added mysql source for win build

--HG--
branch : trunk

externals/mysql/extlib/regex/cclass.h

@@ -0,0 +1,22 @@
+/* character-class table */
+
+#define CCLASS_ALNUM	0
+#define CCLASS_ALPHA	1
+#define CCLASS_BLANK	2
+#define CCLASS_CNTRL	3
+#define CCLASS_DIGIT	4
+#define CCLASS_GRAPH	5
+#define CCLASS_LOWER	6
+#define CCLASS_PRINT	7
+#define CCLASS_PUNCT	8
+#define CCLASS_SPACE	9
+#define CCLASS_UPPER	10
+#define CCLASS_XDIGIT	11
+#define CCLASS_LAST	12
+
+extern struct cclass {
+	const char *name;
+	const char *chars;
+	const char *multis;
+	uint  mask;
+} cclasses[];

externals/mysql/extlib/regex/cname.h

@@ -0,0 +1,104 @@
+/* character-name table */
+static struct cname {
+  const char *name;
+  const char code;
+} cnames[] = {
+	{"NUL",	        '\0'},
+	{"SOH",	        '\001'},
+	{"STX",	        '\002'},
+	{"ETX",	        '\003'},
+	{"EOT",	        '\004'},
+	{"ENQ",	        '\005'},
+	{"ACK",	        '\006'},
+	{"BEL",	        '\007'},
+	{"alert",	'\007'},
+	{"BS",		'\010'},
+	{"backspace",	'\b'},
+	{"HT",		'\011'},
+	{"tab",       	'\t'},
+	{"LF",		'\012'},
+	{"newline",	'\n'},
+	{"VT",		'\013'},
+	{"vertical-tab",'\v'},
+	{"FF",		'\014'},
+	{"form-feed",	'\f'},
+	{"CR",		'\015'},
+	{"carriage-return",	'\r'},
+	{"SO",	                '\016'},
+	{"SI",	                '\017'},
+	{"DLE",	        '\020'},
+	{"DC1",	        '\021'},
+	{"DC2",	        '\022'},
+	{"DC3",	        '\023'},
+	{"DC4",	        '\024'},
+	{"NAK",	        '\025'},
+	{"SYN",	        '\026'},
+	{"ETB",	        '\027'},
+	{"CAN",	        '\030'},
+	{"EM",	        '\031'},
+	{"SUB",	        '\032'},
+	{"ESC",	        '\033'},
+	{"IS4",	        '\034'},
+	{"FS",	        '\034'},
+	{"IS3",	        '\035'},
+	{"GS",	        '\035'},
+	{"IS2",	        '\036'},
+	{"RS",	        '\036'},
+	{"IS1",	        '\037'},
+	{"US",	        '\037'},
+	{"space",	' '},
+	{"exclamation-mark",	'!'},
+	{"quotation-mark",	'"'},
+	{"number-sign",       	'#'},
+	{"dollar-sign",       	'$'},
+	{"percent-sign",	'%'},
+	{"ampersand",		'&'},
+	{"apostrophe",		'\''},
+	{"left-parenthesis",	'('},
+	{"right-parenthesis",	')'},
+	{"asterisk",	        '*'},
+	{"plus-sign",	        '+'},
+	{"comma",	        ','},
+	{"hyphen",	        '-'},
+	{"hyphen-minus",	'-'},
+	{"period",		'.'},
+	{"full-stop",		'.'},
+	{"slash",		'/'},
+	{"solidus",		'/'},
+	{"zero",		'0'},
+	{"one",			'1'},
+	{"two",			'2'},
+	{"three",		'3'},
+	{"four",		'4'},
+	{"five",		'5'},
+	{"six",			'6'},
+	{"seven",		'7'},
+	{"eight",		'8'},
+	{"nine",		'9'},
+	{"colon",		':'},
+	{"semicolon",		';'},
+	{"less-than-sign",	'<'},
+	{"equals-sign",		'='},
+	{"greater-than-sign",	'>'},
+	{"question-mark",	'?'},
+	{"commercial-at",	'@'},
+	{"left-square-bracket",	'['},
+	{"backslash",		'\\'},
+	{"reverse-solidus",	'\\'},
+	{"right-square-bracket",']'},
+	{"circumflex",		'^'},
+	{"circumflex-accent",	'^'},
+	{"underscore",		'_'},
+	{"low-line",		'_'},
+	{"grave-accent",	'`'},
+	{"left-brace",		'{'},
+	{"left-curly-bracket",	'{'},
+	{"vertical-line",	'|'},
+	{"right-brace",		'}'},
+	{"right-curly-bracket",	'}'},
+	{"tilde",		'~'},
+	{"DEL",	        	'\177'},
+ 	{NULL,			0},
+};
+
+

