Mailing-List: contact derby-dev-help@db.apache.org; run by ezmlm
Precedence: bulk
Reply-To: <derby-dev@db.apache.org>
Message-ID: <18696141.1181156846297.JavaMail.jira@brutus>
Date: Wed, 6 Jun 2007 12:07:26 -0700 (PDT)
From: "Mike Matrigali (JIRA)" <jira@apache.org>
To: derby-dev@db.apache.org
Subject: [jira] Updated: (DERBY-2377) Document language based ordering which
 will be implemented by code related sub-tasks of DERBY-1478.
In-Reply-To: <12099102.1172592065504.JavaMail.jira@brutus>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


     [ https://issues.apache.org/jira/browse/DERBY-2377?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Mike Matrigali updated DERBY-2377:
----------------------------------


Here is some updated text which I think would better fit at the top of the document.  I don't think we should try to describe the mechanics of the algorithm, just that we use the java support for this.  We are not providing support in this release for people to change these rules, we use the rules and the logic based on these rules that the JVM provides:

Character-based collation in Derby

A character set is a set of symbols and encodings.  Derby supports a wide
range of character sets and encodes all of them using the
Unicode support provided by the java.lang.Character
class in the JVM in which it runs. See JVM specific documentation for this
class for exact level of Unicode Standard supported.

A collation is a set of rules for comparing characters in a character set.
In Derby collation rules affect comparisons of the CHAR and VARCHAR data
type.  Collation rules also affect how the LIKE Boolean operator processes
CHAR, VARCHAR, CLOB, and LONG VARCHAR data.

The default Derby collation rule is based on the binary Unicode values of
the characters.  So a character is <, =, or > based on the numeric comparison
of the Unicode value.  This rule allows for very efficient comparison of
strings.

Derby also supports the ability to define collation rules based on those
appropriate to a territory, and is referred to as "territory based collation".
One can specifically set the territory of a database when it is created, or
it will default to the default territory of the JVM in which the database
is created.  Non default collation support is provided through the JVM's
java.text.RuleBasedCollator class and the JVM's set of rules associated with
the specified territory.   Details of how these rules are used to provide
territory specific collation can be seen by referring to the JVM specification.
Derby currently only supports running those rules that can be loaded dynamically
from the running JVM based on the territory attribute, user overrides to these
rules are not supported.  Territory based
collation will add extra procession overhead to all character based comparison
operations versus the default collation performance.

> Document language based ordering which will be implemented by code related sub-tasks of DERBY-1478.
> ---------------------------------------------------------------------------------------------------
>
>                 Key: DERBY-2377
>                 URL: https://issues.apache.org/jira/browse/DERBY-2377
>             Project: Derby
>          Issue Type: Sub-task
>          Components: Documentation
>    Affects Versions: 10.3.0.0
>            Reporter: Mamta A. Satoor
>            Assignee: Laura Stewart
>         Attachments: cdevcollation.html, cdevcollation.html, cdevcollation.html, cdevdvlp51654.html, rrefattribcollation.html, tdevdvlpcollation.html
>
>
> There is an effort going on to implement language based ordering. This will be implemented by various code-related subtasks of DERBY-1478. The functional spec is attached to DERBY-1478. This feature will need accompanying documentation changes and this Jira entry is to track that work.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.