MARC 21 Specifications for Record Structure, Character Sets, and Exchange Media

Summary List of MARC 21 Characters Assigned to the Private Use Area (PUA)

Contents

• Introduction
• Issues Related to PUA Characters
• Ancient Korean hangul characters
• Component characters used in RLIN's CJK input method

Introduction

This document brings together the remaining mappings for 61 MARC-8 East Asian Character Code (EACC) values that were mapped to Unicode character values in the Private Use Area (PUA). The other former PUA characters have been identified in Unicode and the mappings changed to regular Unicode encodings. (See Changes to Unicode Mappings for MARC 21 Characters formerly assigned to the Private Use Area.) Alternative non-PUA character mappings to the geta character are indicated in this document to facilitate implementation of the MARC character repertoire in systems where it is not possible or desirable to include characters from the Private Use Area.

The characters are divided into two groups. The groups include:

• 26 ancient Korean hangul characters, and
• 35 component characters used in RLIN's proprietary CJK input method.

These characters are all listed in the MARC 21 character set tables. They are also separately listed here for ease of identification of current PUA characters valid in MARC records.

Issues Related to PUA Characters

MARC-8 mappings to PUA values are provided for users that must have a high level of roundtrip compatibility of data being converted to UCS/Unicode and back to MARC-8, or where there is a need to retain the distinctiveness of any of the 61 characters in this group from other Unicode characters.

Implementers who choose to retain PUA characters in their UCS/Unicode data should note the following. Characters assigned to PUA values, by definition, are locally defined and not natively supported by off-the-shelf Unicode-compliant software and hardware. This can be problematic for display and printing of data since fonts may not include a meaningful glyph for any PUA character value. Character shapes and meanings are necessarily ambiguous for PUA character values, thus they usually display and print as some sort of fill character, if they are supported at all. No special normalization, folding, indexing, or retrieval capability can be assumed for PUA characters. Implementers should also recognize that once data has been converted from MARC-8 to UCS/Unicode using the alternative geta, is will not be possible to return to the original EACC values.

In the future, some MARC-8 characters currently mapped to the PUA may become mappable to non-PUA characters due to the addition of new East Asian characters. This list will be evaluated occasionally by experts to determine whether any of these characters have been added to Unicode.

Ancient Korean hangul characters

Mapped to geta character: 26

Component characters used in RLIN's CJK input method

These characters should not appear in communications data.

Count: 35

MARC 21 HOME >> Specifications >> Character Sets >> Part 5 > > East Asian Code Tables