externals/mysql/extlib/regex/debug.c

@@ -0,0 +1,246 @@
+#include <my_global.h>
+#include <m_ctype.h>
+#include <m_string.h>
+#include <sys/types.h>
+
+#include "my_regex.h"
+#include "utils.h"
+#include "regex2.h"
+#include "debug.ih"
+
+/* Added extra paramter to regchar to remove static buffer ; Monty 96.11.27 */
+
+/*
+ - regprint - print a regexp for debugging
+ == void regprint(regex_t *r, FILE *d);
+ */
+void
+regprint(r, d)
+my_regex_t *r;
+FILE *d;
+{
+	register struct re_guts *g = r->re_g;
+	register int i;
+	register int c;
+	register int last;
+	int nincat[NC];
+	char buf[10];
+
+	fprintf(d, "%ld states, %d categories", (long)g->nstates,
+							g->ncategories);
+	fprintf(d, ", first %ld last %ld", (long)g->firststate,
+						(long)g->laststate);
+	if (g->iflags&USEBOL)
+		fprintf(d, ", USEBOL");
+	if (g->iflags&USEEOL)
+		fprintf(d, ", USEEOL");
+	if (g->iflags&BAD)
+		fprintf(d, ", BAD");
+	if (g->nsub > 0)
+		fprintf(d, ", nsub=%ld", (long)g->nsub);
+	if (g->must != NULL)
+		fprintf(d, ", must(%ld) `%*s'", (long)g->mlen, (int)g->mlen,
+								g->must);
+	if (g->backrefs)
+		fprintf(d, ", backrefs");
+	if (g->nplus > 0)
+		fprintf(d, ", nplus %ld", (long)g->nplus);
+	fprintf(d, "\n");
+	s_print(r->charset, g, d);
+	for (i = 0; i < g->ncategories; i++) {
+		nincat[i] = 0;
+		for (c = CHAR_MIN; c <= CHAR_MAX; c++)
+			if (g->categories[c] == i)
+				nincat[i]++;
+	}
+	fprintf(d, "cc0#%d", nincat[0]);
+	for (i = 1; i < g->ncategories; i++)
+		if (nincat[i] == 1) {
+			for (c = CHAR_MIN; c <= CHAR_MAX; c++)
+				if (g->categories[c] == i)
+					break;
+			fprintf(d, ", %d=%s", i, regchar(r->charset,c,buf));
+		}
+	fprintf(d, "\n");
+	for (i = 1; i < g->ncategories; i++)
+		if (nincat[i] != 1) {
+			fprintf(d, "cc%d\t", i);
+			last = -1;
+			for (c = CHAR_MIN; c <= CHAR_MAX+1; c++)	/* +1 does flush */
+				if (c <= CHAR_MAX && g->categories[c] == i) {
+					if (last < 0) {
+						fprintf(d, "%s", regchar(r->charset,c,buf));
+						last = c;
+					}
+				} else {
+					if (last >= 0) {
+						if (last != c-1)
+							fprintf(d, "-%s",
+								regchar(r->charset,c-1,buf));
+						last = -1;
+					}
+				}
+			fprintf(d, "\n");
+		}
+}
+
+/*
+ - s_print - print the strip for debugging
+ == static void s_print(register struct re_guts *g, FILE *d);
+ */
+static void
+s_print(charset, g, d)
+CHARSET_INFO *charset;
+register struct re_guts *g;
+FILE *d;
+{
+	register sop *s;
+	register cset *cs;
+	register int i;
+	register int done = 0;
+	register sop opnd;
+	register int col = 0;
+	register int last;
+	register sopno offset = 2;
+	char buf[10];
+#	define	GAP()	{	if (offset % 5 == 0) { \
+					if (col > 40) { \
+						fprintf(d, "\n\t"); \
+						col = 0; \
+					} else { \
+						fprintf(d, " "); \
+						col++; \
+					} \
+				} else \
+					col++; \
+				offset++; \
+			}
+
+	if (OP(g->strip[0]) != OEND)
+		fprintf(d, "missing initial OEND!\n");
+	for (s = &g->strip[1]; !done; s++) {
+		opnd = OPND(*s);
+		switch (OP(*s)) {
+		case OEND:
+			fprintf(d, "\n");
+			done = 1;
+			break;
+		case OCHAR:
+			if (strchr("\\|()^$.[+*?{}!<> ", (char)opnd) != NULL)
+				fprintf(d, "\\%c", (char)opnd);
+			else
+				fprintf(d, "%s", regchar(charset,(char)opnd,buf));
+			break;
+		case OBOL:
+			fprintf(d, "^");
+			break;
+		case OEOL:
+			fprintf(d, "$");
+			break;
+		case OBOW:
+			fprintf(d, "\\{");
+			break;
+		case OEOW:
+			fprintf(d, "\\}");
+			break;
+		case OANY:
+			fprintf(d, ".");
+			break;
+		case OANYOF:
+			fprintf(d, "[(%ld)", (long)opnd);
+			cs = &g->sets[opnd];
+			last = -1;
+			for (i = 0; i < g->csetsize+1; i++)	/* +1 flushes */
+				if (CHIN(cs, i) && i < g->csetsize) {
+					if (last < 0) {
+						fprintf(d, "%s", regchar(charset,i,buf));
+						last = i;
+					}
+				} else {
+					if (last >= 0) {
+						if (last != i-1)
+							fprintf(d, "-%s",
+								regchar(charset,i-1,buf));
+						last = -1;
+					}
+				}
+			fprintf(d, "]");
+			break;
+		case OBACK_:
+			fprintf(d, "(\\<%ld>", (long)opnd);
+			break;
+		case O_BACK:
+			fprintf(d, "<%ld>\\)", (long)opnd);
+			break;
+		case OPLUS_:
+			fprintf(d, "(+");
+			if (OP(*(s+opnd)) != O_PLUS)
+				fprintf(d, "<%ld>", (long)opnd);
+			break;
+		case O_PLUS:
+			if (OP(*(s-opnd)) != OPLUS_)
+				fprintf(d, "<%ld>", (long)opnd);
+			fprintf(d, "+)");
+			break;
+		case OQUEST_:
+			fprintf(d, "(?");
+			if (OP(*(s+opnd)) != O_QUEST)
+				fprintf(d, "<%ld>", (long)opnd);
+			break;
+		case O_QUEST:
+			if (OP(*(s-opnd)) != OQUEST_)
+				fprintf(d, "<%ld>", (long)opnd);
+			fprintf(d, "?)");
+			break;
+		case OLPAREN:
+			fprintf(d, "((<%ld>", (long)opnd);
+			break;
+		case ORPAREN:
+			fprintf(d, "<%ld>))", (long)opnd);
+			break;
+		case OCH_:
+			fprintf(d, "<");
+			if (OP(*(s+opnd)) != OOR2)
+				fprintf(d, "<%ld>", (long)opnd);
+			break;
+		case OOR1:
+			if (OP(*(s-opnd)) != OOR1 && OP(*(s-opnd)) != OCH_)
+				fprintf(d, "<%ld>", (long)opnd);
+			fprintf(d, "|");
+			break;
+		case OOR2:
+			fprintf(d, "|");
+			if (OP(*(s+opnd)) != OOR2 && OP(*(s+opnd)) != O_CH)
+				fprintf(d, "<%ld>", (long)opnd);
+			break;
+		case O_CH:
+			if (OP(*(s-opnd)) != OOR1)
+				fprintf(d, "<%ld>", (long)opnd);
+			fprintf(d, ">");
+			break;
+		default:
+			fprintf(d, "!%ld(%ld)!", OP(*s), opnd);
+			break;
+		}
+		if (!done)
+			GAP();
+	}
+}
+
+/*
+ - regchar - make a character printable
+ == static char *regchar(int ch);
+ */
+static char *			/* -> representation */
+regchar(charset,ch,buf)
+CHARSET_INFO *charset;
+int ch;
+char *buf;
+{
+
+	if (my_isprint(charset,ch) || ch == ' ')
+		sprintf(buf, "%c", ch);
+	else
+		sprintf(buf, "\\%o", ch);
+	return(buf);
+}

