNUM			23

static const char * const tok_alias[] = {
58
	"",
59
	"asciiword",
60
	"word",
61
	"numword",
62 63 64 65 66
	"email",
	"url",
	"host",
	"sfloat",
	"version",
67 68 69
	"hword_numpart",
	"hword_part",
	"hword_asciipart",
70 71 72
	"blank",
	"tag",
	"protocol",
73 74
	"numhword",
	"asciihword",
75
	"hword",
76
	"url_path",
77 78 79 80 81 82 83
	"file",
	"float",
	"int",
	"uint",
	"entity"
};

84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102
static const char * const lex_descr[] = {
	"",
	"Word, all ASCII",
	"Word, all letters",
	"Word, letters and digits",
	"Email address",
	"URL",
	"Host",
	"Scientific notation",
	"Version number",
	"Hyphenated word part, letters and digits",
	"Hyphenated word part, all letters",
	"Hyphenated word part, all ASCII",
	"Space symbols",
	"HTML tag",
	"Protocol head",
	"Hyphenated word, letters and digits",
	"Hyphenated word, all ASCII",
	"Hyphenated word, all letters",
103
	"URL path",
104 105 106 107 108 109 110 111 112 113
	"File or path name",
	"Decimal notation",
	"Signed integer",
	"Unsigned integer",
	"HTML entity"
};


/* Parser states */

114 115 116
typedef enum
{
	TPS_Base = 0,
117 118 119
	TPS_InNumWord,
	TPS_InAsciiWord,
	TPS_InWord,
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
	TPS_InUnsignedInt,
	TPS_InSignedIntFirst,
	TPS_InSignedInt,
	TPS_InSpace,
	TPS_InUDecimalFirst,
	TPS_InUDecimal,
	TPS_InDecimalFirst,
	TPS_InDecimal,
	TPS_InVerVersion,
	TPS_InSVerVersion,
	TPS_InVersionFirst,
	TPS_InVersion,
	TPS_InMantissaFirst,
	TPS_InMantissaSign,
	TPS_InMantissa,
	TPS_InHTMLEntityFirst,
	TPS_InHTMLEntity,
	TPS_InHTMLEntityNumFirst,
	TPS_InHTMLEntityNum,
	TPS_InHTMLEntityEnd,
	TPS_InTagFirst,
	TPS_InXMLBegin,
	TPS_InTagCloseFirst,
	TPS_InTagName,
	TPS_InTagBeginEnd,
	TPS_InTag,
	TPS_InTagEscapeK,
	TPS_InTagEscapeKK,
	TPS_InTagBackSleshed,
	TPS_InTagEnd,
	TPS_InCommentFirst,
	TPS_InCommentLast,
	TPS_InComment,
	TPS_InCloseCommentFirst,
	TPS_InCloseCommentLast,
	TPS_InCommentEnd,
	TPS_InHostFirstDomain,
	TPS_InHostDomainSecond,
	TPS_InHostDomain,
	TPS_InPortFirst,
	TPS_InPort,
	TPS_InHostFirstAN,
	TPS_InHost,
	TPS_InEmail,
	TPS_InFileFirst,
	TPS_InFileTwiddle,
	TPS_InPathFirst,
	TPS_InPathFirstFirst,
	TPS_InPathSecond,
	TPS_InFile,
	TPS_InFileNext,
171 172 173
	TPS_InURLPathFirst,
	TPS_InURLPathStart,
	TPS_InURLPath,
174 175 176 177
	TPS_InFURL,
	TPS_InProtocolFirst,
	TPS_InProtocolSecond,
	TPS_InProtocolEnd,
178 179 180 181 182 183
	TPS_InHyphenAsciiWordFirst,
	TPS_InHyphenAsciiWord,
	TPS_InHyphenWordFirst,
	TPS_InHyphenWord,
	TPS_InHyphenNumWordFirst,
	TPS_InHyphenNumWord,
184
	TPS_InHyphenDigitLookahead,
185 186
	TPS_InParseHyphen,
	TPS_InParseHyphenHyphen,
187 188 189
	TPS_InHyphenWordPart,
	TPS_InHyphenAsciiWordPart,
	TPS_InHyphenNumWordPart,
190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211
	TPS_InHyphenUnsignedInt,
	TPS_Null					/* last state (fake value) */
} TParserState;

/* forward declaration */
struct TParser;

typedef int (*TParserCharTest) (struct TParser *);		/* any p_is* functions
														 * except p_iseq */
typedef void (*TParserSpecial) (struct TParser *);		/* special handler for
														 * special cases... */

typedef struct
{
	TParserCharTest isclass;
	char		c;
	uint16		flags;
	TParserState tostate;
	int			type;
	TParserSpecial special;
} TParserStateActionItem;

212 213 214 215 216 217 218 219 220 221
/* Flag bits in TParserStateActionItem.flags */
#define A_NEXT		0x0000
#define A_BINGO		0x0001
#define A_POP		0x0002
#define A_PUSH		0x0004
#define A_RERUN		0x0008
#define A_CLEAR		0x0010
#define A_MERGE		0x0020
#define A_CLRALL	0x0040

222 223 224
typedef struct TParserPosition
{
	int			posbyte;		/* position of parser in bytes */
225
	int			poschar;		/* position of parser in characters */
226
	int			charlen;		/* length of current char */
227 228
	int			lenbytetoken;	/* length of token-so-far in bytes */
	int			lenchartoken;	/* and in chars */
229 230
	TParserState state;
	struct TParserPosition *prev;
231
	const TParserStateActionItem *pushedAtAction;
232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254
} TParserPosition;

typedef struct TParser
{
	/* string and position information */
	char	   *str;			/* multibyte string */
	int			lenstr;			/* length of mbstring */
#ifdef TS_USE_WIDE
	wchar_t    *wstr;			/* wide character string */
	int			lenwstr;		/* length of wsting */
#endif

	/* State of parse */
	int			charmaxlen;
	bool		usewide;
	TParserPosition *state;
	bool		ignore;
	bool		wanthost;

	/* silly char */
	char		c;

	/* out */
255 256 257
	char	   *token;
	int			lenbytetoken;
	int			lenchartoken;
258 259 260
	int			type;
} TParser;

261 262 263 264 265

/* forward decls here */
static bool TParserGet(TParser * prs);


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
static TParserPosition *
newTParserPosition(TParserPosition * prev)
{
	TParserPosition *res = (TParserPosition *) palloc(sizeof(TParserPosition));

	if (prev)
		memcpy(res, prev, sizeof(TParserPosition));
	else
		memset(res, 0, sizeof(TParserPosition));

	res->prev = prev;

	res->pushedAtAction = NULL;

	return res;
}

static TParser *
TParserInit(char *str, int len)
{
	TParser    *prs = (TParser *) palloc0(sizeof(TParser));

	prs->charmaxlen = pg_database_encoding_max_length();
	prs->str = str;
	prs->lenstr = len;

#ifdef TS_USE_WIDE

	/*
	 * Use wide char code only when max encoding length > 1.
	 */
	if (prs->charmaxlen > 1)
	{
		prs->usewide = true;
		prs->wstr = (wchar_t *) palloc(sizeof(wchar_t) * (prs->lenstr + 1));
301 302
		prs->lenwstr = char2wchar(prs->wstr, prs->lenstr + 1,
								  prs->str, prs->lenstr);
303 304 305 306 307 308 309 310
	}
	else
#endif
		prs->usewide = false;

	prs->state = newTParserPosition(NULL);
	prs->state->state = TPS_Base;

311 312 313 314
#ifdef WPARSER_TRACE
	fprintf(stderr, "parsing \"%.*s\"\n", len, str);
#endif

315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337
	return prs;
}

static void
TParserClose(TParser * prs)
{
	while (prs->state)
	{
		TParserPosition *ptr = prs->state->prev;

		pfree(prs->state);
		prs->state = ptr;
	}

#ifdef TS_USE_WIDE
	if (prs->wstr)
		pfree(prs->wstr);
#endif

	pfree(prs);
}

/*
338
 * Character-type support functions, equivalent to is* macros, but
339 340
 * working with any possible encodings and locales. Note,
 * that with multibyte encoding and C-locale isw* function may fail
341
 * or give wrong result. Note 2: multibyte encoding and C-locale
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
 * often are used for Asian languages
 */

#ifdef TS_USE_WIDE

#define p_iswhat(type)														\
static int																	\
p_is##type(TParser *prs) {													\
	Assert( prs->state );													\
	if ( prs->usewide )														\
	{																		\
		if ( lc_ctype_is_c() )												\
			return is##type( 0xff & *( prs->wstr + prs->state->poschar) );	\
																			\
		return isw##type( *(wint_t*)( prs->wstr + prs->state->poschar ) );	\
	}																		\
																			\
	return is##type( *(unsigned char*)( prs->str + prs->state->posbyte ) ); \
}	\
																			\
static int																	\
p_isnot##type(TParser *prs) {												\
	return !p_is##type(prs);												\
}

static int
p_isalnum(TParser * prs)
{
	Assert(prs->state);

	if (prs->usewide)
	{
		if (lc_ctype_is_c())
		{
			unsigned int c = *(prs->wstr + prs->state->poschar);

