Mailing-List: contact user-help@cassandra.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@cassandra.apache.org
Received-SPF: pass (nike.apache.org: domain of jbellis@gmail.com designates
 74.125.82.172 as permitted sender)
DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=gmail.com; s=gamma;
        h=mime-version:in-reply-to:references:date:message-id:subject:from:to
         :content-type:content-transfer-encoding;
        b=fmjPaHgqWDRbtqafTFwVXcHeWG/afkQaDd9Xo3yRYejqhn40WcpSRTs96ddQK2yO3t
         SQIDFS3wPKhHa3gN2EPDmxqur3c+2qjCL0IUIsnRzB12yjPL7b8W9c4zqND68yxvRL7C
         L983k8tpbQXQsWxc5qmvkwQiLnSMHb15+BGqI=
MIME-Version: 1.0
In-Reply-To: <AANLkTikJH8tTb780w1QYHid14NsNLB2FCKwJBpbusdmG@mail.gmail.com>
References: <AANLkTik7Mrgs+bseNKBh-WYc7Zu2ycOAS287V-ZxecVy@mail.gmail.com>
	<AANLkTi=ZZeLZ4pGWhpmFKJytxqC=MYJdjcQAzf74z7gh@mail.gmail.com>
	<AANLkTim7DQree2X-M24Wib_UuckBLPO7D71NEXUqOzSP@mail.gmail.com>
	<AANLkTin28Z7N2eFTnLzNwseziY-q=0KTn79qLriK7Fp=@mail.gmail.com>
	<AANLkTik7Mo=86EdwNiDqNXuZB+xxZ2C5eZKF9dtrMVcc@mail.gmail.com>
	<AANLkTikJH8tTb780w1QYHid14NsNLB2FCKwJBpbusdmG@mail.gmail.com>
Date: Fri, 8 Oct 2010 12:21:23 -0500
Message-ID: <AANLkTinNbqPCb9M_kYzi1geG17CJRXnoFMdt2U4u7Fm6@mail.gmail.com>
Subject: Re: Creating and using indices
From: Jonathan Ellis <jbellis@gmail.com>
To: user <user@cassandra.apache.org>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

On Fri, Oct 8, 2010 at 9:27 AM, Christian Decker
<decker.christian@gmail.com> wrote:
> On Thu, Oct 7, 2010 at 6:25 PM, Jonathan Ellis <jbellis@gmail.com> wrote:
>> No, LT[E] is not on the roadmap for primary index clauses (GT[E] is,
>> for 0.7.1). =A0So you would want to create an index with an inverted
>> comparator, to turn LTE into GTE.
>
> Hm, sounds rather strange, but that would mean that I'd either be able to
> select GT[E] or LT[E], not combining both, right? So expressions like 123=
 <
> uid < 321 would be completely impossible.

No, you only need one primary index clause to avoid the full CF scan.
After that you can have subordinate clauses on basically anything.  so
uid > 123 would be the primary and uid < 321 would be subordinate, and
that would be fine.  (just like now you can have uid=3D123 and x > foo.)

> How would one proceed to create an
> index with an inverted comparator? I'd have to create my own Sub-Type on
> which I'd invert compare function.

Right, although I think this would be common enough that we'd either
provide reversed comparators or add a "reversed" option to the index
definition.

--=20
Jonathan Ellis
Project Chair, Apache Cassandra
co-founder of Riptano, the source for professional Cassandra support
http://riptano.com