externals/mysql/extlib/regex/main.c

@@ -0,0 +1,509 @@
+#include <my_global.h>
+#include <m_string.h>
+#include <sys/types.h>
+#include <assert.h>
+
+#include "my_regex.h"
+#include "main.ih"
+
+char *progname;
+int debug = 0;
+int line = 0;
+int status = 0;
+
+int copts = REG_EXTENDED;
+int eopts = 0;
+regoff_t startoff = 0;
+regoff_t endoff = 0;
+
+
+extern int split();
+extern void regprint();
+
+/*
+ - main - do the simple case, hand off to regress() for regression
+ */
+int main(argc, argv)
+int argc;
+char *argv[];
+{
+	my_regex_t re;
+#	define	NS	10
+	my_regmatch_t subs[NS];
+	char erbuf[100];
+	int err;
+	size_t len;
+	int c;
+	int errflg = 0;
+	register int i;
+	extern int optind;
+	extern char *optarg;
+
+	progname = argv[0];
+
+	while ((c = getopt(argc, argv, "c:e:S:E:x")) != EOF)
+		switch (c) {
+		case 'c':	/* compile options */
+			copts = options('c', optarg);
+			break;
+		case 'e':	/* execute options */
+			eopts = options('e', optarg);
+			break;
+		case 'S':	/* start offset */
+			startoff = (regoff_t)atoi(optarg);
+			break;
+		case 'E':	/* end offset */
+			endoff = (regoff_t)atoi(optarg);
+			break;
+		case 'x':	/* Debugging. */
+			debug++;
+			break;
+		case '?':
+		default:
+			errflg++;
+			break;
+		}
+	if (errflg) {
+		fprintf(stderr, "usage: %s ", progname);
+		fprintf(stderr, "[-c copt][-C][-d] [re]\n");
+		exit(2);
+	}
+
+	if (optind >= argc) {
+		regress(stdin);
+		exit(status);
+	}
+
+	err = my_regcomp(&re, argv[optind++], copts, &my_charset_latin1);
+	if (err) {
+		len = my_regerror(err, &re, erbuf, sizeof(erbuf));
+		fprintf(stderr, "error %s, %d/%d `%s'\n",
+			eprint(err), (int) len, (int) sizeof(erbuf), erbuf);
+		exit(status);
+	}
+	regprint(&re, stdout);
+
+	if (optind >= argc) {
+		my_regfree(&re);
+		exit(status);
+	}
+
+	if (eopts&REG_STARTEND) {
+		subs[0].rm_so = startoff;
+		subs[0].rm_eo = strlen(argv[optind]) - endoff;
+	}
+	err = my_regexec(&re, argv[optind], (size_t)NS, subs, eopts);
+	if (err) {
+		len = my_regerror(err, &re, erbuf, sizeof(erbuf));
+		fprintf(stderr, "error %s, %d/%d `%s'\n",
+			eprint(err), (int) len, (int) sizeof(erbuf), erbuf);
+		exit(status);
+	}
+	if (!(copts&REG_NOSUB)) {
+		len = (int)(subs[0].rm_eo - subs[0].rm_so);
+		if (subs[0].rm_so != -1) {
+			if (len != 0)
+				printf("match `%.*s'\n", (int)len,
+					argv[optind] + subs[0].rm_so);
+			else
+				printf("match `'@%.1s\n",
+					argv[optind] + subs[0].rm_so);
+		}
+		for (i = 1; i < NS; i++)
+			if (subs[i].rm_so != -1)
+				printf("(%d) `%.*s'\n", i,
+					(int)(subs[i].rm_eo - subs[i].rm_so),
+					argv[optind] + subs[i].rm_so);
+	}
+	exit(status);
+}
+
+/*
+ - regress - main loop of regression test
+ == void regress(FILE *in);
+ */
+void
+regress(in)
+FILE *in;
+{
+	char inbuf[1000];
+#	define	MAXF	10
+	char *f[MAXF];
+	int nf;
+	int i;
+	char erbuf[100];
+	size_t ne;
+	const char *badpat = "invalid regular expression";
+#	define	SHORT	10
+	const char *bpname = "REG_BADPAT";
+	my_regex_t re;
+
+	while (fgets(inbuf, sizeof(inbuf), in) != NULL) {
+		line++;
+		if (inbuf[0] == '#' || inbuf[0] == '\n')
+			continue;			/* NOTE CONTINUE */
+		inbuf[strlen(inbuf)-1] = '\0';	/* get rid of stupid \n */
+		if (debug)
+			fprintf(stdout, "%d:\n", line);
+		nf = split(inbuf, f, MAXF, "\t\t");
+		if (nf < 3) {
+			fprintf(stderr, "bad input, line %d\n", line);
+			exit(1);
+		}
+		for (i = 0; i < nf; i++)
+			if (strcmp(f[i], "\"\"") == 0)
+				f[i] = (char*) "";
+		if (nf <= 3)
+			f[3] = NULL;
+		if (nf <= 4)
+			f[4] = NULL;
+		rx_try(f[0], f[1], f[2], f[3], f[4], options('c', f[1]));
+		if (opt('&', f[1]))	/* try with either type of RE */
+			rx_try(f[0], f[1], f[2], f[3], f[4],
+					options('c', f[1]) &~ REG_EXTENDED);
+	}
+
+	ne = my_regerror(REG_BADPAT, (my_regex_t *)NULL, erbuf, sizeof(erbuf));
+	if (strcmp(erbuf, badpat) != 0 || ne != strlen(badpat)+1) {
+		fprintf(stderr, "end: regerror() test gave `%s' not `%s'\n",
+							erbuf, badpat);
+		status = 1;
+	}
+	ne = my_regerror(REG_BADPAT, (my_regex_t *)NULL, erbuf, (size_t)SHORT);
+	if (strncmp(erbuf, badpat, SHORT-1) != 0 || erbuf[SHORT-1] != '\0' ||
+						ne != strlen(badpat)+1) {
+		fprintf(stderr, "end: regerror() short test gave `%s' not `%.*s'\n",
+						erbuf, SHORT-1, badpat);
+		status = 1;
+	}
+	ne = my_regerror(REG_ITOA|REG_BADPAT, (my_regex_t *)NULL, erbuf, sizeof(erbuf));
+	if (strcmp(erbuf, bpname) != 0 || ne != strlen(bpname)+1) {
+		fprintf(stderr, "end: regerror() ITOA test gave `%s' not `%s'\n",
+						erbuf, bpname);
+		status = 1;
+	}
+	re.re_endp = bpname;
+	ne = my_regerror(REG_ATOI, &re, erbuf, sizeof(erbuf));
+	if (atoi(erbuf) != (int)REG_BADPAT) {
+		fprintf(stderr, "end: regerror() ATOI test gave `%s' not `%ld'\n",
+						erbuf, (long)REG_BADPAT);
+		status = 1;
+	} else if (ne != strlen(erbuf)+1) {
+		fprintf(stderr, "end: regerror() ATOI test len(`%s') = %ld\n",
+						erbuf, (long)REG_BADPAT);
+		status = 1;
+	}
+}
+
+/*
+ - rx_try - try it, and report on problems
+ == void rx_try(char *f0, char *f1, char *f2, char *f3, char *f4, int opts);
+ */
+void
+rx_try(f0, f1, f2, f3, f4, opts)
+char *f0;
+char *f1;
+char *f2;
+char *f3;
+char *f4;
+int opts;			/* may not match f1 */
+{
+	my_regex_t re;
+#	define	NSUBS	10
+	my_regmatch_t subs[NSUBS];
+#	define	NSHOULD	15
+	char *should[NSHOULD];
+	int nshould;
+	char erbuf[100];
+	int err;
+	int len;
+	const char *type = (opts & REG_EXTENDED) ? "ERE" : "BRE";
+	register int i;
+	char *grump;
+	char f0copy[1000];
+	char f2copy[1000];
+
+	strcpy(f0copy, f0);
+	re.re_endp = (opts&REG_PEND) ? f0copy + strlen(f0copy) : NULL;
+	fixstr(f0copy);
+	err = my_regcomp(&re, f0copy, opts, &my_charset_latin1);
+	if (err != 0 && (!opt('C', f1) || err != efind(f2))) {
+		/* unexpected error or wrong error */
+		len = my_regerror(err, &re, erbuf, sizeof(erbuf));
+		fprintf(stderr, "%d: %s error %s, %d/%d `%s'\n",
+					line, type, eprint(err), len,
+					(int) sizeof(erbuf), erbuf);
+		status = 1;
+	} else if (err == 0 && opt('C', f1)) {
+		/* unexpected success */
+		fprintf(stderr, "%d: %s should have given REG_%s\n",
+						line, type, f2);
+		status = 1;
+		err = 1;	/* so we won't try regexec */
+	}
+
+	if (err != 0) {
+		my_regfree(&re);
+		return;
+	}
+
+	strcpy(f2copy, f2);
+	fixstr(f2copy);
+
+	if (options('e', f1)&REG_STARTEND) {
+		if (strchr(f2, '(') == NULL || strchr(f2, ')') == NULL)
+			fprintf(stderr, "%d: bad STARTEND syntax\n", line);
+		subs[0].rm_so = strchr(f2, '(') - f2 + 1;
+		subs[0].rm_eo = strchr(f2, ')') - f2;
+	}
+	err = my_regexec(&re, f2copy, NSUBS, subs, options('e', f1));
+
+	if (err != 0 && (f3 != NULL || err != REG_NOMATCH)) {
+		/* unexpected error or wrong error */
+		len = my_regerror(err, &re, erbuf, sizeof(erbuf));
+		fprintf(stderr, "%d: %s exec error %s, %d/%d `%s'\n",
+					line, type, eprint(err), len,
+					(int) sizeof(erbuf), erbuf);
+		status = 1;
+	} else if (err != 0) {
+		/* nothing more to check */
+	} else if (f3 == NULL) {
+		/* unexpected success */
+		fprintf(stderr, "%d: %s exec should have failed\n",
+						line, type);
+		status = 1;
+		err = 1;		/* just on principle */
+	} else if (opts&REG_NOSUB) {
+		/* nothing more to check */
+	} else if ((grump = check(f2, subs[0], f3)) != NULL) {
+		fprintf(stderr, "%d: %s %s\n", line, type, grump);
+		status = 1;
+		err = 1;
+	}
+
+	if (err != 0 || f4 == NULL) {
+		my_regfree(&re);
+		return;
+	}
+
+	for (i = 1; i < NSHOULD; i++)
+		should[i] = NULL;
+	nshould = split(f4, should+1, NSHOULD-1, ",");
+	if (nshould == 0) {
+		nshould = 1;
+		should[1] = (char*) "";
+	}
+	for (i = 1; i < NSUBS; i++) {
+		grump = check(f2, subs[i], should[i]);
+		if (grump != NULL) {
+			fprintf(stderr, "%d: %s $%d %s\n", line,
+							type, i, grump);
+			status = 1;
+			err = 1;
+		}
+	}
+
+	my_regfree(&re);
+}
+
+/*
+ - options - pick options out of a regression-test string
+ == int options(int type, char *s);
+ */
+int
+options(type, s)
+int type;			/* 'c' compile, 'e' exec */
+char *s;
+{
+	register char *p;
+	register int o = (type == 'c') ? copts : eopts;
+	register const char *legal = (type == 'c') ? "bisnmp" : "^$#tl";
+
+	for (p = s; *p != '\0'; p++)
+		if (strchr(legal, *p) != NULL)
+			switch (*p) {
+			case 'b':
+				o &= ~REG_EXTENDED;
+				break;
+			case 'i':
+				o |= REG_ICASE;
+				break;
+			case 's':
+				o |= REG_NOSUB;
+				break;
+			case 'n':
+				o |= REG_NEWLINE;
+				break;
+			case 'm':
+				o &= ~REG_EXTENDED;
+				o |= REG_NOSPEC;
+				break;
+			case 'p':
+				o |= REG_PEND;
+				break;
+			case '^':
+				o |= REG_NOTBOL;
+				break;
+			case '$':
+				o |= REG_NOTEOL;
+				break;
+			case '#':
+				o |= REG_STARTEND;
+				break;
+			case 't':	/* trace */
+				o |= REG_TRACE;
+				break;
+			case 'l':	/* force long representation */
+				o |= REG_LARGE;
+				break;
+			case 'r':	/* force backref use */
+				o |= REG_BACKR;
+				break;
+			}
+	return(o);
+}
+
+/*
+ - opt - is a particular option in a regression string?
+ == int opt(int c, char *s);
+ */
+int				/* predicate */
+opt(c, s)
+int c;
+char *s;
+{
+	return(strchr(s, c) != NULL);
+}
+
+/*
+ - fixstr - transform magic characters in strings
+ == void fixstr(register char *p);
+ */
+void
+fixstr(p)
+register char *p;
+{
+	if (p == NULL)
+		return;
+
+	for (; *p != '\0'; p++)
+		if (*p == 'N')
+			*p = '\n';
+		else if (*p == 'T')
+			*p = '\t';
+		else if (*p == 'S')
+			*p = ' ';
+		else if (*p == 'Z')
+			*p = '\0';
+}
+
+/*
+ - check - check a substring match
+ == char *check(char *str, regmatch_t sub, char *should);
+ */
+char *				/* NULL or complaint */
+check(str, sub, should)
+char *str;
+my_regmatch_t sub;
+char *should;
+{
+	register int len;
+	register int shlen;
+	register char *p;
+	static char grump[500];
+	register char *at = NULL;
+
+	if (should != NULL && strcmp(should, "-") == 0)
+		should = NULL;
+	if (should != NULL && should[0] == '@') {
+		at = should + 1;
+		should = (char*) "";
+	}
+
+	/* check rm_so and rm_eo for consistency */
+	if (sub.rm_so > sub.rm_eo || (sub.rm_so == -1 && sub.rm_eo != -1) ||
+				(sub.rm_so != -1 && sub.rm_eo == -1) ||
+				(sub.rm_so != -1 && sub.rm_so < 0) ||
+				(sub.rm_eo != -1 && sub.rm_eo < 0) ) {
+		sprintf(grump, "start %ld end %ld", (long)sub.rm_so,
+							(long)sub.rm_eo);
+		return(grump);
+	}
+
+	/* check for no match */
+	if (sub.rm_so == -1 && should == NULL)
+		return(NULL);
+	if (sub.rm_so == -1)
+		return((char*) "did not match");
+
+	/* check for in range */
+	if ((int) sub.rm_eo > (int) strlen(str)) {
+		sprintf(grump, "start %ld end %ld, past end of string",
+					(long)sub.rm_so, (long)sub.rm_eo);
+		return(grump);
+	}
+
+	len = (int)(sub.rm_eo - sub.rm_so);
+	shlen = (int)strlen(should);
+	p = str + sub.rm_so;
+
+	/* check for not supposed to match */
+	if (should == NULL) {
+		sprintf(grump, "matched `%.*s'", len, p);
+		return(grump);
+	}
+
+	/* check for wrong match */
+	if (len != shlen || strncmp(p, should, (size_t)shlen) != 0) {
+		sprintf(grump, "matched `%.*s' instead", len, p);
+		return(grump);
+	}
+	if (shlen > 0)
+		return(NULL);
+
+	/* check null match in right place */
+	if (at == NULL)
+		return(NULL);
+	shlen = strlen(at);
+	if (shlen == 0)
+		shlen = 1;	/* force check for end-of-string */
+	if (strncmp(p, at, shlen) != 0) {
+		sprintf(grump, "matched null at `%.20s'", p);
+		return(grump);
+	}
+	return(NULL);
+}
+
+/*
+ - eprint - convert error number to name
+ == static char *eprint(int err);
+ */
+static char *
+eprint(err)
+int err;
+{
+	static char epbuf[100];
+	size_t len;
+
+	len = my_regerror(REG_ITOA|err, (my_regex_t *)NULL, epbuf, sizeof(epbuf));
+	assert(len <= sizeof(epbuf));
+	return(epbuf);
+}
+
+/*
+ - efind - convert error name to number
+ == static int efind(char *name);
+ */
+static int
+efind(name)
+char *name;
+{
+	static char efbuf[100];
+	my_regex_t re;
+
+	sprintf(efbuf, "REG_%s", name);
+	assert(strlen(efbuf) < sizeof(efbuf));
+	re.re_endp = efbuf;
+	(void) my_regerror(REG_ATOI, &re, efbuf, sizeof(efbuf));
+	return(atoi(efbuf));
+}