			/*
			 * any non-ascii symbol with multibyte encoding with C-locale is
			 * an alpha character
			 */
			if (c > 0x7f)
				return 1;

			return isalnum(0xff & c);
		}

		return iswalnum((wint_t) *(prs->wstr + prs->state->poschar));
	}

	return isalnum(*(unsigned char *) (prs->str + prs->state->posbyte));
}
static int
p_isnotalnum(TParser * prs)
{
	return !p_isalnum(prs);
}

static int
p_isalpha(TParser * prs)
{
	Assert(prs->state);

	if (prs->usewide)
	{
		if (lc_ctype_is_c())
		{
			unsigned int c = *(prs->wstr + prs->state->poschar);

			/*
			 * any non-ascii symbol with multibyte encoding with C-locale is
			 * an alpha character
			 */
			if (c > 0x7f)
				return 1;

			return isalpha(0xff & c);
		}

		return iswalpha((wint_t) *(prs->wstr + prs->state->poschar));
	}

	return isalpha(*(unsigned char *) (prs->str + prs->state->posbyte));
}

static int
p_isnotalpha(TParser * prs)
{
	return !p_isalpha(prs);
}

/* p_iseq should be used only for ascii symbols */

static int
p_iseq(TParser * prs, char c)
{
	Assert(prs->state);
	return ((prs->state->charlen == 1 && *(prs->str + prs->state->posbyte) == c)) ? 1 : 0;
}
#else							/* TS_USE_WIDE */

#define p_iswhat(type)														\
static int																	\
p_is##type(TParser *prs) {													\
	Assert( prs->state );													\
	return is##type( (unsigned char)*( prs->str + prs->state->posbyte ) );	\
}	\
																			\
static int																	\
p_isnot##type(TParser *prs) {												\
	return !p_is##type(prs);												\
}


static int
p_iseq(TParser * prs, char c)
{
	Assert(prs->state);
	return (*(prs->str + prs->state->posbyte) == c) ? 1 : 0;
}

p_iswhat(alnum)
p_iswhat(alpha)
#endif   /* TS_USE_WIDE */

p_iswhat(digit)
p_iswhat(lower)
p_iswhat(print)
p_iswhat(punct)
p_iswhat(space)
p_iswhat(upper)
p_iswhat(xdigit)

static int
p_isEOF(TParser * prs)
{
	Assert(prs->state);
	return (prs->state->posbyte == prs->lenstr || prs->state->charlen == 0) ? 1 : 0;
}

static int
p_iseqC(TParser * prs)
{
	return p_iseq(prs, prs->c);
}

static int
p_isneC(TParser * prs)
{
	return !p_iseq(prs, prs->c);
}

static int
p_isascii(TParser * prs)
{
	return (prs->state->charlen == 1 && isascii((unsigned char) *(prs->str + prs->state->posbyte))) ? 1 : 0;
}

static int
500
p_isasclet(TParser * prs)
501
{
502
	return (p_isascii(prs) && p_isalpha(prs)) ? 1 : 0;
503 504 505
}


506
/* deliberately suppress unused-function complaints for the above */
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
void		_make_compiler_happy(void);
void
_make_compiler_happy(void)
{
	p_isalnum(NULL);
	p_isnotalnum(NULL);
	p_isalpha(NULL);
	p_isnotalpha(NULL);
	p_isdigit(NULL);
	p_isnotdigit(NULL);
	p_islower(NULL);
	p_isnotlower(NULL);
	p_isprint(NULL);
	p_isnotprint(NULL);
	p_ispunct(NULL);
	p_isnotpunct(NULL);
	p_isspace(NULL);
	p_isnotspace(NULL);
	p_isupper(NULL);
	p_isnotupper(NULL);
	p_isxdigit(NULL);
	p_isnotxdigit(NULL);
	p_isEOF(NULL);
	p_iseqC(NULL);
	p_isneC(NULL);
}


static void
SpecialTags(TParser * prs)
{
538
	switch (prs->state->lenchartoken)
539 540
	{
		case 8:			/* </script */
541
			if (pg_strncasecmp(prs->token, "</script", 8) == 0)
542 543 544
				prs->ignore = false;
			break;
		case 7:			/* <script || </style */
545
			if (pg_strncasecmp(prs->token, "</style", 7) == 0)
546
				prs->ignore = false;
547
			else if (pg_strncasecmp(prs->token, "<script", 7) == 0)
548 549 550
				prs->ignore = true;
			break;
		case 6:			/* <style */
551
			if (pg_strncasecmp(prs->token, "<style", 6) == 0)
552 553 554 555 556 557 558 559 560 561 562
				prs->ignore = true;
			break;
		default:
			break;
	}
}

static void
SpecialFURL(TParser * prs)
{
	prs->wanthost = true;
563 564
	prs->state->posbyte -= prs->state->lenbytetoken;
	prs->state->poschar -= prs->state->lenchartoken;
565 566 567 568 569
}

static void
SpecialHyphen(TParser * prs)
{
570 571
	prs->state->posbyte -= prs->state->lenbytetoken;
	prs->state->poschar -= prs->state->lenchartoken;
572 573 574 575 576
}

static void
SpecialVerVersion(TParser * prs)
{
577 578 579 580
	prs->state->posbyte -= prs->state->lenbytetoken;
	prs->state->poschar -= prs->state->lenchartoken;
	prs->state->lenbytetoken = 0;
	prs->state->lenchartoken = 0;
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
}

static int
p_isstophost(TParser * prs)
{
	if (prs->wanthost)
	{
		prs->wanthost = false;
		return 1;
	}
	return 0;
}

static int
p_isignore(TParser * prs)
{
	return (prs->ignore) ? 1 : 0;
}

static int
p_ishost(TParser * prs)
{
	TParser    *tmpprs = TParserInit(prs->str + prs->state->posbyte, prs->lenstr - prs->state->posbyte);
	int			res = 0;

	if (TParserGet(tmpprs) && tmpprs->type == HOST)
	{
608 609 610 611
		prs->state->posbyte += tmpprs->lenbytetoken;
		prs->state->poschar += tmpprs->lenchartoken;
		prs->state->lenbytetoken += tmpprs->lenbytetoken;
		prs->state->lenchartoken += tmpprs->lenchartoken;
612 613 614 615 616 617 618 619 620
		prs->state->charlen = tmpprs->state->charlen;
		res = 1;
	}
	TParserClose(tmpprs);

	return res;
}

static int
621
p_isURLPath(TParser * prs)
622 623 624 625 626 627 628
{
	TParser    *tmpprs = TParserInit(prs->str + prs->state->posbyte, prs->lenstr - prs->state->posbyte);
	int			res = 0;

	tmpprs->state = newTParserPosition(tmpprs->state);
	tmpprs->state->state = TPS_InFileFirst;

629
	if (TParserGet(tmpprs) && (tmpprs->type == URLPATH || tmpprs->type == FILEPATH))
630
	{
631 632 633 634
		prs->state->posbyte += tmpprs->lenbytetoken;
		prs->state->poschar += tmpprs->lenchartoken;
		prs->state->lenbytetoken += tmpprs->lenbytetoken;
		prs->state->lenchartoken += tmpprs->lenchartoken;
635 636 637 638 639 640 641 642 643 644 645 646
		prs->state->charlen = tmpprs->state->charlen;
		res = 1;
	}
	TParserClose(tmpprs);

	return res;
}

/*
 * Table of state/action of parser
 */

647
static const TParserStateActionItem actionTPS_Base[] = {
648 649 650
	{p_isEOF, 0, A_NEXT, TPS_Null, 0, NULL},
	{p_iseqC, '<', A_PUSH, TPS_InTagFirst, 0, NULL},
	{p_isignore, 0, A_NEXT, TPS_InSpace, 0, NULL},
651 652
	{p_isasclet, 0, A_NEXT, TPS_InAsciiWord, 0, NULL},
	{p_isalpha, 0, A_NEXT, TPS_InWord, 0, NULL},
653 654 655 656 657 658 659 660 661 662 663
	{p_isdigit, 0, A_NEXT, TPS_InUnsignedInt, 0, NULL},
	{p_iseqC, '-', A_PUSH, TPS_InSignedIntFirst, 0, NULL},
	{p_iseqC, '+', A_PUSH, TPS_InSignedIntFirst, 0, NULL},
	{p_iseqC, '&', A_PUSH, TPS_InHTMLEntityFirst, 0, NULL},
	{p_iseqC, '~', A_PUSH, TPS_InFileTwiddle, 0, NULL},
	{p_iseqC, '/', A_PUSH, TPS_InFileFirst, 0, NULL},
	{p_iseqC, '.', A_PUSH, TPS_InPathFirstFirst, 0, NULL},
	{NULL, 0, A_NEXT, TPS_InSpace, 0, NULL}
};


664
static const TParserStateActionItem actionTPS_InNumWord[] = {
665 666
	{p_isEOF, 0, A_BINGO, TPS_Base, NUMWORD, NULL},
	{p_isalnum, 0, A_NEXT, TPS_InNumWord, 0, NULL},
667 668 669
	{p_iseqC, '@', A_PUSH, TPS_InEmail, 0, NULL},
	{p_iseqC, '/', A_PUSH, TPS_InFileFirst, 0, NULL},
	{p_iseqC, '.', A_PUSH, TPS_InFileNext, 0, NULL},
670 671
	{p_iseqC, '-', A_PUSH, TPS_InHyphenNumWordFirst, 0, NULL},
	{NULL, 0, A_BINGO, TPS_Base, NUMWORD, NULL}
672 673
};

