Unfortunately, that interpretation is not accepted by many systems,
including OCLC's.  Our interpretation, and the way our software is
implemented, is that once you have introduced another character set with
an ISO 2022 escape sequence such as ESC$1 (1B 34 21), you must return to
ASCII with the ISO 2022 sequence ESC(B (1B 28 42). The !B 73 sequence
would be treated as an error in that context.

Gary L. Smith
Software Architect
OCLC
[log in to unmask]

-----Original Message-----
From: MARC [mailto:[log in to unmask]] On Behalf Of Karen Coyle
Sent: Friday, March 23, 2012 04:49
To: [log in to unmask]
Subject: Re: {ASCII} within EACC string

To embed ASCII in the midst of an escaped character sequence, you need
to escape back to ASCII and then back again to the other character set.
From the documentation:

To redesignate ASCII, the following two-character escape sequence is
used:

    ESCs (ASCII 1B(hex) 73(hex)) for ASCII default character set


So this string should have a 1B 73 before the first ASCII character, and
then another 1B and the code for the character set that follows.
Basically, escape in/escape out for each change in character set. I
suspect this was translated from Unicode where the characters can
commingle freely.