externals/mysql/extlib/regex/my_regex.h

@@ -0,0 +1,86 @@
+#ifndef _REGEX_H_
+#define	_REGEX_H_	/* never again */
+/* ========= begin header generated by ./mkh ========= */
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#include "m_ctype.h"
+
+/* === regex2.h === */
+#ifdef _WIN64
+typedef __int64 regoff_t;
+#else
+typedef off_t regoff_t;
+#endif
+struct re_guts;			/* none of your business :-) */
+typedef struct {
+	int re_magic;
+	size_t re_nsub;		/* number of parenthesized subexpressions */
+	const char *re_endp;	/* end pointer for REG_PEND */
+	struct re_guts *re_g;	/* none of your business :-) */
+	CHARSET_INFO *charset;	/* For ctype things */
+} my_regex_t;
+typedef struct {
+	regoff_t rm_so;		/* start of match */
+	regoff_t rm_eo;		/* end of match */
+} my_regmatch_t;
+
+
+/* === regcomp.c === */
+extern int my_regcomp(my_regex_t *, const char *, int, CHARSET_INFO *charset);
+#define	REG_BASIC	0000
+#define	REG_EXTENDED	0001
+#define	REG_ICASE	0002
+#define	REG_NOSUB	0004
+#define	REG_NEWLINE	0010
+#define	REG_NOSPEC	0020
+#define	REG_PEND	0040
+#define	REG_DUMP	0200
+
+
+/* === regerror.c === */
+#define	REG_NOMATCH	 1
+#define	REG_BADPAT	 2
+#define	REG_ECOLLATE	 3
+#define	REG_ECTYPE	 4
+#define	REG_EESCAPE	 5
+#define	REG_ESUBREG	 6
+#define	REG_EBRACK	 7
+#define	REG_EPAREN	 8
+#define	REG_EBRACE	 9
+#define	REG_BADBR	10
+#define	REG_ERANGE	11
+#define	REG_ESPACE	12
+#define	REG_BADRPT	13
+#define	REG_EMPTY	14
+#define	REG_ASSERT	15
+#define	REG_INVARG	16
+#define	REG_ATOI	255	/* convert name to number (!) */
+#define	REG_ITOA	0400	/* convert number to name (!) */
+extern size_t my_regerror(int, const my_regex_t *, char *, size_t);
+
+
+/* === regexec.c === */
+extern int my_regexec(const my_regex_t *, const char *, size_t, my_regmatch_t [], int);
+#define	REG_NOTBOL	00001
+#define	REG_NOTEOL	00002
+#define	REG_STARTEND	00004
+#define	REG_TRACE	00400	/* tracing of execution */
+#define	REG_LARGE	01000	/* force large representation */
+#define	REG_BACKR	02000	/* force use of backref code */
+
+
+/* === regfree.c === */
+extern void my_regfree(my_regex_t *);
+
+/* === reginit.c === */
+
+extern void my_regex_init(CHARSET_INFO *cs);	/* Should be called for multithread progs */
+extern void my_regex_end(void);	/* If one wants a clean end */
+
+#ifdef __cplusplus
+}
+#endif
+/* ========= end header generated by ./mkh ========= */
+#endif

