Click here to Skip to main content
Click here to Skip to main content
Add your own
alternative version

Henry Spencer's Regexp Engine Revisited

, 2 Jul 2003
A small, Unicode-aware regular expression engine based on Henry Spencer's early work
/*
 * Simple timing program for regcomp().
 * Usage: timer ncomp nexec nsub
 *	or
 *	timer ncomp nexec nsub regexp string [ answer [ sub ] ]
 *
 * The second form is for timing repetitions of a single test case.
 * The first form's test data is a compiled-in copy of the "tests" file.
 * Ncomp, nexec, nsub are how many times to do each regcomp, regexec,
 * and regsub.  The way to time an operation individually is to do something
 * like "timer 1 50 1".
 */
#include <stdio.h>

struct try {
	char *re, *str, *ans, *src, *dst;
} tests[] = {
#include "timer.t.h"
{ NULL, NULL, NULL, NULL, NULL }
};

#include <regexp.h>

int errreport = 0;		/* Report errors via errseen? */
char *errseen = NULL;		/* Error message. */

char *progname;

/* ARGSUSED */
main(argc, argv)
int argc;
char *argv[];
{
	int ncomp, nexec, nsub;
	struct try one;
	char dummy[512];

	if (argc < 4) {
		ncomp = 1;
		nexec = 1;
		nsub = 1;
	} else {
		ncomp = atoi(argv[1]);
		nexec = atoi(argv[2]);
		nsub = atoi(argv[3]);
	}
	
	progname = argv[0];
	if (argc > 5) {
		one.re = argv[4];
		one.str = argv[5];
		if (argc > 6)
			one.ans = argv[6];
		else
			one.ans = "y";
		if (argc > 7) {	
			one.src = argv[7];
			one.dst = "xxx";
		} else {
			one.src = "x";
			one.dst = "x";
		}
		errreport = 1;
		try(one, ncomp, nexec, nsub);
	} else
		multiple(ncomp, nexec, nsub);
	exit(0);
}

void
regerror(s)
char *s;
{
	if (errreport)
		errseen = s;
	else
		error(s, "");
}

#ifndef ERRAVAIL
error(s1, s2)
char *s1;
char *s2;
{
	fprintf(stderr, "regexp: ");
	fprintf(stderr, s1, s2);
	fprintf(stderr, "\n");
	exit(1);
}
#endif

int lineno = 0;

multiple(ncomp, nexec, nsub)
int ncomp, nexec, nsub;
{
	register int i;
	extern char *strchr();

	errreport = 1;
	for (i = 0; tests[i].re != NULL; i++) {
		lineno++;
		try(tests[i], ncomp, nexec, nsub);
	}
}

try(fields, ncomp, nexec, nsub)
struct try fields;
int ncomp, nexec, nsub;
{
	regexp *r;
	char dbuf[BUFSIZ];
	register int i;

	errseen = NULL;
	r = regcomp(fields.re);
	if (r == NULL) {
		if (*fields.ans != 'c')
			complain("regcomp failure in `%s'", fields.re);
		return;
	}
	if (*fields.ans == 'c') {
		complain("unexpected regcomp success in `%s'", fields.re);
		free((char *)r);
		return;
	}
	for (i = ncomp-1; i > 0; i--) {
		free((char *)r);
		r = regcomp(fields.re);
	}
	if (!regexec(r, fields.str)) {
		if (*fields.ans != 'n')
			complain("regexec failure in `%s'", "");
		free((char *)r);
		return;
	}
	if (*fields.ans == 'n') {
		complain("unexpected regexec success", "");
		free((char *)r);
		return;
	}
	for (i = nexec-1; i > 0; i--)
		(void) regexec(r, fields.str);
	errseen = NULL;
	for (i = nsub; i > 0; i--)
		regsub(r, fields.src, dbuf);
	if (errseen != NULL) {	
		complain("regsub complaint", "");
		free((char *)r);
		return;
	}
	if (strcmp(dbuf, fields.dst) != 0)
		complain("regsub result `%s' wrong", dbuf);
	free((char *)r);
}

complain(s1, s2)
char *s1;
char *s2;
{
	fprintf(stderr, "try: %d: ", lineno);
	fprintf(stderr, s1, s2);
	fprintf(stderr, " (%s)\n", (errseen != NULL) ? errseen : "");
}

By viewing downloads associated with this article you agree to the Terms of Service and the article's licence.

If a file you wish to view isn't highlighted, and is a text file (not binary), please let us know and we'll add colourisation support for it.

License

This article has no explicit license attached to it but may contain usage terms in the article text or the download files themselves. If in doubt please contact the author via the discussion board below.

A list of licenses authors might use can be found here

Share

About the Author

Sunspot
Web Developer
Canada Canada
I'm a senior software developer, working at Silanis Technology (http://www.silanis.com). I've acquired quite a bit of experience (usually the hard way!) in Win32 and raw COM programming on the job. In my spare time, I like to monkey around with POSIX code.
 
I'm mostly interested in portable C++ libraries. I'm happiest when I develop portable C++ code--C++ being such a powerful language as long as one keeps clear of the rather nasty subtleties of the language.
 
I hope the articles I contribute will be of some help to someone. If even one person gains a few hours through use of that code, I'll be very happy.
 
When not coding, I like to listen to Anime and try to learn Japanese. It's not working too well so far, unfortunately. :{)

| Advertise | Privacy | Mobile
Web02 | 2.8.140827.1 | Last Updated 3 Jul 2003
Article Copyright 2003 by Sunspot
Everything else Copyright © CodeProject, 1999-2014
Terms of Service
Layout: fixed | fluid