Mailing-List: contact common-commits-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Reply-To: common-dev@hadoop.apache.org
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable
From: Apache Wiki <wikidiffs@apache.org>
To: Apache Wiki <wikidiffs@apache.org>
Date: Tue, 01 Feb 2011 20:48:50 -0000
Message-ID: <20110201204850.50909.87620@eosnew.apache.org>
Subject: 
 =?utf-8?q?=5BHadoop_Wiki=5D_Update_of_=22Hive/PartitionedViews=22_by_John?=
 =?utf-8?q?Sichi?=

Dear Wiki user,

You have subscribed to a wiki page or wiki category on "Hadoop Wiki" for ch=
ange notification.

The "Hive/PartitionedViews" page has been changed by JohnSichi.
http://wiki.apache.org/hadoop/Hive/PartitionedViews?action=3Ddiff&rev1=3D2&=
rev2=3D3

--------------------------------------------------

   1. One possible approach mentioned in [[https://issues.apache.org/jira/b=
rowse/HIVE-1079|HIVE-1079]] is to infer view partitions automatically based=
 on the partitions of the underlying tables.  A command such as SHOW PARTIT=
IONS could then synthesize virtual partition descriptors on the fly.  This =
is fairly easy to do for use case #1, but potentially very difficult for us=
e cases #2 and #3.  So for now, we are punting on this approach.
   1. Instead, we will require users to explicitly declare view partitionin=
g as part of CREATE VIEW, and explicitly manage partition metadata via ALTE=
R VIEW {ADD|DROP} PARTITION.  This allows all of the use cases to be satisf=
ied (while placing more burden on the user, and taking up more metastore sp=
ace).
  =

+ =3D Syntax =3D
+ =

+ {{{
+ CREATE VIEW [IF NOT EXISTS] view_name [ (column_name [COMMENT column_comm=
ent], ...) ]
+ [COMMENT table_comment]
+ [ PARTITIONED ON (col1, col2, ...) ]
+ [ TBLPROPERTIES ... ]
+ AS SELECT ...
+ =

+ ALTER VIEW view_name ADD [IF NOT EXISTS] partition_spec partition_spec ...
+ =

+ ALTER VIEW view_name DROP [IF EXISTS] partition_spec, partition_spec, ...
+ =

+ partition_spec:
+   : PARTITION (partition_col =3D partition_col_value, partition_col =3D p=
artiton_col_value, ...)
+ }}}
+ =

+ Notes:
+ =

+  * Whereas CREATE TABLE uses PARTITIONED BY, CREATE VIEW uses PARTITIONED=
 ON.  This difference is intentional because in CREATE TABLE, the PARTITION=
ED BY clause specifies additional column definitions which are appended to =
the non-partitioning columns.  With CREATE VIEW, the PARTITIONED ON clause =
references (by name) columns already produced by the view definition.  Only=
 column names appear in PARTITIONED ON; no types etc.  However, to match th=
e CREATE TABLE convention of trailing partitioning columns, the columns ref=
erenced by the PARTITIONED ON clause must be the last columns in the view d=
efinition, and their order in the PARTITIONED ON clause must match their or=
der in the view definition.
+  * The ALTER VIEW ADD/DROP partition syntax is identical to ALTER TABLE, =
except that it is illegal to specify a LOCATION clause.
+  * Other ALTER TABLE commands which operate on partitions (e.g. TOUCH/ARC=
HIVE) are not supported.  (But maybe we need to support TOUCH?)
+=20