674
static const TParserStateActionItem actionTPS_InAsciiWord[] = {
675 676
	{p_isEOF, 0, A_BINGO, TPS_Base, ASCIIWORD, NULL},
	{p_isasclet, 0, A_NEXT, TPS_Null, 0, NULL},
677 678 679
	{p_iseqC, '.', A_PUSH, TPS_InHostFirstDomain, 0, NULL},
	{p_iseqC, '.', A_PUSH, TPS_InFileNext, 0, NULL},
	{p_iseqC, '-', A_PUSH, TPS_InHostFirstAN, 0, NULL},
680
	{p_iseqC, '-', A_PUSH, TPS_InHyphenAsciiWordFirst, 0, NULL},
681 682 683 684
	{p_iseqC, '@', A_PUSH, TPS_InEmail, 0, NULL},
	{p_iseqC, ':', A_PUSH, TPS_InProtocolFirst, 0, NULL},
	{p_iseqC, '/', A_PUSH, TPS_InFileFirst, 0, NULL},
	{p_isdigit, 0, A_PUSH, TPS_InHost, 0, NULL},
685 686 687
	{p_isdigit, 0, A_NEXT, TPS_InNumWord, 0, NULL},
	{p_isalpha, 0, A_NEXT, TPS_InWord, 0, NULL},
	{NULL, 0, A_BINGO, TPS_Base, ASCIIWORD, NULL}
688 689
};

690
static const TParserStateActionItem actionTPS_InWord[] = {
691 692 693 694 695
	{p_isEOF, 0, A_BINGO, TPS_Base, WORD_T, NULL},
	{p_isalpha, 0, A_NEXT, TPS_Null, 0, NULL},
	{p_isdigit, 0, A_NEXT, TPS_InNumWord, 0, NULL},
	{p_iseqC, '-', A_PUSH, TPS_InHyphenWordFirst, 0, NULL},
	{NULL, 0, A_BINGO, TPS_Base, WORD_T, NULL}
696 697
};

698
static const TParserStateActionItem actionTPS_InUnsignedInt[] = {
699 700 701 702 703 704
	{p_isEOF, 0, A_BINGO, TPS_Base, UNSIGNEDINT, NULL},
	{p_isdigit, 0, A_NEXT, TPS_Null, 0, NULL},
	{p_iseqC, '.', A_PUSH, TPS_InHostFirstDomain, 0, NULL},
	{p_iseqC, '.', A_PUSH, TPS_InUDecimalFirst, 0, NULL},
	{p_iseqC, 'e', A_PUSH, TPS_InMantissaFirst, 0, NULL},
	{p_iseqC, 'E', A_PUSH, TPS_InMantissaFirst, 0, NULL},
705 706
	{p_isasclet, 0, A_PUSH, TPS_InHost, 0, NULL},
	{p_isalpha, 0, A_NEXT, TPS_InNumWord, 0, NULL},
707 708 709 710
	{p_iseqC, '/', A_PUSH, TPS_InFileFirst, 0, NULL},
	{NULL, 0, A_BINGO, TPS_Base, UNSIGNEDINT, NULL}
};

711
static const TParserStateActionItem actionTPS_InSignedIntFirst[] = {
712 713 714 715 716
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_isdigit, 0, A_NEXT | A_CLEAR, TPS_InSignedInt, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

717
static const TParserStateActionItem actionTPS_InSignedInt[] = {
718 719 720 721 722 723 724 725
	{p_isEOF, 0, A_BINGO, TPS_Base, SIGNEDINT, NULL},
	{p_isdigit, 0, A_NEXT, TPS_Null, 0, NULL},
	{p_iseqC, '.', A_PUSH, TPS_InDecimalFirst, 0, NULL},
	{p_iseqC, 'e', A_PUSH, TPS_InMantissaFirst, 0, NULL},
	{p_iseqC, 'E', A_PUSH, TPS_InMantissaFirst, 0, NULL},
	{NULL, 0, A_BINGO, TPS_Base, SIGNEDINT, NULL}
};

726
static const TParserStateActionItem actionTPS_InSpace[] = {
727 728 729 730 731 732 733 734 735 736 737
	{p_isEOF, 0, A_BINGO, TPS_Base, SPACE, NULL},
	{p_iseqC, '<', A_BINGO, TPS_Base, SPACE, NULL},
	{p_isignore, 0, A_NEXT, TPS_Null, 0, NULL},
	{p_iseqC, '-', A_BINGO, TPS_Base, SPACE, NULL},
	{p_iseqC, '+', A_BINGO, TPS_Base, SPACE, NULL},
	{p_iseqC, '&', A_BINGO, TPS_Base, SPACE, NULL},
	{p_iseqC, '/', A_BINGO, TPS_Base, SPACE, NULL},
	{p_isnotalnum, 0, A_NEXT, TPS_InSpace, 0, NULL},
	{NULL, 0, A_BINGO, TPS_Base, SPACE, NULL}
};

738
static const TParserStateActionItem actionTPS_InUDecimalFirst[] = {
739 740 741 742 743
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_isdigit, 0, A_CLEAR, TPS_InUDecimal, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

744
static const TParserStateActionItem actionTPS_InUDecimal[] = {
745 746 747 748 749 750 751 752
	{p_isEOF, 0, A_BINGO, TPS_Base, DECIMAL, NULL},
	{p_isdigit, 0, A_NEXT, TPS_InUDecimal, 0, NULL},
	{p_iseqC, '.', A_PUSH, TPS_InVersionFirst, 0, NULL},
	{p_iseqC, 'e', A_PUSH, TPS_InMantissaFirst, 0, NULL},
	{p_iseqC, 'E', A_PUSH, TPS_InMantissaFirst, 0, NULL},
	{NULL, 0, A_BINGO, TPS_Base, DECIMAL, NULL}
};

753
static const TParserStateActionItem actionTPS_InDecimalFirst[] = {
754 755 756 757 758
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_isdigit, 0, A_CLEAR, TPS_InDecimal, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

759
static const TParserStateActionItem actionTPS_InDecimal[] = {
760 761 762 763 764 765 766 767
	{p_isEOF, 0, A_BINGO, TPS_Base, DECIMAL, NULL},
	{p_isdigit, 0, A_NEXT, TPS_InDecimal, 0, NULL},
	{p_iseqC, '.', A_PUSH, TPS_InVerVersion, 0, NULL},
	{p_iseqC, 'e', A_PUSH, TPS_InMantissaFirst, 0, NULL},
	{p_iseqC, 'E', A_PUSH, TPS_InMantissaFirst, 0, NULL},
	{NULL, 0, A_BINGO, TPS_Base, DECIMAL, NULL}
};

768
static const TParserStateActionItem actionTPS_InVerVersion[] = {
769 770 771 772 773
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_isdigit, 0, A_RERUN, TPS_InSVerVersion, 0, SpecialVerVersion},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

774
static const TParserStateActionItem actionTPS_InSVerVersion[] = {
775 776 777 778 779 780
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_isdigit, 0, A_BINGO | A_CLRALL, TPS_InUnsignedInt, SPACE, NULL},
	{NULL, 0, A_NEXT, TPS_Null, 0, NULL}
};


781
static const TParserStateActionItem actionTPS_InVersionFirst[] = {
782 783 784 785 786
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_isdigit, 0, A_CLEAR, TPS_InVersion, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

787
static const TParserStateActionItem actionTPS_InVersion[] = {
788 789 790 791 792 793
	{p_isEOF, 0, A_BINGO, TPS_Base, VERSIONNUMBER, NULL},
	{p_isdigit, 0, A_NEXT, TPS_InVersion, 0, NULL},
	{p_iseqC, '.', A_PUSH, TPS_InVersionFirst, 0, NULL},
	{NULL, 0, A_BINGO, TPS_Base, VERSIONNUMBER, NULL}
};

794
static const TParserStateActionItem actionTPS_InMantissaFirst[] = {
795 796 797 798 799 800 801
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_isdigit, 0, A_CLEAR, TPS_InMantissa, 0, NULL},
	{p_iseqC, '+', A_NEXT, TPS_InMantissaSign, 0, NULL},
	{p_iseqC, '-', A_NEXT, TPS_InMantissaSign, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

802
static const TParserStateActionItem actionTPS_InMantissaSign[] = {
803 804 805 806 807
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_isdigit, 0, A_CLEAR, TPS_InMantissa, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

808
static const TParserStateActionItem actionTPS_InMantissa[] = {
809 810 811 812 813
	{p_isEOF, 0, A_BINGO, TPS_Base, SCIENTIFIC, NULL},
	{p_isdigit, 0, A_NEXT, TPS_InMantissa, 0, NULL},
	{NULL, 0, A_BINGO, TPS_Base, SCIENTIFIC, NULL}
};

814
static const TParserStateActionItem actionTPS_InHTMLEntityFirst[] = {
815 816
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_iseqC, '#', A_NEXT, TPS_InHTMLEntityNumFirst, 0, NULL},
817
	{p_isasclet, 0, A_NEXT, TPS_InHTMLEntity, 0, NULL},
818 819 820
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

821
static const TParserStateActionItem actionTPS_InHTMLEntity[] = {
822
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
823
	{p_isasclet, 0, A_NEXT, TPS_InHTMLEntity, 0, NULL},
824 825 826 827
	{p_iseqC, ';', A_NEXT, TPS_InHTMLEntityEnd, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

