[
https://issues.apache.org/jira/browse/CASSANDRA-2474?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13095913#comment-13095913
]
Pavel Yaskevich edited comment on CASSANDRA-2474 at 9/2/11 11:16 AM:
---------------------------------------------------------------------
bq. We already have the transposition as a visual "hint" as to what is going
on. I don't think it's difficult to say that "you need to use componentX when
using transposition," and :colon tokens are usually used to represent bind
variables in SQL drivers.
I didn't use ":transposed" keyword any where but I'm file using it instead of
:colon tokens. I don't like "componentX" notation because we can simply
preserve ordering of the arguments to make the same thing which is much clearer
syntax on my opinion
bq. Maybe it is better if we will support (..,..,..) notation for composite
columns
Can be used as a replacement for "componentX"
bq. Sorry, I don't follow.
You wrote:
bq. If I wanted to add a location tweet field, I'd need to delete (e.g.) the
2e1c3308,cscotta column and replace it with 2e1c3308,cscotta,<location>. So,
generally not recommended... but the composite column spec implies we should
support it.) Which brings us to: how to deal with insert/update in composite
columns.
I mean when we want to extend "2e1c3308,cscotta" with <location> we can simply
update old column's name to "2e1c3308,cscotta,<location>", for querying we can
use "1/0/-1" as <'end-of-component' byte> from CompositeType doc when we want
to search by specific components of the column name.
e.g. (if I understand CompositeType annotation correctly)
{noformat}
SELECT name AS (tweet_id, username), value AS body
FROM timeline
WHERE tweet_id = '95a789a' AND user_id = 'cscotta'
{noformat}
start = "<7>95a789a.getBytes()<0><7>cscotta.getBytes()<0>"
end = "<7>95a789a.getBytes()<0><7>cscotta.getBytes()<1>"
Which should give as *all* columns where column name starts with 95a789a,cscotta
{noformat}
SELECT name AS (tweet_id, username, location), value AS body
FROM timeline
WHERE tweet_id = '95a789a' AND user_id = 'cscotta' AND location = 'USA'
{noformat}
start = "<7>95a789a.getBytes()<0><7>cscotta.getBytes()<0><3>USA.getBytes()<0>"
end = "<7>95a789a.getBytes()<0><7>cscotta.getBytes()<0><3>USA.getBytes()<1>"
Which should give as *all* columns where column name starts with
95a789a,cscotta,USA
was (Author: xedin):
bq. We already have the transposition as a visual "hint" as to what is
going on. I don't think it's difficult to say that "you need to use componentX
when using transposition," and :colon tokens are usually used to represent bind
variables in SQL drivers.
I didn't use ":transposed" keyword any where but I'm file using it instead of
:colon tokens. I don't like "componentX" notation because we can simply
preserve ordering of the arguments to make the same thing which is much clearer
syntax on my opinion
bq. Maybe it is better if we will support (..,..,..) notation for composite
columns
Can be used as a replacement for "componentX"
bq. Sorry, I don't follow.
You wrote:
bq. If I wanted to add a location tweet field, I'd need to delete (e.g.) the
2e1c3308,cscotta column and replace it with 2e1c3308,cscotta,<location>. So,
generally not recommended... but the composite column spec implies we should
support it.) Which brings us to: how to deal with insert/update in composite
columns.
I mean when we want to extend "2e1c3308,cscotta" with <location> we can simply
update old column's name to "2e1c3308,cscotta,<location>", for querying we can
use "1/0/-1" as <'end-of-component' byte> from CompositeType doc when we want
to search by specific components of the column name.
e.g. (if I understand CompositeType annotation correctly)
{noformat}
SELECT name AS (tweet_id, username), value AS body
FROM timeline
WHERE tweet_id = '95a789a' AND user_id = 'cscotta'
{noformat}
start = "<7>95a789a.getBytes()<0><7>cscotta.getBytes()<0>"
end = "<7>95a789a.getBytes()<0><7>cscotta.getBytes()<1>"
Which should give as *all* columns where column name starts with 95a789a,cscotta
{noformat}
SELECT name AS (tweet_id, username, location), value AS body
FROM timeline
WHERE tweet_id = '95a789a' AND user_id = 'cscotta' AND location = 'USA'
{noformat}
start = "<7>95a789a.getBytes()<0><7>cscotta.getBytes()<0><3>USA.getBytes()<0>"
end = "<7>95a789a.getBytes()<0><7>cscotta.getBytes()<1><3>USA.getBytes()<1>"
Which should give as *all* columns where column name starts with
95a789a,cscotta,USA
> CQL support for compound columns
> --------------------------------
>
> Key: CASSANDRA-2474
> URL: https://issues.apache.org/jira/browse/CASSANDRA-2474
> Project: Cassandra
> Issue Type: Sub-task
> Components: API, Core
> Reporter: Eric Evans
> Assignee: Pavel Yaskevich
> Labels: cql
> Fix For: 1.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg
>
>
> For the most part, this boils down to supporting the specification of
> compound column names (the CQL syntax is colon-delimted terms), and then
> teaching the decoders (drivers) to create structures from the results.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
