Oddity in regular expressions

[Andrew Koenig]

Here's a program:

       BEGIN {
           s = "[::]"
           print "Before: " s
           gsub(/\[/, "\\\\&\\f(CW", s)
           print "After: " s
       }

Running under GNU Awk 3.0.6 yields this output:

       Before: [::]
       After: \[\f(CW::]

Running under AT&T awk yields this output

       Before: [::]
       After: \&\f(CW::]

The problem appears to be that GNU Awk is treating the & as
a regular-expression replacement character even though it is
preceded by a backslash.

I can see the point -- it view the replacement as \ \ & \ f ( CW
and treats the \ \ as an escaped backslash.  Nevertheless, it's
an incompatibility with the original.

Is this an intentional incompatibility, or a bug?

-- 
Andrew Koenig, address@hidden, http://www.research.att.com/info/ark
 -=- MIME -=- 
This is a MIME-encapsulated message.

--308004CE05.987171428/mail-blue.research.att.com
Content-Description: Notification
Content-Type: text/plain

This is the Postfix program at host mail-blue.research.att.com.

I'm sorry to have to inform you that the message returned
below could not be delivered to one or more destinations.

For further assistance, please contact <address@hidden>

If you do so, please include this problem report. You can
delete your own text from the message returned below.

                        The Postfix program

<address@hidden>: Name service error for domain gnu.org: Host found but no
    data record of requested type


--308004CE05.987171428/mail-blue.research.att.com
Content-Description: Undelivered Message
Content-Type: message/rfc822

Received: from europa.research.att.com (europa.research.att.com 
[135.207.23.117])
        by mail-blue.research.att.com (Postfix) with ESMTP id 308004CE05
        for <address@hidden>; Fri, 13 Apr 2001 10:17:08 -0400 (EDT)
Received: (from address@hidden)
        by europa.research.att.com (8.8.8/8.8.7) id KAA02813;
        Fri, 13 Apr 2001 10:17:07 -0400 (EDT)
Date: Fri, 13 Apr 2001 10:17:07 -0400 (EDT)
Message-Id: <address@hidden>
X-Authentication-Warning: europa.research.att.com: ark set sender to 
address@hidden using -f
From: Andrew Koenig <address@hidden>
To: address@hidden
Subject: Oddity in regular expressions

Here's a program:

       BEGIN {
           s = "[::]"
           print "Before: " s
           gsub(/\[/, "\\\\&\\f(CW", s)
           print "After: " s
       }

Running under GNU Awk 3.0.6 yields this output:

       Before: [::]
       After: \[\f(CW::]

Running under AT&T awk yields this output

       Before: [::]
       After: \&\f(CW::]

The problem appears to be that GNU Awk is treating the & as
a regular-expression replacement character even though it is
preceded by a backslash.

I can see the point -- it view the replacement as \ \ & \ f ( CW
and treats the \ \ as an escaped backslash.  Nevertheless, it's
an incompatibility with the original.

Is this an intentional incompatibility, or a bug?

-- 
Andrew Koenig, address@hidden, http://www.research.att.com/info/ark

--308004CE05.987171428/mail-blue.research.att.com--