828
static const TParserStateActionItem actionTPS_InHTMLEntityNumFirst[] = {
829 830 831 832 833
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_isdigit, 0, A_NEXT, TPS_InHTMLEntityNum, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

834
static const TParserStateActionItem actionTPS_InHTMLEntityNum[] = {
835 836 837 838 839 840
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_isdigit, 0, A_NEXT, TPS_InHTMLEntityNum, 0, NULL},
	{p_iseqC, ';', A_NEXT, TPS_InHTMLEntityEnd, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

841
static const TParserStateActionItem actionTPS_InHTMLEntityEnd[] = {
842 843 844
	{NULL, 0, A_BINGO | A_CLEAR, TPS_Base, HTMLENTITY, NULL}
};

845
static const TParserStateActionItem actionTPS_InTagFirst[] = {
846 847 848 849
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_iseqC, '/', A_PUSH, TPS_InTagCloseFirst, 0, NULL},
	{p_iseqC, '!', A_PUSH, TPS_InCommentFirst, 0, NULL},
	{p_iseqC, '?', A_PUSH, TPS_InXMLBegin, 0, NULL},
850
	{p_isasclet, 0, A_PUSH, TPS_InTagName, 0, NULL},
851 852 853
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

854
static const TParserStateActionItem actionTPS_InXMLBegin[] = {
855 856 857 858 859 860 861
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	/* <?xml ... */
	{p_iseqC, 'x', A_NEXT, TPS_InTag, 0, NULL},
	{p_iseqC, 'X', A_NEXT, TPS_InTag, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

862
static const TParserStateActionItem actionTPS_InTagCloseFirst[] = {
863
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
864
	{p_isasclet, 0, A_NEXT, TPS_InTagName, 0, NULL},
865 866 867
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

868
static const TParserStateActionItem actionTPS_InTagName[] = {
869 870 871 872 873
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	/* <br/> case */
	{p_iseqC, '/', A_NEXT, TPS_InTagBeginEnd, 0, NULL},
	{p_iseqC, '>', A_NEXT, TPS_InTagEnd, 0, SpecialTags},
	{p_isspace, 0, A_NEXT, TPS_InTag, 0, SpecialTags},
874
	{p_isasclet, 0, A_NEXT, TPS_Null, 0, NULL},
875 876 877
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

878
static const TParserStateActionItem actionTPS_InTagBeginEnd[] = {
879 880 881 882 883
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_iseqC, '>', A_NEXT, TPS_InTagEnd, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

884
static const TParserStateActionItem actionTPS_InTag[] = {
885 886 887 888
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_iseqC, '>', A_NEXT, TPS_InTagEnd, 0, SpecialTags},
	{p_iseqC, '\'', A_NEXT, TPS_InTagEscapeK, 0, NULL},
	{p_iseqC, '"', A_NEXT, TPS_InTagEscapeKK, 0, NULL},
889
	{p_isasclet, 0, A_NEXT, TPS_Null, 0, NULL},
890 891 892 893 894 895 896 897 898 899 900 901 902 903 904
	{p_isdigit, 0, A_NEXT, TPS_Null, 0, NULL},
	{p_iseqC, '=', A_NEXT, TPS_Null, 0, NULL},
	{p_iseqC, '-', A_NEXT, TPS_Null, 0, NULL},
	{p_iseqC, '#', A_NEXT, TPS_Null, 0, NULL},
	{p_iseqC, '/', A_NEXT, TPS_Null, 0, NULL},
	{p_iseqC, ':', A_NEXT, TPS_Null, 0, NULL},
	{p_iseqC, '.', A_NEXT, TPS_Null, 0, NULL},
	{p_iseqC, '&', A_NEXT, TPS_Null, 0, NULL},
	{p_iseqC, '?', A_NEXT, TPS_Null, 0, NULL},
	{p_iseqC, '%', A_NEXT, TPS_Null, 0, NULL},
	{p_iseqC, '~', A_NEXT, TPS_Null, 0, NULL},
	{p_isspace, 0, A_NEXT, TPS_Null, 0, SpecialTags},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

905
static const TParserStateActionItem actionTPS_InTagEscapeK[] = {
906 907 908 909 910 911
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_iseqC, '\\', A_PUSH, TPS_InTagBackSleshed, 0, NULL},
	{p_iseqC, '\'', A_NEXT, TPS_InTag, 0, NULL},
	{NULL, 0, A_NEXT, TPS_InTagEscapeK, 0, NULL}
};

912
static const TParserStateActionItem actionTPS_InTagEscapeKK[] = {
913 914 915 916 917 918
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_iseqC, '\\', A_PUSH, TPS_InTagBackSleshed, 0, NULL},
	{p_iseqC, '"', A_NEXT, TPS_InTag, 0, NULL},
	{NULL, 0, A_NEXT, TPS_InTagEscapeKK, 0, NULL}
};

919
static const TParserStateActionItem actionTPS_InTagBackSleshed[] = {
920 921 922 923
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{NULL, 0, A_MERGE, TPS_Null, 0, NULL}
};

924
static const TParserStateActionItem actionTPS_InTagEnd[] = {
925
	{NULL, 0, A_BINGO | A_CLRALL, TPS_Base, TAG_T, NULL}
926 927
};

928
static const TParserStateActionItem actionTPS_InCommentFirst[] = {
929 930 931 932 933 934 935 936
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_iseqC, '-', A_NEXT, TPS_InCommentLast, 0, NULL},
	/* <!DOCTYPE ...> */
	{p_iseqC, 'D', A_NEXT, TPS_InTag, 0, NULL},
	{p_iseqC, 'd', A_NEXT, TPS_InTag, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

937
static const TParserStateActionItem actionTPS_InCommentLast[] = {
938 939 940 941 942
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_iseqC, '-', A_NEXT, TPS_InComment, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

943
static const TParserStateActionItem actionTPS_InComment[] = {
944 945 946 947 948
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_iseqC, '-', A_NEXT, TPS_InCloseCommentFirst, 0, NULL},
	{NULL, 0, A_NEXT, TPS_Null, 0, NULL}
};

949
static const TParserStateActionItem actionTPS_InCloseCommentFirst[] = {
950 951 952 953 954
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_iseqC, '-', A_NEXT, TPS_InCloseCommentLast, 0, NULL},
	{NULL, 0, A_NEXT, TPS_InComment, 0, NULL}
};

955
static const TParserStateActionItem actionTPS_InCloseCommentLast[] = {
956 957 958 959 960 961
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_iseqC, '-', A_NEXT, TPS_Null, 0, NULL},
	{p_iseqC, '>', A_NEXT, TPS_InCommentEnd, 0, NULL},
	{NULL, 0, A_NEXT, TPS_InComment, 0, NULL}
};

962
static const TParserStateActionItem actionTPS_InCommentEnd[] = {
963
	{NULL, 0, A_BINGO | A_CLRALL, TPS_Base, TAG_T, NULL}
964 965
};

966
static const TParserStateActionItem actionTPS_InHostFirstDomain[] = {
967
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
968
	{p_isasclet, 0, A_NEXT, TPS_InHostDomainSecond, 0, NULL},
969 970 971 972
	{p_isdigit, 0, A_NEXT, TPS_InHost, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

973
static const TParserStateActionItem actionTPS_InHostDomainSecond[] = {
974
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
975
	{p_isasclet, 0, A_NEXT, TPS_InHostDomain, 0, NULL},
976 977 978 979 980 981 982
	{p_isdigit, 0, A_PUSH, TPS_InHost, 0, NULL},
	{p_iseqC, '-', A_PUSH, TPS_InHostFirstAN, 0, NULL},
	{p_iseqC, '.', A_PUSH, TPS_InHostFirstDomain, 0, NULL},
	{p_iseqC, '@', A_PUSH, TPS_InEmail, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

983
static const TParserStateActionItem actionTPS_InHostDomain[] = {
984
	{p_isEOF, 0, A_BINGO | A_CLRALL, TPS_Base, HOST, NULL},
985
	{p_isasclet, 0, A_NEXT, TPS_InHostDomain, 0, NULL},
986 987 988 989 990 991
	{p_isdigit, 0, A_PUSH, TPS_InHost, 0, NULL},
	{p_iseqC, ':', A_PUSH, TPS_InPortFirst, 0, NULL},
	{p_iseqC, '-', A_PUSH, TPS_InHostFirstAN, 0, NULL},
	{p_iseqC, '.', A_PUSH, TPS_InHostFirstDomain, 0, NULL},
	{p_iseqC, '@', A_PUSH, TPS_InEmail, 0, NULL},
	{p_isdigit, 0, A_POP, TPS_Null, 0, NULL},
992
	{p_isstophost, 0, A_BINGO | A_CLRALL, TPS_InURLPathStart, HOST, NULL},
993 994 995 996
	{p_iseqC, '/', A_PUSH, TPS_InFURL, 0, NULL},
	{NULL, 0, A_BINGO | A_CLRALL, TPS_Base, HOST, NULL}
};

997
static const TParserStateActionItem actionTPS_InPortFirst[] = {
998 999 1000 1001 1002
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_isdigit, 0, A_NEXT, TPS_InPort, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

