This marker designates the full name from the beginning up to the marker as the Major. The comma character is an example of this type. For example, the name:

SMITH, JOHN WILLIAM
CASEY JONES, HENRY

are reordered to:

JOHN WILLIAM SMITH
HENRY CASEY JONES

This marker designates the full name from the marker to the end of the name as the Major. This marker causes no name reordering. For example, the % character in the names:

JOHN WILLIAM %SMITH
HENRY %CASEY JONES

This marker designates the part of the name from the marker until the matching closing marker as the Major. This marker designates two characters (as the leading and trailing delimiter). For example, the () characters in the name:

JOHN WILLIAM (SMITH)

These markers are user defined in the Edit-list. Refer to the Edit List chapter of the

DEFINITION and CUSTOMIZATION GUIDE FOR SSA-NAME3 SERVICE GROUPS

.

If the name contains more than one Major Marker then only the first one is processed. If a marker designates an empty string then it is ignored (and removed) from the name.

Delete Between Markers allows all of the text between two markers to be deleted. For example, if

()

were defined as ’Delete Between’ markers, the name:

ALPHA PROCESSING CO (DISTRIBUTION)

would become;

ALPHA PROCESSING CO

Delete Before/After Markers allows all of the text before or after a marker to be deleted and optionally replaced with another word or phrase. For example, if ’SEE DOC’ was defined as a ’Delete After’ marker, the name:

ALPHA PROCESSING CO SEE DOC NO 36541

would become;

ALPHA PROCESSING CO

Cleaning editing is the next phase, again driven by Edit-list rules. This phase allows simple userdefined string/character replacements to be put into effect before the normal cleaning rules are processed.

For example, with no cleaning editing, and using the tables as defined in the Fast-start, the string,

ME T/A YOU

cleans to

ME T A YOU

because the / character is defined as a delimiter and is removed. Because cleaning editing is invoked before the normal cleaning rules it can be used to trap such events. For example adding the following rule to the Edit-list definition file,

*S >T/A<
*W >TRADING AS<

will trap the

T/A

and replace it with

TRADING AS

.

Cleaning Editing processes the rules in order from longest to shortest. For example, BV and BVBA are two company legal endings in the Netherlands (similar to

INC.

in the USA). Defining the following Cleaning Editing rules in the Edit-list:

*S >B.V.<BA
*W >BV<
*S >B.V.B.A.<BA
*W >BVBA<

means that the name, WORLDGROUP HOLDINGS B.V.B.A. would be correctly translated to WORLDGROUP HOLDINGS BVBA. If Cleaning Editing processed rules in order from shortest to longest, it would have become WORLDGROUP HOLDINGS BV B A (which is not what was required) because the B.V. rule would have been processed first and the remaining characters not recognized by the other rule.

Comma characters are removed from the input name. If the name format is ’R’ (i.e. the major name is at the right of the name and the Algorithm definition has

NAME-FORMAT=R

) and if the name contains a comma then the part of the name prior to the comma is considered to be the last name. It is then moved to the end of the name (e.g. SMITH, JOHN is cleaned into JOHN SMITH). If the comma has no words before it then it is ignored. After the first comma is processed then all other commas are treated as delimiters. The comma (,) is of this type but some systems may treat the slash (/) character for names like SMITH/JOHN B. This comma processing can be turned off by setting

CLEANING-OPTIONS #2

in the Algorithm Definition.

Delimiters are removed from the input name. If a delimiter is embedded in a word then the word is broken at that position as if there was a blank (e.g. VAN-DAM is cleaned into VAN DAM). Most special symbols are delimiters, as well as the blank character.

The Token is a special character which causes cleaning to add a blank both before and after the Token character. e.g. If the character " is defined as a Token, and the name "HELLO" was to be cleaned, the output name would be <blank>"<blank>HELLO<blank"<blank>

The Self type is not converted. It is a shorthand notation to express the fact the character is to be kept as itself.

All other characters are replaced with the value in the table (e.g. ’e’ is replaced with ’E’). Alphabetic characters and the numerals are of this type, as well as accented characters.