From mboxrd@z Thu Jan  1 00:00:00 1970
X-Msuck: nntp://news.gmane.org/gmane.linux.lib.musl.general/7123
Path: news.gmane.org!not-for-mail
From: Sergey Dmitrouk <sdmitrouk@accesssoftek.com>
Newsgroups: gmane.linux.lib.musl.general
Subject: [PATCH] conforming strverscmp() implementation
Date: Tue, 3 Mar 2015 12:45:07 +0200
Message-ID: <20150303104507.GA5094@zx-spectrum>
Reply-To: musl@lists.openwall.com
NNTP-Posting-Host: plane.gmane.org
Mime-Version: 1.0
Content-Type: multipart/mixed; boundary="uAKRQypu60I7Lcqm"
X-Trace: ger.gmane.org 1425379511 1359 80.91.229.3 (3 Mar 2015 10:45:11 GMT)
X-Complaints-To: usenet@ger.gmane.org
NNTP-Posting-Date: Tue, 3 Mar 2015 10:45:11 +0000 (UTC)
To: <musl@lists.openwall.com>
Original-X-From: musl-return-7136-gllmg-musl=m.gmane.org@lists.openwall.com Tue Mar 03 11:45:11 2015
Return-path: <musl-return-7136-gllmg-musl=m.gmane.org@lists.openwall.com>
Envelope-to: gllmg-musl@m.gmane.org
Original-Received: from mother.openwall.net ([195.42.179.200])
	by plane.gmane.org with smtp (Exim 4.69)
	(envelope-from <musl-return-7136-gllmg-musl=m.gmane.org@lists.openwall.com>)
	id 1YSkK7-0002js-B1
	for gllmg-musl@m.gmane.org; Tue, 03 Mar 2015 11:45:11 +0100
Original-Received: (qmail 7877 invoked by uid 550); 3 Mar 2015 10:45:09 -0000
Mailing-List: contact musl-help@lists.openwall.com; run by ezmlm
Precedence: bulk
List-Post: <mailto:musl@lists.openwall.com>
List-Help: <mailto:musl-help@lists.openwall.com>
List-Unsubscribe: <mailto:musl-unsubscribe@lists.openwall.com>
List-Subscribe: <mailto:musl-subscribe@lists.openwall.com>
Original-Received: (qmail 7839 invoked from network); 3 Mar 2015 10:45:05 -0000
Content-Disposition: inline
Xref: news.gmane.org gmane.linux.lib.musl.general:7123
Archived-At: <http://permalink.gmane.org/gmane.linux.lib.musl.general/7123>

--uAKRQypu60I7Lcqm
Content-Type: text/plain; charset="us-ascii"
Content-Disposition: inline

Hi,

Attached is strverscmp() that should conform to what is described in
manual pages.  It's not actually a diff as it would be harder to read
in this case.

The idea is just as described in manuals (once you get what they mean):

1. Find point where strings differ as in current code, but keep track
   of locations of last seen numbers.
2. If numbers are both integers (that is they don't start with zero or
   it's a single zero) compare as in current implementation.
3. If only one of numbers is integer, the other one is smaller.
4. Otherwise, number with longer leading sequence of zeroes is smaller.
5. If leading zeroes match, compare characters following last zero in
   both strings.

I didn't see much comments in musl, so there is none in this file, but
can add some if requested.

Implementation is admittedly more sophisticated than current version for
the reason that all simpler approaches I tried fail for at least one of
possible comparison cases.

Regards,
Sergey

--uAKRQypu60I7Lcqm
Content-Type: text/plain; charset="us-ascii"
Content-Disposition: attachment; filename="strverscmp.c"

#define _GNU_SOURCE
#include <ctype.h>
#include <string.h>

int strverscmp(const char *l, const char *r)
{
	const char *ln=(isdigit(*l) ? l : NULL), *rn=(isdigit(*r) ? r : NULL);
	while (*l==*r) {
		if (!*l) return 0;

		if (isdigit(*l)) {
			if (ln == NULL) {
				ln = l; rn = r;
			}
		} else {
			ln = NULL; rn = NULL;
		}

		l++; r++;
	}

	if ((*l != '\0' && !isdigit(*l)) || (*r != '\0' && !isdigit(*r))) {
		ln = NULL; rn = NULL;
	}

	if (ln != NULL) {
		int intl=(*ln != '0' || !isdigit(*(ln + 1)));
		int intr=(*rn != '0' || !isdigit(*(rn + 1)));

		if (intl ^ intr) {
			return intl ? 1 : -1;
		} else if (intl) {
			size_t lenl=0, lenr=0;
			while (isdigit(l[lenl])) lenl++;
			while (isdigit(r[lenr])) lenr++;
			if (lenl==lenr) {
				return (*l - *r);
			} else if (lenl>lenr) {
				return 1;
			} else {
				return -1;
			}
		} else {
			size_t zl=0, zr=0;
			while (ln[zl]=='0') zl++;
			while (rn[zr]=='0') zr++;
			if (zl>zr) {
				return -1;
			} else if (zl<zr) {
				return 1;
			}
		}
	}

	return (*l - *r);
}

--uAKRQypu60I7Lcqm--