externals/mysql/extlib/regex/regerror.c

@@ -0,0 +1,116 @@
+#include <my_global.h>
+#include <m_string.h>
+#include <m_ctype.h>
+
+#include "my_regex.h"
+#include "utils.h"
+#include "regerror.ih"
+
+/*
+ = #define	REG_NOMATCH	 1
+ = #define	REG_BADPAT	 2
+ = #define	REG_ECOLLATE	 3
+ = #define	REG_ECTYPE	 4
+ = #define	REG_EESCAPE	 5
+ = #define	REG_ESUBREG	 6
+ = #define	REG_EBRACK	 7
+ = #define	REG_EPAREN	 8
+ = #define	REG_EBRACE	 9
+ = #define	REG_BADBR	10
+ = #define	REG_ERANGE	11
+ = #define	REG_ESPACE	12
+ = #define	REG_BADRPT	13
+ = #define	REG_EMPTY	14
+ = #define	REG_ASSERT	15
+ = #define	REG_INVARG	16
+ = #define	REG_ATOI	255	// convert name to number (!)
+ = #define	REG_ITOA	0400	// convert number to name (!)
+ */
+static struct rerr {
+	int code;
+	const char *name;
+        const char *explain;
+} rerrs[] = {
+	{REG_NOMATCH,	"REG_NOMATCH",	"regexec() failed to match"},
+	{REG_BADPAT,	"REG_BADPAT",	"invalid regular expression"},
+	{REG_ECOLLATE,	"REG_ECOLLATE",	"invalid collating element"},
+	{REG_ECTYPE,	"REG_ECTYPE",	"invalid character class"},
+	{REG_EESCAPE,	"REG_EESCAPE",	"trailing backslash (\\)"},
+	{REG_ESUBREG,	"REG_ESUBREG",	"invalid backreference number"},
+	{REG_EBRACK,	"REG_EBRACK",	"brackets ([ ]) not balanced"},
+	{REG_EPAREN,	"REG_EPAREN",	"parentheses not balanced"},
+	{REG_EBRACE,	"REG_EBRACE",	"braces not balanced"},
+	{REG_BADBR,	"REG_BADBR",	"invalid repetition count(s)"},
+	{REG_ERANGE,	"REG_ERANGE",	"invalid character range"},
+	{REG_ESPACE,	"REG_ESPACE",	"out of memory"},
+	{REG_BADRPT,	"REG_BADRPT",	"repetition-operator operand invalid"},
+	{REG_EMPTY,	"REG_EMPTY",	"empty (sub)expression"},
+	{REG_ASSERT,	"REG_ASSERT",	"\"can't happen\" -- you found a bug"},
+	{REG_INVARG,	"REG_INVARG",	"invalid argument to regex routine"},
+	{0,		"",		"*** unknown regexp error code ***"},
+};
+
+/*
+ - regerror - the interface to error numbers
+ = extern size_t regerror(int, const regex_t *, char *, size_t);
+ */
+/* ARGSUSED */
+size_t
+my_regerror(int errcode, const my_regex_t *preg, char *errbuf, size_t errbuf_size)
+{
+	register struct rerr *r;
+	register size_t len;
+	register int target = errcode &~ REG_ITOA;
+	register char *s;
+	char convbuf[50];
+
+	if (errcode == REG_ATOI)
+		s = regatoi(preg, convbuf);
+	else {
+		for (r = rerrs; r->code != 0; r++)
+			if (r->code == target)
+				break;
+
+		if (errcode&REG_ITOA) {
+			if (r->code != 0)
+				(void) strcpy(convbuf, r->name);
+			else
+				sprintf(convbuf, "REG_0x%x", target);
+			assert(strlen(convbuf) < sizeof(convbuf));
+			s = convbuf;
+		} else
+			s = (char*) r->explain;
+	}
+
+	len = strlen(s) + 1;
+	if (errbuf_size > 0) {
+		if (errbuf_size > len)
+			(void) strcpy(errbuf, s);
+		else {
+			(void) strncpy(errbuf, s, errbuf_size-1);
+			errbuf[errbuf_size-1] = '\0';
+		}
+	}
+
+	return(len);
+}
+
+/*
+ - regatoi - internal routine to implement REG_ATOI
+ == static char *regatoi(const regex_t *preg, char *localbuf);
+ */
+static char *
+regatoi(preg, localbuf)
+const my_regex_t *preg;
+char *localbuf;
+{
+	register struct rerr *r;
+	for (r = rerrs; r->code != 0; r++)
+		if (strcmp(r->name, preg->re_endp) == 0)
+			break;
+	if (r->code == 0)
+		return((char*) "0");
+
+	sprintf(localbuf, "%d", r->code);
+	return(localbuf);
+}

externals/mysql/extlib/regex/regex2.h

@@ -0,0 +1,145 @@
+/*
+ * First, the stuff that ends up in the outside-world include file
+ = typedef off_t regoff_t;
+ = typedef struct {
+ = 	int re_magic;
+ = 	size_t re_nsub;		// number of parenthesized subexpressions
+ = 	const char *re_endp;	// end pointer for REG_PEND
+ = 	struct re_guts *re_g;	// none of your business :-)
+ = } regex_t;
+ = typedef struct {
+ = 	regoff_t rm_so;		// start of match
+ = 	regoff_t rm_eo;		// end of match
+ = } regmatch_t;
+ */
+/*
+ * internals of regex_t
+ */
+#ifndef	__regex2_h__
+#define __regex2_h__
+
+#define	MAGIC1	((('r'^0200)<<8) | 'e')
+
+/*
+ * The internal representation is a *strip*, a sequence of
+ * operators ending with an endmarker.  (Some terminology etc. is a
+ * historical relic of earlier versions which used multiple strips.)
+ * Certain oddities in the representation are there to permit running
+ * the machinery backwards; in particular, any deviation from sequential
+ * flow must be marked at both its source and its destination.  Some
+ * fine points:
+ *
+ * - OPLUS_ and O_PLUS are *inside* the loop they create.
+ * - OQUEST_ and O_QUEST are *outside* the bypass they create.
+ * - OCH_ and O_CH are *outside* the multi-way branch they create, while
+ *   OOR1 and OOR2 are respectively the end and the beginning of one of
+ *   the branches.  Note that there is an implicit OOR2 following OCH_
+ *   and an implicit OOR1 preceding O_CH.
+ *
+ * In state representations, an operator's bit is on to signify a state
+ * immediately *preceding* "execution" of that operator.
+ */
+typedef unsigned long sop;	/* strip operator */
+typedef long sopno;
+#define	OPRMASK	0xf8000000
+#define	OPDMASK	0x07ffffff
+#define	OPSHIFT	((unsigned long)27)
+#define	OP(n)	((unsigned long) ((n)&OPRMASK))
+#define	OPND(n)	((unsigned long) ((n)&OPDMASK))
+#define	SOP(op, opnd)	(unsigned long) ((op)|(opnd))
+#define OSHIFT(A) ((unsigned long) (A) << OPSHIFT)
+/* operators			   meaning	operand			*/
+/*						(back, fwd are offsets)	*/
+#define	OEND	OSHIFT(1)	/* endmarker	-			*/
+#define	OCHAR	OSHIFT(2)	/* character	unsigned char		*/
+#define	OBOL	OSHIFT(3)	/* left anchor	-			*/
+#define	OEOL	OSHIFT(4)	/* right anchor	-			*/
+#define	OANY	OSHIFT(5)	/* .		-			*/
+#define	OANYOF	OSHIFT(6)	/* [...]	set number		*/
+#define	OBACK_	OSHIFT(7)	/* begin \d	paren number		*/
+#define	O_BACK	OSHIFT(8)	/* end \d	paren number		*/
+#define	OPLUS_	OSHIFT(9)	/* + prefix	fwd to suffix		*/
+#define	O_PLUS	OSHIFT(10)	/* + suffix	back to prefix		*/
+#define	OQUEST_	OSHIFT(11)	/* ? prefix	fwd to suffix		*/
+#define	O_QUEST	OSHIFT(12)	/* ? suffix	back to prefix		*/
+#define	OLPAREN	OSHIFT(13)	/* (		fwd to )		*/
+#define	ORPAREN	OSHIFT(14)	/* )		back to (		*/
+#define	OCH_	OSHIFT(15)	/* begin choice	fwd to OOR2		*/
+#define	OOR1	OSHIFT(16)	/* | pt. 1	back to OOR1 or OCH_	*/
+#define	OOR2	OSHIFT(17)	/* | pt. 2	fwd to OOR2 or O_CH	*/
+#define	O_CH	OSHIFT(18)	/* end choice	back to OOR1		*/
+#define	OBOW	OSHIFT(19)	/* begin word	-			*/
+#define	OEOW	OSHIFT(20)	/* end word	-			*/
+
+/*
+ * Structure for [] character-set representation.  Character sets are
+ * done as bit vectors, grouped 8 to a byte vector for compactness.
+ * The individual set therefore has both a pointer to the byte vector
+ * and a mask to pick out the relevant bit of each byte.  A hash code
+ * simplifies testing whether two sets could be identical.
+ *
+ * This will get trickier for multicharacter collating elements.  As
+ * preliminary hooks for dealing with such things, we also carry along
+ * a string of multi-character elements, and decide the size of the
+ * vectors at run time.
+ */
+#ifdef __WIN__
+typedef unsigned char uch ;
+#endif
+
+typedef struct {
+	uch *ptr;		/* -> uch [csetsize] */
+	uch mask;		/* bit within array */
+	uch hash;		/* hash code */
+	size_t smultis;
+	char *multis;		/* -> char[smulti]  ab\0cd\0ef\0\0 */
+} cset;
+/* note that CHadd and CHsub are unsafe, and CHIN doesn't yield 0/1 */
+#define	CHadd(cs, c)	((cs)->ptr[(uch)(c)] |= (cs)->mask, (cs)->hash += (uch) (c))
+#define	CHsub(cs, c)	((cs)->ptr[(uch)(c)] &= ~(cs)->mask, (cs)->hash -= (uch)(c))
+#define	CHIN(cs, c)	((cs)->ptr[(uch)(c)] & (cs)->mask)
+#define	MCadd(p, cs, cp)	mcadd(p, cs, cp)	/* regcomp() internal fns */
+#define	MCsub(p, cs, cp)	mcsub(p, cs, cp)
+#define	MCin(p, cs, cp)	mcin(p, cs, cp)
+
+/* stuff for character categories */
+typedef unsigned char cat_t;
+
+/*
+ * main compiled-expression structure
+ */
+struct re_guts {
+	int magic;
+#		define	MAGIC2	((('R'^0200)<<8)|'E')
+	sop *strip;		/* malloced area for strip */
+	int csetsize;		/* number of bits in a cset vector */
+	int ncsets;		/* number of csets in use */
+	cset *sets;		/* -> cset [ncsets] */
+	uch *setbits;		/* -> uch[csetsize][ncsets/CHAR_BIT] */
+	int cflags;		/* copy of regcomp() cflags argument */
+	sopno nstates;		/* = number of sops */
+	sopno firststate;	/* the initial OEND (normally 0) */
+	sopno laststate;	/* the final OEND */
+	int iflags;		/* internal flags */
+#		define	USEBOL	01	/* used ^ */
+#		define	USEEOL	02	/* used $ */
+#		define	BAD	04	/* something wrong */
+	int nbol;		/* number of ^ used */
+	int neol;		/* number of $ used */
+	int ncategories;	/* how many character categories */
+	cat_t *categories;	/* ->catspace[-CHAR_MIN] */
+	char *must;		/* match must contain this string */
+	int mlen;		/* length of must */
+	size_t nsub;		/* copy of re_nsub */
+	int backrefs;		/* does it use back references? */
+	sopno nplus;		/* how deep does it nest +s? */
+	/* catspace must be last */
+	cat_t catspace[1];	/* actually [NC] */
+};
+
+/* misc utilities */
+#undef OUT			/* May be defined in windows */
+#define	OUT	(CHAR_MAX+1)	/* a non-character value */
+#define	ISWORD(s,c)	(my_isalnum(s,c) || (c) == '_')
+
+#endif /* __regex2_h__ */

