Re: property, character, and sequence name loose matching

From: Asmus Freytag (
Date: Wed Mar 10 2010 - 15:02:18 CST

  • Next message: Mark Davis ☕: "Re: property, character, and sequence name loose matching"

    This whole discussion represents a fundamental misunderstanding of the
    matching rules.

    First, there are the naming rules for characters (etc.). These define
    how legal formal names are *created*. (Those rules are spelled out in an
    annex to 10646, for example).

    Then there are the matching rules. They impact legal formal names only
    insofar as accidental matches (under those rules )between otherwise
    legal namer are disallowed, therefore effectively disallowing a small
    slice of possible future names.

    But, the matching rules define something else: they allow certain
    modifications of names to fit the strictures of a given syntax (where
    spaces generally are not allowed, but - or _ may be allowed).

    Yes, you can create your own syntax, where a space is inserted after
    nearly every character, and if you do that correctly, character names in
    your modified syntax can still be matched against formal names. Whether
    that's advisable is another issue. What is useful, is the ability to
    have a well-defined match of
    LatinLetterCapitalA and latin_letter_capital_a to LATIN CAPITAL LETTER A.


    This archive was generated by hypermail 2.1.5 : Wed Mar 10 2010 - 15:05:19 CST