1003
static const TParserStateActionItem actionTPS_InPort[] = {
1004 1005
	{p_isEOF, 0, A_BINGO | A_CLRALL, TPS_Base, HOST, NULL},
	{p_isdigit, 0, A_NEXT, TPS_InPort, 0, NULL},
1006
	{p_isstophost, 0, A_BINGO | A_CLRALL, TPS_InURLPathStart, HOST, NULL},
1007 1008 1009 1010
	{p_iseqC, '/', A_PUSH, TPS_InFURL, 0, NULL},
	{NULL, 0, A_BINGO | A_CLRALL, TPS_Base, HOST, NULL}
};

1011
static const TParserStateActionItem actionTPS_InHostFirstAN[] = {
1012 1013
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_isdigit, 0, A_NEXT, TPS_InHost, 0, NULL},
1014
	{p_isasclet, 0, A_NEXT, TPS_InHost, 0, NULL},
1015 1016 1017
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

1018
static const TParserStateActionItem actionTPS_InHost[] = {
1019 1020
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_isdigit, 0, A_NEXT, TPS_InHost, 0, NULL},
1021
	{p_isasclet, 0, A_NEXT, TPS_InHost, 0, NULL},
1022 1023 1024 1025 1026 1027
	{p_iseqC, '@', A_PUSH, TPS_InEmail, 0, NULL},
	{p_iseqC, '.', A_PUSH, TPS_InHostFirstDomain, 0, NULL},
	{p_iseqC, '-', A_PUSH, TPS_InHostFirstAN, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

1028
static const TParserStateActionItem actionTPS_InEmail[] = {
1029 1030 1031 1032
	{p_ishost, 0, A_BINGO | A_CLRALL, TPS_Base, EMAIL, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

1033
static const TParserStateActionItem actionTPS_InFileFirst[] = {
1034
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
1035
	{p_isasclet, 0, A_NEXT, TPS_InFile, 0, NULL},
1036 1037 1038
	{p_isdigit, 0, A_NEXT, TPS_InFile, 0, NULL},
	{p_iseqC, '.', A_NEXT, TPS_InPathFirst, 0, NULL},
	{p_iseqC, '_', A_NEXT, TPS_InFile, 0, NULL},
1039
	{p_iseqC, '?', A_PUSH, TPS_InURLPathFirst, 0, NULL},
1040 1041 1042 1043
	{p_iseqC, '~', A_PUSH, TPS_InFileTwiddle, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

1044
static const TParserStateActionItem actionTPS_InFileTwiddle[] = {
1045
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
1046
	{p_isasclet, 0, A_NEXT, TPS_InFile, 0, NULL},
1047 1048 1049 1050 1051 1052
	{p_isdigit, 0, A_NEXT, TPS_InFile, 0, NULL},
	{p_iseqC, '_', A_NEXT, TPS_InFile, 0, NULL},
	{p_iseqC, '/', A_NEXT, TPS_InFileFirst, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

1053
static const TParserStateActionItem actionTPS_InPathFirst[] = {
1054
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
1055
	{p_isasclet, 0, A_NEXT, TPS_InFile, 0, NULL},
1056 1057 1058 1059 1060 1061 1062
	{p_isdigit, 0, A_NEXT, TPS_InFile, 0, NULL},
	{p_iseqC, '_', A_NEXT, TPS_InFile, 0, NULL},
	{p_iseqC, '.', A_NEXT, TPS_InPathSecond, 0, NULL},
	{p_iseqC, '/', A_NEXT, TPS_InFileFirst, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

1063
static const TParserStateActionItem actionTPS_InPathFirstFirst[] = {
1064 1065 1066 1067 1068 1069
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_iseqC, '.', A_NEXT, TPS_InPathSecond, 0, NULL},
	{p_iseqC, '/', A_NEXT, TPS_InFileFirst, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

1070
static const TParserStateActionItem actionTPS_InPathSecond[] = {
1071 1072 1073 1074 1075 1076 1077
	{p_isEOF, 0, A_BINGO | A_CLEAR, TPS_Base, FILEPATH, NULL},
	{p_iseqC, '/', A_NEXT | A_PUSH, TPS_InFileFirst, 0, NULL},
	{p_iseqC, '/', A_BINGO | A_CLEAR, TPS_Base, FILEPATH, NULL},
	{p_isspace, 0, A_BINGO | A_CLEAR, TPS_Base, FILEPATH, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

1078
static const TParserStateActionItem actionTPS_InFile[] = {
1079
	{p_isEOF, 0, A_BINGO, TPS_Base, FILEPATH, NULL},
1080
	{p_isasclet, 0, A_NEXT, TPS_InFile, 0, NULL},
1081 1082 1083 1084 1085
	{p_isdigit, 0, A_NEXT, TPS_InFile, 0, NULL},
	{p_iseqC, '.', A_PUSH, TPS_InFileNext, 0, NULL},
	{p_iseqC, '_', A_NEXT, TPS_InFile, 0, NULL},
	{p_iseqC, '-', A_NEXT, TPS_InFile, 0, NULL},
	{p_iseqC, '/', A_PUSH, TPS_InFileFirst, 0, NULL},
1086
	{p_iseqC, '?', A_PUSH, TPS_InURLPathFirst, 0, NULL},
1087 1088 1089
	{NULL, 0, A_BINGO, TPS_Base, FILEPATH, NULL}
};

1090
static const TParserStateActionItem actionTPS_InFileNext[] = {
1091
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
1092
	{p_isasclet, 0, A_CLEAR, TPS_InFile, 0, NULL},
1093 1094 1095 1096 1097
	{p_isdigit, 0, A_CLEAR, TPS_InFile, 0, NULL},
	{p_iseqC, '_', A_CLEAR, TPS_InFile, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

1098
static const TParserStateActionItem actionTPS_InURLPathFirst[] = {
1099 1100 1101
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_iseqC, '"', A_POP, TPS_Null, 0, NULL},
	{p_iseqC, '\'', A_POP, TPS_Null, 0, NULL},
1102
	{p_isnotspace, 0, A_CLEAR, TPS_InURLPath, 0, NULL},
1103 1104 1105
	{NULL, 0, A_POP, TPS_Null, 0, NULL},
};

1106
static const TParserStateActionItem actionTPS_InURLPathStart[] = {
1107
	{NULL, 0, A_NEXT, TPS_InURLPath, 0, NULL}
1108 1109
};

1110
static const TParserStateActionItem actionTPS_InURLPath[] = {
1111 1112 1113 1114 1115
	{p_isEOF, 0, A_BINGO, TPS_Base, URLPATH, NULL},
	{p_iseqC, '"', A_BINGO, TPS_Base, URLPATH, NULL},
	{p_iseqC, '\'', A_BINGO, TPS_Base, URLPATH, NULL},
	{p_isnotspace, 0, A_NEXT, TPS_InURLPath, 0, NULL},
	{NULL, 0, A_BINGO, TPS_Base, URLPATH, NULL}
1116 1117
};

1118
static const TParserStateActionItem actionTPS_InFURL[] = {
1119
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
1120
	{p_isURLPath, 0, A_BINGO | A_CLRALL, TPS_Base, URL_T, SpecialFURL},
1121 1122 1123
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

1124
static const TParserStateActionItem actionTPS_InProtocolFirst[] = {
1125 1126 1127 1128 1129
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_iseqC, '/', A_NEXT, TPS_InProtocolSecond, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

1130
static const TParserStateActionItem actionTPS_InProtocolSecond[] = {
1131 1132 1133 1134 1135
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_iseqC, '/', A_NEXT, TPS_InProtocolEnd, 0, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

1136
static const TParserStateActionItem actionTPS_InProtocolEnd[] = {
1137 1138 1139
	{NULL, 0, A_BINGO | A_CLRALL, TPS_Base, PROTOCOL, NULL}
};

1140
static const TParserStateActionItem actionTPS_InHyphenAsciiWordFirst[] = {
1141
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
1142 1143
	{p_isasclet, 0, A_NEXT, TPS_InHyphenAsciiWord, 0, NULL},
	{p_isalpha, 0, A_NEXT, TPS_InHyphenWord, 0, NULL},
1144
	{p_isdigit, 0, A_NEXT, TPS_InHyphenDigitLookahead, 0, NULL},
1145 1146 1147
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

1148
static const TParserStateActionItem actionTPS_InHyphenAsciiWord[] = {
1149 1150 1151 1152 1153 1154
	{p_isEOF, 0, A_BINGO | A_CLRALL, TPS_InParseHyphen, ASCIIHWORD, SpecialHyphen},
	{p_isasclet, 0, A_NEXT, TPS_InHyphenAsciiWord, 0, NULL},
	{p_isalpha, 0, A_NEXT, TPS_InHyphenWord, 0, NULL},
	{p_isdigit, 0, A_NEXT, TPS_InHyphenNumWord, 0, NULL},
	{p_iseqC, '-', A_PUSH, TPS_InHyphenAsciiWordFirst, 0, NULL},
	{NULL, 0, A_BINGO | A_CLRALL, TPS_InParseHyphen, ASCIIHWORD, SpecialHyphen}
1155 1156
};

1157
static const TParserStateActionItem actionTPS_InHyphenWordFirst[] = {
1158
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
1159
	{p_isalpha, 0, A_NEXT, TPS_InHyphenWord, 0, NULL},
1160
	{p_isdigit, 0, A_NEXT, TPS_InHyphenDigitLookahead, 0, NULL},
1161 1162 1163
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