externals/mysql/extlib/regex/regexec.c

@@ -0,0 +1,139 @@
+/*
+ * the outer shell of regexec()
+ *
+ * This file includes engine.c *twice*, after muchos fiddling with the
+ * macros that code uses.  This lets the same code operate on two different
+ * representations for state sets.
+ */
+#include <my_global.h>
+#include <m_string.h>
+#include <m_ctype.h>
+#ifdef __WIN__
+#include  <limits.h>
+#endif
+#include "my_regex.h"
+#include "utils.h"
+#include "regex2.h"
+
+/* for use in asserts */
+#define nope 0
+
+/* macros for manipulating states, small version */
+#define	states	long
+#define	states1	long		/* for later use in regexec() decision. Ensure Win64 definition is correct.*/
+#define	CLEAR(v)	((v) = 0)
+#define	SET0(v, n)	((v) &= ~((states) 1 << (n)))
+#define	SET1(v, n)	((v) |= (states) 1 << (n))
+#define	ISSET(v, n)	((v) & ((states) 1 << (n)))
+#define	ASSIGN(d, s)	((d) = (s))
+#define	EQ(a, b)	((a) == (b))
+#define	STATEVARS	int dummy	/* dummy version */
+#define	STATESETUP(m, n)	/* nothing */
+#define	STATETEARDOWN(m)	/* nothing */
+#define	SETUP(v)	((v) = 0)
+#define	onestate	long			/* Changed from int by Monty */
+#define	INIT(o, n)	((o) = (unsigned states)1 << (n))
+#define	INC(o)	((o) <<= 1)
+#define	ISSTATEIN(v, o)	((v) & (o))
+/* some abbreviations; note that some of these know variable names! */
+/* do "if I'm here, I can also be there" etc without branches */
+#define	FWD(dst, src, n)	((dst) |= ((unsigned states)(src)&(here)) << (n))
+#define	BACK(dst, src, n)	((dst) |= ((unsigned states)(src)&(here)) >> (n))
+#define	ISSETBACK(v, n)	((v) & ((unsigned states)here >> (n)))
+/* function names */
+#define SNAMES			/* engine.c looks after details */
+
+#include "engine.c"
+
+/* now undo things */
+#undef	states
+#undef	CLEAR
+#undef	SET0
+#undef	SET1
+#undef	ISSET
+#undef	ASSIGN
+#undef	EQ
+#undef	STATEVARS
+#undef	STATESETUP
+#undef	STATETEARDOWN
+#undef	SETUP
+#undef	onestate
+#undef	INIT
+#undef	INC
+#undef	ISSTATEIN
+#undef	FWD
+#undef	BACK
+#undef	ISSETBACK
+#undef	SNAMES
+
+/* macros for manipulating states, large version */
+#define	states	char *
+#define	CLEAR(v)	memset(v, 0, m->g->nstates)
+#define	SET0(v, n)	((v)[n] = 0)
+#define	SET1(v, n)	((v)[n] = 1)
+#define	ISSET(v, n)	((v)[n])
+#define	ASSIGN(d, s)	memcpy(d, s, m->g->nstates)
+#define	EQ(a, b)	(memcmp(a, b, m->g->nstates) == 0)
+#define	STATEVARS	int vn; char *space
+#define	STATESETUP(m, nv)	{ (m)->space = malloc((nv)*(m)->g->nstates); \
+				if ((m)->space == NULL) return(REG_ESPACE); \
+				(m)->vn = 0; }
+#define	STATETEARDOWN(m)	{ free((m)->space); }
+#define	SETUP(v)	((v) = &m->space[m->vn++ * m->g->nstates])
+#define	onestate	int
+#define	INIT(o, n)	((o) = (n))
+#define	INC(o)	((o)++)
+#define	ISSTATEIN(v, o)	((v)[o])
+/* some abbreviations; note that some of these know variable names! */
+/* do "if I'm here, I can also be there" etc without branches */
+#define	FWD(dst, src, n)	((dst)[here+(n)] |= (src)[here])
+#define	BACK(dst, src, n)	((dst)[here-(n)] |= (src)[here])
+#define	ISSETBACK(v, n)	((v)[here - (n)])
+/* function names */
+#define	LNAMES			/* flag */
+
+#include "engine.c"
+
+/*
+ - regexec - interface for matching
+ = extern int regexec(const regex_t *, const char *, size_t, \
+ =					regmatch_t [], int);
+ = #define	REG_NOTBOL	00001
+ = #define	REG_NOTEOL	00002
+ = #define	REG_STARTEND	00004
+ = #define	REG_TRACE	00400	// tracing of execution
+ = #define	REG_LARGE	01000	// force large representation
+ = #define	REG_BACKR	02000	// force use of backref code
+ *
+ * We put this here so we can exploit knowledge of the state representation
+ * when choosing which matcher to call.  Also, by this point the matchers
+ * have been prototyped.
+ */
+int				/* 0 success, REG_NOMATCH failure */
+my_regexec(preg, str, nmatch, pmatch, eflags)
+const my_regex_t *preg;
+const char *str;
+size_t nmatch;
+my_regmatch_t pmatch[];
+int eflags;
+{
+	register struct re_guts *g = preg->re_g;
+#ifdef REDEBUG
+#	define	GOODFLAGS(f)	(f)
+#else
+#	define	GOODFLAGS(f)	((f)&(REG_NOTBOL|REG_NOTEOL|REG_STARTEND))
+#endif
+
+	if (preg->re_magic != MAGIC1 || g->magic != MAGIC2)
+		return(REG_BADPAT);
+	assert(!(g->iflags&BAD));
+	if (g->iflags&BAD)		/* backstop for no-debug case */
+		return(REG_BADPAT);
+	eflags = GOODFLAGS(eflags);
+
+	if ((size_t) g->nstates <= CHAR_BIT*sizeof(states1) &&
+	    !(eflags&REG_LARGE))
+		return(smatcher(preg->charset, g, (char *)str, nmatch, pmatch, eflags));
+	else
+		return(lmatcher(preg->charset, g, (char *)str, nmatch, pmatch, eflags));
+}

