Re: Bidirectional Algorithm Problems

From: schererm@us.ibm.com
Date: Wed Nov 17 1999 - 14:58:40 EST


i agree.
take this as a reordering example and ignore where the levels come from.
mark leisher may be right - maybe this should be made clearer in the tr.

markus

Dave <davidt@picsel.com> on 99-11-17 03:21:13

To: "Unicode List" <unicode@unicode.org>
cc:
Subject: Bidirectional Algorithm Problems

First of all, thanks Rick and Markus for taking the time to respond.

However, I'm still not clear on this:

> The example you cited should have no explicit embedding codes in it at
> all.

DID YOU SAY 'he said "car MEANS CAR"'? has embedding levels:
11111111111112222222224443333333333211

Forgive my stupidity if I'm missing something but surely an embedding
level greater than 2 is impossible without explicit embeddings or
overrides.

Aren't embedding levels only changed when resolving explicit levels and
directions in X1 to X9 and when resolving implicit levels in I1 and I2?
Am I right in assuming that if there are no explicit changes then changes
to levels can only occur in I1 and I2, and so the greatest change in
level is +2 maximum?

Thanks again,
Dave.



This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:56 EDT