1164
static const TParserStateActionItem actionTPS_InHyphenWord[] = {
1165 1166 1167 1168 1169
	{p_isEOF, 0, A_BINGO | A_CLRALL, TPS_InParseHyphen, HWORD, SpecialHyphen},
	{p_isalpha, 0, A_NEXT, TPS_InHyphenWord, 0, NULL},
	{p_isdigit, 0, A_NEXT, TPS_InHyphenNumWord, 0, NULL},
	{p_iseqC, '-', A_PUSH, TPS_InHyphenWordFirst, 0, NULL},
	{NULL, 0, A_BINGO | A_CLRALL, TPS_InParseHyphen, HWORD, SpecialHyphen}
1170 1171
};

1172
static const TParserStateActionItem actionTPS_InHyphenNumWordFirst[] = {
1173
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
1174
	{p_isalpha, 0, A_NEXT, TPS_InHyphenNumWord, 0, NULL},
1175
	{p_isdigit, 0, A_NEXT, TPS_InHyphenDigitLookahead, 0, NULL},
1176 1177 1178
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

1179
static const TParserStateActionItem actionTPS_InHyphenNumWord[] = {
1180 1181 1182 1183
	{p_isEOF, 0, A_BINGO | A_CLRALL, TPS_InParseHyphen, NUMHWORD, SpecialHyphen},
	{p_isalnum, 0, A_NEXT, TPS_InHyphenNumWord, 0, NULL},
	{p_iseqC, '-', A_PUSH, TPS_InHyphenNumWordFirst, 0, NULL},
	{NULL, 0, A_BINGO | A_CLRALL, TPS_InParseHyphen, NUMHWORD, SpecialHyphen}
1184 1185
};

1186
static const TParserStateActionItem actionTPS_InHyphenDigitLookahead[] = {
1187
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
1188
	{p_isdigit, 0, A_NEXT, TPS_InHyphenDigitLookahead, 0, NULL},
1189
	{p_isalpha, 0, A_NEXT, TPS_InHyphenNumWord, 0, NULL},
1190
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
1191 1192
};

1193
static const TParserStateActionItem actionTPS_InParseHyphen[] = {
1194
	{p_isEOF, 0, A_RERUN, TPS_Base, 0, NULL},
1195 1196
	{p_isasclet, 0, A_NEXT, TPS_InHyphenAsciiWordPart, 0, NULL},
	{p_isalpha, 0, A_NEXT, TPS_InHyphenWordPart, 0, NULL},
1197
	{p_isdigit, 0, A_PUSH, TPS_InHyphenUnsignedInt, 0, NULL},
1198 1199 1200 1201
	{p_iseqC, '-', A_PUSH, TPS_InParseHyphenHyphen, 0, NULL},
	{NULL, 0, A_RERUN, TPS_Base, 0, NULL}
};

1202
static const TParserStateActionItem actionTPS_InParseHyphenHyphen[] = {
1203 1204 1205 1206 1207
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
	{p_isalnum, 0, A_BINGO | A_CLEAR, TPS_InParseHyphen, SPACE, NULL},
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

1208
static const TParserStateActionItem actionTPS_InHyphenWordPart[] = {
1209 1210 1211 1212
	{p_isEOF, 0, A_BINGO, TPS_Base, PARTHWORD, NULL},
	{p_isalpha, 0, A_NEXT, TPS_InHyphenWordPart, 0, NULL},
	{p_isdigit, 0, A_NEXT, TPS_InHyphenNumWordPart, 0, NULL},
	{NULL, 0, A_BINGO, TPS_InParseHyphen, PARTHWORD, NULL}
1213 1214
};

1215
static const TParserStateActionItem actionTPS_InHyphenAsciiWordPart[] = {
1216 1217 1218 1219 1220
	{p_isEOF, 0, A_BINGO, TPS_Base, ASCIIPARTHWORD, NULL},
	{p_isasclet, 0, A_NEXT, TPS_InHyphenAsciiWordPart, 0, NULL},
	{p_isalpha, 0, A_NEXT, TPS_InHyphenWordPart, 0, NULL},
	{p_isdigit, 0, A_NEXT, TPS_InHyphenNumWordPart, 0, NULL},
	{NULL, 0, A_BINGO, TPS_InParseHyphen, ASCIIPARTHWORD, NULL}
1221 1222
};

1223
static const TParserStateActionItem actionTPS_InHyphenNumWordPart[] = {
1224 1225 1226
	{p_isEOF, 0, A_BINGO, TPS_Base, NUMPARTHWORD, NULL},
	{p_isalnum, 0, A_NEXT, TPS_InHyphenNumWordPart, 0, NULL},
	{NULL, 0, A_BINGO, TPS_InParseHyphen, NUMPARTHWORD, NULL}
1227 1228
};

1229
static const TParserStateActionItem actionTPS_InHyphenUnsignedInt[] = {
1230
	{p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
1231 1232
	{p_isdigit, 0, A_NEXT, TPS_Null, 0, NULL},
	{p_isalpha, 0, A_CLEAR, TPS_InHyphenNumWordPart, 0, NULL},
1233 1234 1235
	{NULL, 0, A_POP, TPS_Null, 0, NULL}
};

1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256

/*
 * main table of per-state parser actions
 */
typedef struct
{
	const TParserStateActionItem *action;	/* the actual state info */
	TParserState state;			/* only for Assert crosscheck */
#ifdef WPARSER_TRACE
	const char *state_name;		/* only for debug printout */
#endif
} TParserStateAction;

#ifdef WPARSER_TRACE
#define TPARSERSTATEACTION(state) \
	{ CppConcat(action,state), state, CppAsString(state) }
#else
#define TPARSERSTATEACTION(state) \
	{ CppConcat(action,state), state }
#endif

1257
/*
1258
 * order must be the same as in typedef enum {} TParserState!!
1259 1260 1261
 */

static const TParserStateAction Actions[] = {
1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329
	TPARSERSTATEACTION(TPS_Base),
	TPARSERSTATEACTION(TPS_InNumWord),
	TPARSERSTATEACTION(TPS_InAsciiWord),
	TPARSERSTATEACTION(TPS_InWord),
	TPARSERSTATEACTION(TPS_InUnsignedInt),
	TPARSERSTATEACTION(TPS_InSignedIntFirst),
	TPARSERSTATEACTION(TPS_InSignedInt),
	TPARSERSTATEACTION(TPS_InSpace),
	TPARSERSTATEACTION(TPS_InUDecimalFirst),
	TPARSERSTATEACTION(TPS_InUDecimal),
	TPARSERSTATEACTION(TPS_InDecimalFirst),
	TPARSERSTATEACTION(TPS_InDecimal),
	TPARSERSTATEACTION(TPS_InVerVersion),
	TPARSERSTATEACTION(TPS_InSVerVersion),
	TPARSERSTATEACTION(TPS_InVersionFirst),
	TPARSERSTATEACTION(TPS_InVersion),
	TPARSERSTATEACTION(TPS_InMantissaFirst),
	TPARSERSTATEACTION(TPS_InMantissaSign),
	TPARSERSTATEACTION(TPS_InMantissa),
	TPARSERSTATEACTION(TPS_InHTMLEntityFirst),
	TPARSERSTATEACTION(TPS_InHTMLEntity),
	TPARSERSTATEACTION(TPS_InHTMLEntityNumFirst),
	TPARSERSTATEACTION(TPS_InHTMLEntityNum),
	TPARSERSTATEACTION(TPS_InHTMLEntityEnd),
	TPARSERSTATEACTION(TPS_InTagFirst),
	TPARSERSTATEACTION(TPS_InXMLBegin),
	TPARSERSTATEACTION(TPS_InTagCloseFirst),
	TPARSERSTATEACTION(TPS_InTagName),
	TPARSERSTATEACTION(TPS_InTagBeginEnd),
	TPARSERSTATEACTION(TPS_InTag),
	TPARSERSTATEACTION(TPS_InTagEscapeK),
	TPARSERSTATEACTION(TPS_InTagEscapeKK),
	TPARSERSTATEACTION(TPS_InTagBackSleshed),
	TPARSERSTATEACTION(TPS_InTagEnd),
	TPARSERSTATEACTION(TPS_InCommentFirst),
	TPARSERSTATEACTION(TPS_InCommentLast),
	TPARSERSTATEACTION(TPS_InComment),
	TPARSERSTATEACTION(TPS_InCloseCommentFirst),
	TPARSERSTATEACTION(TPS_InCloseCommentLast),
	TPARSERSTATEACTION(TPS_InCommentEnd),
	TPARSERSTATEACTION(TPS_InHostFirstDomain),
	TPARSERSTATEACTION(TPS_InHostDomainSecond),
	TPARSERSTATEACTION(TPS_InHostDomain),
	TPARSERSTATEACTION(TPS_InPortFirst),
	TPARSERSTATEACTION(TPS_InPort),
	TPARSERSTATEACTION(TPS_InHostFirstAN),
	TPARSERSTATEACTION(TPS_InHost),
	TPARSERSTATEACTION(TPS_InEmail),
	TPARSERSTATEACTION(TPS_InFileFirst),
	TPARSERSTATEACTION(TPS_InFileTwiddle),
	TPARSERSTATEACTION(TPS_InPathFirst),
	TPARSERSTATEACTION(TPS_InPathFirstFirst),
	TPARSERSTATEACTION(TPS_InPathSecond),
	TPARSERSTATEACTION(TPS_InFile),
	TPARSERSTATEACTION(TPS_InFileNext),
	TPARSERSTATEACTION(TPS_InURLPathFirst),
	TPARSERSTATEACTION(TPS_InURLPathStart),
	TPARSERSTATEACTION(TPS_InURLPath),
	TPARSERSTATEACTION(TPS_InFURL),
	TPARSERSTATEACTION(TPS_InProtocolFirst),
	TPARSERSTATEACTION(TPS_InProtocolSecond),
	TPARSERSTATEACTION(TPS_InProtocolEnd),
	TPARSERSTATEACTION(TPS_InHyphenAsciiWordFirst),
	TPARSERSTATEACTION(TPS_InHyphenAsciiWord),
	TPARSERSTATEACTION(TPS_InHyphenWordFirst),
	TPARSERSTATEACTION(TPS_InHyphenWord),
	TPARSERSTATEACTION(TPS_InHyphenNumWordFirst),
	TPARSERSTATEACTION(TPS_InHyphenNumWord),
1330
	TPARSERSTATEACTION(TPS_InHyphenDigitLookahead),
1331 1332 1333 1334 1335
	TPARSERSTATEACTION(TPS_InParseHyphen),
	TPARSERSTATEACTION(TPS_InParseHyphenHyphen),
	TPARSERSTATEACTION(TPS_InHyphenWordPart),
	TPARSERSTATEACTION(TPS_InHyphenAsciiWordPart),
	TPARSERSTATEACTION(TPS_InHyphenNumWordPart),
1336
	TPARSERSTATEACTION(TPS_InHyphenUnsignedInt)
1337 1338 1339 1340 1341 1342
};