externals/mysql/extlib/regex/regfree.c

@@ -0,0 +1,38 @@
+#include <my_global.h>
+#include <sys/types.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include "my_regex.h"
+
+#include "utils.h"
+#include "regex2.h"
+
+/*
+ - regfree - free everything
+ = extern void regfree(regex_t *);
+ */
+void
+my_regfree(preg)
+my_regex_t *preg;
+{
+	register struct re_guts *g;
+
+	if (preg->re_magic != MAGIC1)	/* oops */
+		return;			/* nice to complain, but hard */
+
+	g = preg->re_g;
+	if (g == NULL || g->magic != MAGIC2)	/* oops again */
+		return;
+	preg->re_magic = 0;		/* mark it invalid */
+	g->magic = 0;			/* mark it invalid */
+
+	if (g->strip != NULL)
+		free((char *)g->strip);
+	if (g->sets != NULL)
+		free((char *)g->sets);
+	if (g->setbits != NULL)
+		free((char *)g->setbits);
+	if (g->must != NULL)
+		free(g->must);
+	free((char *)g);
+}

externals/mysql/extlib/regex/reginit.c

@@ -0,0 +1,81 @@
+/* Init cclasses array from ctypes */
+
+#include <my_global.h>
+#include <m_ctype.h>
+#include <m_string.h>
+#include "cclass.h"
+
+static my_bool regex_inited=0;
+
+void my_regex_init(CHARSET_INFO *cs)
+{
+  char buff[CCLASS_LAST][256];
+  int  count[CCLASS_LAST];
+  uint i;
+
+  if (!regex_inited)
+  {
+    regex_inited=1;
+    bzero((uchar*) &count,sizeof(count));
+
+    for (i=1 ; i<= 255; i++)
+    {
+      if (my_isalnum(cs,i))
+	buff[CCLASS_ALNUM][count[CCLASS_ALNUM]++]=(char) i;
+      if (my_isalpha(cs,i))
+	buff[CCLASS_ALPHA][count[CCLASS_ALPHA]++]=(char) i;
+      if (my_iscntrl(cs,i))
+	buff[CCLASS_CNTRL][count[CCLASS_CNTRL]++]=(char) i;
+      if (my_isdigit(cs,i))
+	buff[CCLASS_DIGIT][count[CCLASS_DIGIT]++]=(char) i;
+      if (my_isgraph(cs,i))
+	buff[CCLASS_GRAPH][count[CCLASS_GRAPH]++]=(char) i;
+      if (my_islower(cs,i))
+	buff[CCLASS_LOWER][count[CCLASS_LOWER]++]=(char) i;
+      if (my_isprint(cs,i))
+	buff[CCLASS_PRINT][count[CCLASS_PRINT]++]=(char) i;
+      if (my_ispunct(cs,i))
+	buff[CCLASS_PUNCT][count[CCLASS_PUNCT]++]=(char) i;
+      if (my_isspace(cs,i))
+	buff[CCLASS_SPACE][count[CCLASS_SPACE]++]=(char) i;
+      if (my_isupper(cs,i))
+	buff[CCLASS_UPPER][count[CCLASS_UPPER]++]=(char) i;
+      if (my_isxdigit(cs,i))
+	buff[CCLASS_XDIGIT][count[CCLASS_XDIGIT]++]=(char) i;
+    }
+    buff[CCLASS_BLANK][0]=' ';
+    buff[CCLASS_BLANK][1]='\t';
+    count[CCLASS_BLANK]=2;
+    for (i=0; i < CCLASS_LAST ; i++)
+    {
+      char *tmp=(char*) malloc(count[i]+1);
+      if (!tmp)
+      {
+	/*
+	  This is very unlikely to happen as this function is called once
+	  at program startup
+	*/
+	fprintf(stderr,
+		"Fatal error: Can't allocate memory in regex_init\n");
+	exit(1);
+      }
+      memcpy(tmp,buff[i],count[i]*sizeof(char));
+      tmp[count[i]]=0;
+      cclasses[i].chars=tmp;
+    }
+  }
+  return;
+}
+
+void my_regex_end()
+{
+  if (regex_inited)
+  {
+    int i;
+    for (i=0; i < CCLASS_LAST ; i++)
+      free((char*) cclasses[i].chars);
+    regex_inited=0;
+  }
+}
+
+

