Mailing-List: contact commits-help@cassandra.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@cassandra.apache.org
Date: Mon, 29 Apr 2013 07:48:19 +0000 (UTC)
From: =?utf-8?Q?Micha=C5=82_Michalski_=28JIRA=29?= <jira@apache.org>
To: commits@cassandra.apache.org
Message-ID: <JIRA.12509119.1307041943629.241086.1367221699561@arcas>
In-Reply-To: <JIRA.12509119.1307041943629@arcas>
References: <JIRA.12509119.1307041943629@arcas>
Subject: [jira] [Commented] (CASSANDRA-2737) CQL: support IF EXISTS
 extension for DROP commands (table, keyspace, index)
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable


    [ https://issues.apache.org/jira/browse/CASSANDRA-2737?page=3Dcom.atlas=
sian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=3D=
13644325#comment-13644325 ]=20

Micha=C5=82 Michalski commented on CASSANDRA-2737:
---------------------------------------------

Thanks for reply, Sylvain!

I definitely agree that my solution has some overhead and your is much simp=
ler, but one of the things I aimed for was to distinguish the "critical" ex=
ceptions from the ones that can be handled by such "IF (NOT) EXISTS" extens=
ions, so - in the end - it could be possible to let user know what really h=
appened and, maybe, display some descriptive NOTICE message instead of disp=
laying nothing (I was a bit "inspired" in how it's reported in PostgreSQL i=
n similar case). However, if you say it's not necessary, I'm fine with it :=
-)=20

Additionally, in most of the cases validation checking if KS/CF/Index exist=
s (or not) happens outside *Statement classes (for example for CFs it's don=
e in MigrationManager.announceColumnFamilyDrop()) and I didn't like the ide=
a of passing additional parameter and handling it there - I preferred this =
decision (if exception should be thrown or not) to be done by *Statement cl=
ass itself, basing on what was returned/thrown by MigrationManager.=20

One thing I missed when coding was the fact that for KS/CF, when reporting =
existing / inexistent KS/CF,  ConfigurationException is thrown (index-relat=
ed code throws InvalidRequestException, which looks like a little inconsist=
ency to me, by the way), so I could have reused it, instead of creating new=
 ones.

Anyway, I see your point :-) Tomorrow I'm leaving for three weeks and when =
I'm back I'll reimplement it taking your suggestions into account :-)
               =20
> CQL: support IF EXISTS extension for DROP commands (table, keyspace, inde=
x)
> -------------------------------------------------------------------------=
--
>
>                 Key: CASSANDRA-2737
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-2737
>             Project: Cassandra
>          Issue Type: New Feature
>    Affects Versions: 0.8.0
>            Reporter: Cathy Daw
>            Priority: Trivial
>              Labels: cql
>             Fix For: 2.0
>
>         Attachments: 2737-concept-v1.txt
>
>


--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrato=
rs
For more information on JIRA, see: http://www.atlassian.com/software/jira