static bool
TParserGet(TParser * prs)
{
1343
	const TParserStateActionItem *item = NULL;
1344

1345 1346
	Assert(prs->state);

1347 1348 1349
	if (prs->state->posbyte >= prs->lenstr)
		return false;

1350
	prs->token = prs->str + prs->state->posbyte;
1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365
	prs->state->pushedAtAction = NULL;

	/* look at string */
	while (prs->state->posbyte <= prs->lenstr)
	{
		if (prs->state->posbyte == prs->lenstr)
			prs->state->charlen = 0;
		else
			prs->state->charlen = (prs->charmaxlen == 1) ? prs->charmaxlen :
				pg_mblen(prs->str + prs->state->posbyte);

		Assert(prs->state->posbyte + prs->state->charlen <= prs->lenstr);
		Assert(prs->state->state >= TPS_Base && prs->state->state < TPS_Null);
		Assert(Actions[prs->state->state].state == prs->state->state);

1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376
		if (prs->state->pushedAtAction)
		{
			/* After a POP, pick up at the next test */
			item = prs->state->pushedAtAction + 1;
			prs->state->pushedAtAction = NULL;
		}
		else
		{
			item = Actions[prs->state->state].action;
			Assert(item != NULL);
		}
1377 1378 1379 1380 1381 1382

		/* find action by character class */
		while (item->isclass)
		{
			prs->c = item->c;
			if (item->isclass(prs) != 0)
1383
				break;
1384 1385 1386
			item++;
		}

1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414
#ifdef WPARSER_TRACE
		{
			TParserPosition *ptr;

			fprintf(stderr, "state ");
			/* indent according to stack depth */
			for (ptr = prs->state->prev; ptr; ptr = ptr->prev)
				fprintf(stderr, "  ");
			fprintf(stderr, "%s ", Actions[prs->state->state].state_name);
			if (prs->state->posbyte < prs->lenstr)
				fprintf(stderr, "at %c", *(prs->str + prs->state->posbyte));
			else
				fprintf(stderr, "at EOF");
			fprintf(stderr, " matched rule %d flags%s%s%s%s%s%s%s%s%s%s%s\n",
					(int) (item - Actions[prs->state->state].action),
					(item->flags & A_BINGO) ? " BINGO" : "",
					(item->flags & A_POP) ? " POP" : "",
					(item->flags & A_PUSH) ? " PUSH" : "",
					(item->flags & A_RERUN) ? " RERUN" : "",
					(item->flags & A_CLEAR) ? " CLEAR" : "",
					(item->flags & A_MERGE) ? " MERGE" : "",
					(item->flags & A_CLRALL) ? " CLRALL" : "",
					(item->tostate != TPS_Null) ? " tostate " : "",
					(item->tostate != TPS_Null) ? Actions[item->tostate].state_name : "",
					(item->type > 0) ? " type " : "",
					tok_alias[item->type]);
		}
#endif
1415 1416 1417 1418 1419

		/* call special handler if exists */
		if (item->special)
			item->special(prs);

1420
		/* BINGO, token is found */
1421 1422 1423
		if (item->flags & A_BINGO)
		{
			Assert(item->type > 0);
1424 1425 1426
			prs->lenbytetoken = prs->state->lenbytetoken;
			prs->lenchartoken = prs->state->lenchartoken;
			prs->state->lenbytetoken = prs->state->lenchartoken = 0;
1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473
			prs->type = item->type;
		}

		/* do various actions by flags */
		if (item->flags & A_POP)
		{						/* pop stored state in stack */
			TParserPosition *ptr = prs->state->prev;

			pfree(prs->state);
			prs->state = ptr;
			Assert(prs->state);
		}
		else if (item->flags & A_PUSH)
		{						/* push (store) state in stack */
			prs->state->pushedAtAction = item;	/* remember where we push */
			prs->state = newTParserPosition(prs->state);
		}
		else if (item->flags & A_CLEAR)
		{						/* clear previous pushed state */
			TParserPosition *ptr;

			Assert(prs->state->prev);
			ptr = prs->state->prev->prev;
			pfree(prs->state->prev);
			prs->state->prev = ptr;
		}
		else if (item->flags & A_CLRALL)
		{						/* clear all previous pushed state */
			TParserPosition *ptr;

			while (prs->state->prev)
			{
				ptr = prs->state->prev->prev;
				pfree(prs->state->prev);
				prs->state->prev = ptr;
			}
		}
		else if (item->flags & A_MERGE)
		{						/* merge posinfo with current and pushed state */
			TParserPosition *ptr = prs->state;

			Assert(prs->state->prev);
			prs->state = prs->state->prev;

			prs->state->posbyte = ptr->posbyte;
			prs->state->poschar = ptr->poschar;
			prs->state->charlen = ptr->charlen;
1474 1475
			prs->state->lenbytetoken = ptr->lenbytetoken;
			prs->state->lenchartoken = ptr->lenchartoken;
1476 1477 1478 1479 1480 1481 1482 1483
			pfree(ptr);
		}

		/* set new state if pointed */
		if (item->tostate != TPS_Null)
			prs->state->state = item->tostate;

		/* check for go away */
1484 1485 1486
		if ((item->flags & A_BINGO) ||
			(prs->state->posbyte >= prs->lenstr &&
			 (item->flags & A_RERUN) == 0))
1487 1488
			break;

1489
		/* go to beginning of loop if we should rerun or we just restore state */
1490 1491 1492 1493 1494 1495 1496
		if (item->flags & (A_RERUN | A_POP))
			continue;

		/* move forward */
		if (prs->state->charlen)
		{
			prs->state->posbyte += prs->state->charlen;
1497
			prs->state->lenbytetoken += prs->state->charlen;
1498
			prs->state->poschar++;
1499
			prs->state->lenchartoken++;
1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539
		}
	}

	return (item && (item->flags & A_BINGO)) ? true : false;
}

Datum
prsd_lextype(PG_FUNCTION_ARGS)
{
	LexDescr   *descr = (LexDescr *) palloc(sizeof(LexDescr) * (LASTNUM + 1));
	int			i;

	for (i = 1; i <= LASTNUM; i++)
	{
		descr[i - 1].lexid = i;
		descr[i - 1].alias = pstrdup(tok_alias[i]);
		descr[i - 1].descr = pstrdup(lex_descr[i]);
	}

	descr[LASTNUM].lexid = 0;

	PG_RETURN_POINTER(descr);
}

Datum
prsd_start(PG_FUNCTION_ARGS)
{
	PG_RETURN_POINTER(TParserInit((char *) PG_GETARG_POINTER(0), PG_GETARG_INT32(1)));
}

Datum
prsd_nexttoken(PG_FUNCTION_ARGS)
{
	TParser    *p = (TParser *) PG_GETARG_POINTER(0);
	char	  **t = (char **) PG_GETARG_POINTER(1);
	int		   *tlen = (int *) PG_GETARG_POINTER(2);

	if (!TParserGet(p))
		PG_RETURN_INT32(0);

1540 1541
	*t = p->token;
	*tlen = p->lenbytetoken;
1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554

	PG_RETURN_INT32(p->type);
}

Datum
prsd_end(PG_FUNCTION_ARGS)
{
	TParser    *p = (TParser *) PG_GETARG_POINTER(0);

	TParserClose(p);
	PG_RETURN_VOID();
}

1555 1556 1557
#define LEAVETOKEN(x)	( (x)==SPACE )
#define COMPLEXTOKEN(x) ( (x)==URL_T || (x)==NUMHWORD || (x)==ASCIIHWORD || (x)==HWORD )
#define ENDPUNCTOKEN(x) ( (x)==SPACE )
1558