externals/mysql/extlib/regex/split.c

@@ -0,0 +1,316 @@
+#include <stdio.h>
+#include <string.h>
+
+/*
+ - split - divide a string into fields, like awk split()
+ = int split(char *string, char *fields[], int nfields, char *sep);
+ */
+int				/* number of fields, including overflow */
+split(string, fields, nfields, sep)
+char *string;
+char *fields[];			/* list is not NULL-terminated */
+int nfields;			/* number of entries available in fields[] */
+char *sep;			/* "" white, "c" single char, "ab" [ab]+ */
+{
+	register char *p = string;
+	register char c;			/* latest character */
+	register char sepc = sep[0];
+	register char sepc2;
+	register int fn;
+	register char **fp = fields;
+	register char *sepp;
+	register int trimtrail;
+
+	/* white space */
+	if (sepc == '\0') {
+		while ((c = *p++) == ' ' || c == '\t')
+			continue;
+		p--;
+		trimtrail = 1;
+		sep = (char*) " \t";	/* note, code below knows this is 2 long */
+		sepc = ' ';
+	} else
+		trimtrail = 0;
+	sepc2 = sep[1];		/* now we can safely pick this up */
+
+	/* catch empties */
+	if (*p == '\0')
+		return(0);
+
+	/* single separator */
+	if (sepc2 == '\0') {
+		fn = nfields;
+		for (;;) {
+			*fp++ = p;
+			fn--;
+			if (fn == 0)
+				break;
+			while ((c = *p++) != sepc)
+				if (c == '\0')
+					return(nfields - fn);
+			*(p-1) = '\0';
+		}
+		/* we have overflowed the fields vector -- just count them */
+		fn = nfields;
+		for (;;) {
+			while ((c = *p++) != sepc)
+				if (c == '\0')
+					return(fn);
+			fn++;
+		}
+		/* not reached */
+	}
+
+	/* two separators */
+	if (sep[2] == '\0') {
+		fn = nfields;
+		for (;;) {
+			*fp++ = p;
+			fn--;
+			while ((c = *p++) != sepc && c != sepc2)
+				if (c == '\0') {
+					if (trimtrail && **(fp-1) == '\0')
+						fn++;
+					return(nfields - fn);
+				}
+			if (fn == 0)
+				break;
+			*(p-1) = '\0';
+			while ((c = *p++) == sepc || c == sepc2)
+				continue;
+			p--;
+		}
+		/* we have overflowed the fields vector -- just count them */
+		fn = nfields;
+		while (c != '\0') {
+			while ((c = *p++) == sepc || c == sepc2)
+				continue;
+			p--;
+			fn++;
+			while ((c = *p++) != '\0' && c != sepc && c != sepc2)
+				continue;
+		}
+		/* might have to trim trailing white space */
+		if (trimtrail) {
+			p--;
+			while ((c = *--p) == sepc || c == sepc2)
+				continue;
+			p++;
+			if (*p != '\0') {
+				if (fn == nfields+1)
+					*p = '\0';
+				fn--;
+			}
+		}
+		return(fn);
+	}
+
+	/* n separators */
+	fn = 0;
+	for (;;) {
+		if (fn < nfields)
+			*fp++ = p;
+		fn++;
+		for (;;) {
+			c = *p++;
+			if (c == '\0')
+				return(fn);
+			sepp = sep;
+			while ((sepc = *sepp++) != '\0' && sepc != c)
+				continue;
+			if (sepc != '\0')	/* it was a separator */
+				break;
+		}
+		if (fn < nfields)
+			*(p-1) = '\0';
+		for (;;) {
+			c = *p++;
+			sepp = sep;
+			while ((sepc = *sepp++) != '\0' && sepc != c)
+				continue;
+			if (sepc == '\0')	/* it wasn't a separator */
+				break;
+		}
+		p--;
+	}
+
+	/* not reached */
+}
+
+#ifdef TEST_SPLIT
+
+
+/*
+ * test program
+ * pgm		runs regression
+ * pgm sep	splits stdin lines by sep
+ * pgm str sep	splits str by sep
+ * pgm str sep n	splits str by sep n times
+ */
+int
+main(argc, argv)
+int argc;
+char *argv[];
+{
+	char buf[512];
+	register int n;
+#	define	MNF	10
+	char *fields[MNF];
+
+	if (argc > 4)
+		for (n = atoi(argv[3]); n > 0; n--) {
+			(void) strcpy(buf, argv[1]);
+		}
+	else if (argc > 3)
+		for (n = atoi(argv[3]); n > 0; n--) {
+			(void) strcpy(buf, argv[1]);
+			(void) split(buf, fields, MNF, argv[2]);
+		}
+	else if (argc > 2)
+		dosplit(argv[1], argv[2]);
+	else if (argc > 1)
+		while (fgets(buf, sizeof(buf), stdin) != NULL) {
+			buf[strlen(buf)-1] = '\0';	/* stomp newline */
+			dosplit(buf, argv[1]);
+		}
+	else
+		regress();
+
+	exit(0);
+}
+
+dosplit(string, seps)
+char *string;
+char *seps;
+{
+#	define	NF	5
+	char *fields[NF];
+	register int nf;
+
+	nf = split(string, fields, NF, seps);
+	print(nf, NF, fields);
+}
+
+print(nf, nfp, fields)
+int nf;
+int nfp;
+char *fields[];
+{
+	register int fn;
+	register int bound;
+
+	bound = (nf > nfp) ? nfp : nf;
+	printf("%d:\t", nf);
+	for (fn = 0; fn < bound; fn++)
+		printf("\"%s\"%s", fields[fn], (fn+1 < nf) ? ", " : "\n");
+}
+
+#define	RNF	5		/* some table entries know this */
+struct {
+	char *str;
+	char *seps;
+	int nf;
+	char *fi[RNF];
+} tests[] = {
+	"",		" ",	0,	{ "" },
+	" ",		" ",	2,	{ "", "" },
+	"x",		" ",	1,	{ "x" },
+	"xy",		" ",	1,	{ "xy" },
+	"x y",		" ",	2,	{ "x", "y" },
+	"abc def  g ",	" ",	5,	{ "abc", "def", "", "g", "" },
+	"  a bcd",	" ",	4,	{ "", "", "a", "bcd" },
+	"a b c d e f",	" ",	6,	{ "a", "b", "c", "d", "e f" },
+	" a b c d ",	" ",	6,	{ "", "a", "b", "c", "d " },
+
+	"",		" _",	0,	{ "" },
+	" ",		" _",	2,	{ "", "" },
+	"x",		" _",	1,	{ "x" },
+	"x y",		" _",	2,	{ "x", "y" },
+	"ab _ cd",	" _",	2,	{ "ab", "cd" },
+	" a_b  c ",	" _",	5,	{ "", "a", "b", "c", "" },
+	"a b c_d e f",	" _",	6,	{ "a", "b", "c", "d", "e f" },
+	" a b c d ",	" _",	6,	{ "", "a", "b", "c", "d " },
+
+	"",		" _~",	0,	{ "" },
+	" ",		" _~",	2,	{ "", "" },
+	"x",		" _~",	1,	{ "x" },
+	"x y",		" _~",	2,	{ "x", "y" },
+	"ab _~ cd",	" _~",	2,	{ "ab", "cd" },
+	" a_b  c~",	" _~",	5,	{ "", "a", "b", "c", "" },
+	"a b_c d~e f",	" _~",	6,	{ "a", "b", "c", "d", "e f" },
+	"~a b c d ",	" _~",	6,	{ "", "a", "b", "c", "d " },
+
+	"",		" _~-",	0,	{ "" },
+	" ",		" _~-",	2,	{ "", "" },
+	"x",		" _~-",	1,	{ "x" },
+	"x y",		" _~-",	2,	{ "x", "y" },
+	"ab _~- cd",	" _~-",	2,	{ "ab", "cd" },
+	" a_b  c~",	" _~-",	5,	{ "", "a", "b", "c", "" },
+	"a b_c-d~e f",	" _~-",	6,	{ "a", "b", "c", "d", "e f" },
+	"~a-b c d ",	" _~-",	6,	{ "", "a", "b", "c", "d " },
+
+	"",		"  ",	0,	{ "" },
+	" ",		"  ",	2,	{ "", "" },
+	"x",		"  ",	1,	{ "x" },
+	"xy",		"  ",	1,	{ "xy" },
+	"x y",		"  ",	2,	{ "x", "y" },
+	"abc def  g ",	"  ",	4,	{ "abc", "def", "g", "" },
+	"  a bcd",	"  ",	3,	{ "", "a", "bcd" },
+	"a b c d e f",	"  ",	6,	{ "a", "b", "c", "d", "e f" },
+	" a b c d ",	"  ",	6,	{ "", "a", "b", "c", "d " },
+
+	"",		"",	0,	{ "" },
+	" ",		"",	0,	{ "" },
+	"x",		"",	1,	{ "x" },
+	"xy",		"",	1,	{ "xy" },
+	"x y",		"",	2,	{ "x", "y" },
+	"abc def  g ",	"",	3,	{ "abc", "def", "g" },
+	"\t a bcd",	"",	2,	{ "a", "bcd" },
+	"  a \tb\t c ",	"",	3,	{ "a", "b", "c" },
+	"a b c d e ",	"",	5,	{ "a", "b", "c", "d", "e" },
+	"a b\tc d e f",	"",	6,	{ "a", "b", "c", "d", "e f" },
+	" a b c d e f ",	"",	6,	{ "a", "b", "c", "d", "e f " },
+
+	NULL,		NULL,	0,	{ NULL },
+};
+
+regress()
+{
+	char buf[512];
+	register int n;
+	char *fields[RNF+1];
+	register int nf;
+	register int i;
+	register int printit;
+	register char *f;
+
+	for (n = 0; tests[n].str != NULL; n++) {
+		(void) strcpy(buf, tests[n].str);
+		fields[RNF] = NULL;
+		nf = split(buf, fields, RNF, tests[n].seps);
+		printit = 0;
+		if (nf != tests[n].nf) {
+			printf("split `%s' by `%s' gave %d fields, not %d\n",
+				tests[n].str, tests[n].seps, nf, tests[n].nf);
+			printit = 1;
+		} else if (fields[RNF] != NULL) {
+			printf("split() went beyond array end\n");
+			printit = 1;
+		} else {
+			for (i = 0; i < nf && i < RNF; i++) {
+				f = fields[i];
+				if (f == NULL)
+					f = "(NULL)";
+				if (strcmp(f, tests[n].fi[i]) != 0) {
+					printf("split `%s' by `%s', field %d is `%s', not `%s'\n",
+						tests[n].str, tests[n].seps,
+						i, fields[i], tests[n].fi[i]);
+					printit = 1;
+				}
+			}
+		}
+		if (printit)
+			print(nf, RNF, fields);
+	}
+}
+#endif

externals/mysql/extlib/regex/utils.h

@@ -0,0 +1,22 @@
+/* utility definitions */
+#ifdef _POSIX2_RE_DUP_MAX
+#define	DUPMAX		_POSIX2_RE_DUP_MAX	/* xxx is this right? */
+#else
+#define DUPMAX		255
+#endif
+#define	RE_INFINITY	(DUPMAX + 1)
+#define	NC		(CHAR_MAX - CHAR_MIN + 1)
+typedef unsigned char uch;
+
+/* switch off assertions (if not already off) if no REDEBUG */
+#ifndef REDEBUG
+#ifndef NDEBUG
+#define	NDEBUG	/* no assertions please */
+#endif
+#endif
+#include <assert.h>
+
+/* for old systems with bcopy() but no memmove() */
+#ifdef USEBCOPY
+#define	memmove(d, s, c)	bcopy(s, d, c)
+#endif