Mailing-List: contact user-help@cassandra.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@cassandra.apache.org
Received-SPF: pass (athena.apache.org: local policy)
DomainKey-Signature: a=rsa-sha1; c=nofws; d=thelastpickle.com; h=to:from
	:subject:date:message-id:content-type:mime-version:in-reply-to;
	 q=dns; s=thelastpickle.com; b=fY/R1Jw7JTnNm6TlsRMra2P4nDmhqnEEo
	Sa9OfQLttTLKH1hkVNU1WIfp8mVqB9w1sFn+Ot2yPknJu9f4YC8yWmKCEsjGSVkG
	qtop0U46DtVo7YAI3ax1oVRVjIJ15TSYrCPZdwpsAJQlTEPKcoKxmWjRR45/IDLz
	FSSfSPUpD8=
To: user@cassandra.apache.org
From: Aaron Morton <aaron@thelastpickle.com>
Subject: Re: column family names
Date: Tue, 31 Aug 2010 00:55:03 GMT
Message-id: <968b952c-6621-ca3d-af24-974b81e174f7@me.com>
Content-Type: multipart/alternative;
 boundary=Apple-Webmail-42--463acb86-31f0-4aff-8e9f-9e69a318ae18
MIME-Version: 1.0
In-Reply-To: <AANLkTikAJPT9Y0qXFMyrravfoweg-v_T4Yz5R3Vu5gA2@mail.gmail.com>


--Apple-Webmail-42--463acb86-31f0-4aff-8e9f-9e69a318ae18
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
    charset=ISO-8859-1;
    format=flowed

under scores or URL encoding ?=0A=0AAaron=0A=0AOn 31 Aug, 2010,at 12:27 PM=
, Benjamin Black <b@b3k.us> wrote:=0A=0APlease don't do this.=0A=0AOn Mon,=
 Aug 30, 2010 at 5:22 AM, Terje Marthinussen=0A<tmarthinussen@gmail.com> w=
rote:=0A> Ah, sorry, I forgot that underscore was part of \w.=0A> That wil=
l do the trick for now.=0A>=0A> I do not see the big issue with file names=
 though. Why not expand the=0A> allowed characters a bit and escape the fi=
le names? Maybe some sort of URL=0A> like escaping.=0A>=0A> Terje=0A>=0A> =
On Mon, Aug 30, 2010 at 6:29 PM, Aaron Morton <aaron@thelastpickle.com>=0A=
> wrote:=0A>>=0A>> Moving to the user list.=0A>> The new restrictions were=
 added as part of =A0CASSANDRA-1377 for 0.6.5 and=0A>> 0.7, AFAIK it's to =
ensure the file names created for the CFs can be=0A>> correctly parsed. So=
 it's probably not going to change.=0A>> The names have to match the \w re=
g ex class, which includes the underscore=0A>> character.=0A>>=0A>> Aaron=0A=
>>=0A>> On 30 Aug 2010, at 21:01, Terje Marthinussen <tmarthinussen@gmail.=
com>=0A>> wrote:=0A>>=0A>> Hi,=0A>>=0A>> Now that we can make columns fami=
lies on the fly, it gets interesting to=0A>> use=0A>> column families more=
 as part of the data model (can reduce diskspace quite=0A>> a=0A>> bit vs.=
 super columns in some cases).=0A>>=0A>> However, currently, the column fa=
mily name validator is pretty strict=0A>> allowing only word characters an=
d in some cases it is pretty darned nice=0A>> to=0A>> be able to put somet=
hing like a "-" inbetweenallthewords.=0A>>=0A>> Any reason to be this stri=
ct or could it be loosened up a little bit?=0A>>=0A>> Terje=0A>=0A>=0A
--Apple-Webmail-42--463acb86-31f0-4aff-8e9f-9e69a318ae18
Content-Type: multipart/related;
    type="text/html";
    boundary=Apple-Webmail-86--463acb86-31f0-4aff-8e9f-9e69a318ae18


--Apple-Webmail-86--463acb86-31f0-4aff-8e9f-9e69a318ae18
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html;
    charset=ISO-8859-1;

<html><body><div>under scores or URL encoding ?</div><div><br></div><div>Aaron</div><d=
iv><br>On 31 Aug, 2010,at 12:27 PM, Benjamin Black &lt;b@b3k.us&gt; wrote:=
<br><br></div><div><blockquote type=3D"cite"><div><div class=3D"_stretch">=
Please don't do this.<br>=0A<br>=0AOn Mon, Aug 30, 2010 at 5:22 AM, Terje =
Marthinussen<br>=0A&lt;tmarthinussen@gmail.com&gt; wrote:<br>=0A&gt; Ah, s=
orry, I forgot that underscore was part of \w.<br>=0A&gt; That will do the=
 trick for now.<br>=0A&gt;<br>=0A&gt; I do not see the big issue with file=
 names though. Why not expand the<br>=0A&gt; allowed characters a bit and =
escape the file names? Maybe some sort of URL<br>=0A&gt; like escaping.<br=
>=0A&gt;<br>=0A&gt; Terje<br>=0A&gt;<br>=0A&gt; On Mon, Aug 30, 2010 at 6:=
29 PM, Aaron Morton &lt;aaron@thelastpickle.com&gt;<br>=0A&gt; wrote:<br>=0A=
&gt;&gt;<br>=0A&gt;&gt; Moving to the user list.<br>=0A&gt;&gt; The new re=
strictions were added as part of &nbsp;CASSANDRA-1377 for 0.6.5 and<br>=0A=
&gt;&gt; 0.7, AFAIK it's to ensure the file names created for the CFs can =
be<br>=0A&gt;&gt; correctly parsed. So it's probably not going to change.<=
br>=0A&gt;&gt; The names have to match the \w reg ex class, which includes=
 the underscore<br>=0A&gt;&gt; character.<br>=0A&gt;&gt;<br>=0A&gt;&gt; Aa=
ron<br>=0A&gt;&gt;<br>=0A&gt;&gt; On 30 Aug 2010, at 21:01, Terje Marthinu=
ssen &lt;tmarthinussen@gmail.com&gt;<br>=0A&gt;&gt; wrote:<br>=0A&gt;&gt;<=
br>=0A&gt;&gt; Hi,<br>=0A&gt;&gt;<br>=0A&gt;&gt; Now that we can make colu=
mns families on the fly, it gets interesting to<br>=0A&gt;&gt; use<br>=0A&=
gt;&gt; column families more as part of the data model (can reduce diskspa=
ce quite<br>=0A&gt;&gt; a<br>=0A&gt;&gt; bit vs. super columns in some cas=
es).<br>=0A&gt;&gt;<br>=0A&gt;&gt; However, currently, the column family n=
ame validator is pretty strict<br>=0A&gt;&gt; allowing only word character=
s and in some cases it is pretty darned nice<br>=0A&gt;&gt; to<br>=0A&gt;&=
gt; be able to put something like a "-" inbetweenallthewords.<br>=0A&gt;&g=
t;<br>=0A&gt;&gt; Any reason to be this strict or could it be loosened up =
a little bit?<br>=0A&gt;&gt;<br>=0A&gt;&gt; Terje<br>=0A&gt;<br>=0A&gt;<br=
>=0A</div></div></blockquote></div></body></html>
--Apple-Webmail-86--463acb86-31f0-4aff-8e9f-9e69a318ae18--

--Apple-Webmail-42--463acb86-31f0-4aff-8e9f-9e69a318ae18--