1559 1560 1561 1562 1563
#define TS_IDIGNORE(x) ( (x)==TAG_T || (x)==PROTOCOL || (x)==SPACE || (x)==HTMLENTITY )
#define HLIDIGNORE(x) ( (x)==URL_T || (x)==TAG_T || (x)==NUMHWORD || (x)==ASCIIHWORD || (x)==HWORD )
#define HTMLHLIDIGNORE(x) ( (x)==URL_T || (x)==NUMHWORD || (x)==ASCIIHWORD || (x)==HWORD )
#define NONWORDTOKEN(x) ( (x)==SPACE || HLIDIGNORE(x) )
#define NOENDTOKEN(x)	( NONWORDTOKEN(x) || (x)==SCIENTIFIC || (x)==VERSIONNUMBER || (x)==DECIMAL || (x)==SIGNEDINT || (x)==UNSIGNEDINT || TS_IDIGNORE(x) )
1564 1565 1566 1567 1568 1569 1570 1571

typedef struct
{
	HeadlineWordEntry *words;
	int			len;
} hlCheck;

static bool
1572
checkcondition_HL(void *checkval, QueryOperand * val)
1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597
{
	int			i;

	for (i = 0; i < ((hlCheck *) checkval)->len; i++)
	{
		if (((hlCheck *) checkval)->words[i].item == val)
			return true;
	}
	return false;
}


static bool
hlCover(HeadlineParsedText * prs, TSQuery query, int *p, int *q)
{
	int			i,
				j;
	QueryItem  *item = GETQUERY(query);
	int			pos = *p;

	*q = 0;
	*p = 0x7fffffff;

	for (j = 0; j < query->size; j++)
	{
1598
		if (item->type != QI_VAL)
1599 1600 1601 1602 1603 1604
		{
			item++;
			continue;
		}
		for (i = pos; i < prs->curwords; i++)
		{
1605
			if (prs->words[i].item == &item->operand)
1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620
			{
				if (i > *q)
					*q = i;
				break;
			}
		}
		item++;
	}

	if (*q == 0)
		return false;

	item = GETQUERY(query);
	for (j = 0; j < query->size; j++)
	{
1621
		if (item->type != QI_VAL)
1622 1623 1624 1625 1626 1627
		{
			item++;
			continue;
		}
		for (i = *q; i >= pos; i--)
		{
1628
			if (prs->words[i].item == &item->operand)
1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659
			{
				if (i < *p)
					*p = i;
				break;
			}
		}
		item++;
	}

	if (*p <= *q)
	{
		hlCheck		ch;

		ch.words = &(prs->words[*p]);
		ch.len = *q - *p + 1;
		if (TS_execute(GETQUERY(query), &ch, false, checkcondition_HL))
			return true;
		else
		{
			(*p)++;
			return hlCover(prs, query, p, q);
		}
	}

	return false;
}

Datum
prsd_headline(PG_FUNCTION_ARGS)
{
	HeadlineParsedText *prs = (HeadlineParsedText *) PG_GETARG_POINTER(0);
1660
	List	   *prsoptions = (List *) PG_GETARG_POINTER(1);
1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679
	TSQuery		query = PG_GETARG_TSQUERY(2);

	/* from opt + start and and tag */
	int			min_words = 15;
	int			max_words = 35;
	int			shortword = 3;

	int			p = 0,
				q = 0;
	int			bestb = -1,
				beste = -1;
	int			bestlen = -1;
	int			pose = 0,
				posb,
				poslen,
				curlen;

	int			i;
	int			highlight = 0;
1680
	ListCell   *l;
1681 1682 1683 1684

	/* config */
	prs->startsel = NULL;
	prs->stopsel = NULL;
1685
	foreach(l, prsoptions)
1686
	{
1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711
		DefElem    *defel = (DefElem *) lfirst(l);
		char	   *val = defGetString(defel);

		if (pg_strcasecmp(defel->defname, "MaxWords") == 0)
			max_words = pg_atoi(val, sizeof(int32), 0);
		else if (pg_strcasecmp(defel->defname, "MinWords") == 0)
			min_words = pg_atoi(val, sizeof(int32), 0);
		else if (pg_strcasecmp(defel->defname, "ShortWord") == 0)
			shortword = pg_atoi(val, sizeof(int32), 0);
		else if (pg_strcasecmp(defel->defname, "StartSel") == 0)
			prs->startsel = pstrdup(val);
		else if (pg_strcasecmp(defel->defname, "StopSel") == 0)
			prs->stopsel = pstrdup(val);
		else if (pg_strcasecmp(defel->defname, "HighlightAll") == 0)
			highlight = (pg_strcasecmp(val, "1") == 0 ||
						 pg_strcasecmp(val, "on") == 0 ||
						 pg_strcasecmp(val, "true") == 0 ||
						 pg_strcasecmp(val, "t") == 0 ||
						 pg_strcasecmp(val, "y") == 0 ||
						 pg_strcasecmp(val, "yes") == 0);
		else
			ereport(ERROR,
					(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
					 errmsg("unrecognized headline parameter: \"%s\"",
							defel->defname)));
1712 1713 1714 1715
	}

	if (highlight == 0)
	{
1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728
		if (min_words >= max_words)
			ereport(ERROR,
					(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
					 errmsg("MinWords should be less than MaxWords")));
		if (min_words <= 0)
			ereport(ERROR,
					(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
					 errmsg("MinWords should be positive")));
		if (shortword < 0)
			ereport(ERROR,
					(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
					 errmsg("ShortWord should be >= 0")));

1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857
		while (hlCover(prs, query, &p, &q))
		{
			/* find cover len in words */
			curlen = 0;
			poslen = 0;
			for (i = p; i <= q && curlen < max_words; i++)
			{
				if (!NONWORDTOKEN(prs->words[i].type))
					curlen++;
				if (prs->words[i].item && !prs->words[i].repeated)
					poslen++;
				pose = i;
			}

			if (poslen < bestlen && !(NOENDTOKEN(prs->words[beste].type) || prs->words[beste].len <= shortword))
			{
				/* best already finded, so try one more cover */
				p++;
				continue;
			}

			posb = p;
			if (curlen < max_words)
			{					/* find good end */
				for (i = i - 1; i < prs->curwords && curlen < max_words; i++)
				{
					if (i != q)
					{
						if (!NONWORDTOKEN(prs->words[i].type))
							curlen++;
						if (prs->words[i].item && !prs->words[i].repeated)
							poslen++;
					}
					pose = i;
					if (NOENDTOKEN(prs->words[i].type) || prs->words[i].len <= shortword)
						continue;
					if (curlen >= min_words)
						break;
				}
				if (curlen < min_words && i >= prs->curwords)
				{				/* got end of text and our cover is shoter
								 * than min_words */
					for (i = p; i >= 0; i--)
					{
						if (!NONWORDTOKEN(prs->words[i].type))
							curlen++;
						if (prs->words[i].item && !prs->words[i].repeated)
							poslen++;
						if (NOENDTOKEN(prs->words[i].type) || prs->words[i].len <= shortword)
							continue;
						if (curlen >= min_words)
							break;
					}
					posb = (i >= 0) ? i : 0;
				}
			}
			else
			{					/* shorter cover :((( */
				for (; curlen > min_words; i--)
				{
					if (!NONWORDTOKEN(prs->words[i].type))
						curlen--;
					if (prs->words[i].item && !prs->words[i].repeated)
						poslen--;
					pose = i;
					if (NOENDTOKEN(prs->words[i].type) || prs->words[i].len <= shortword)
						continue;
					break;
				}
			}

			if (bestlen < 0 || (poslen > bestlen && !(NOENDTOKEN(prs->words[pose].type) || prs->words[pose].len <= shortword)) ||
				(bestlen >= 0 && !(NOENDTOKEN(prs->words[pose].type) || prs->words[pose].len <= shortword) &&
				 (NOENDTOKEN(prs->words[beste].type) || prs->words[beste].len <= shortword)))
			{
				bestb = posb;
				beste = pose;
				bestlen = poslen;
			}

			p++;
		}

		if (bestlen < 0)
		{
			curlen = 0;
			for (i = 0; i < prs->curwords && curlen < min_words; i++)
			{
				if (!NONWORDTOKEN(prs->words[i].type))
					curlen++;
				pose = i;
			}
			bestb = 0;
			beste = pose;
		}
	}
	else
	{
		bestb = 0;
		beste = prs->curwords - 1;
	}

	for (i = bestb; i <= beste; i++)
	{
		if (prs->words[i].item)
			prs->words[i].selected = 1;
		if (highlight == 0)
		{
			if (HLIDIGNORE(prs->words[i].type))
				prs->words[i].replace = 1;
		}
		else
		{
			if (HTMLHLIDIGNORE(prs->words[i].type))
				prs->words[i].replace = 1;
		}

		prs->words[i].in = (prs->words[i].repeated) ? 0 : 1;
	}

	if (!prs->startsel)
		prs->startsel = pstrdup("<b>");
	if (!prs->stopsel)
		prs->stopsel = pstrdup("</b>");
	prs->startsellen = strlen(prs->startsel);
	prs->stopsellen = strlen(prs->stopsel);

	PG_RETURN_POINTER(prs);
}