[jira] [Commented] (AVRO-2019) Improve documentation for logical type annotations in IDL
[
https://issues.apache.org/jira/browse/AVRO-2019?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16168107#comment-16168107
]
Bridger Howell commented on AVRO-2019:
--
Clarification on my previous comment: I do think this documentation update is a
good idea; I was just trying to say that IDL logical types don't necessarily
need to be treated internally as a feature separate from the generic metadata
facility of IDL annotations.
In order to break {{@logicalType}} annotations in IDL you'd probably either be
changing the avro spec for logical types, or changing the way that annotations
in IDL are processed, either of which is a breaking change already.
On the other hand, if you try to retain the way {{@logicalType}} maps to a
logical type separately from the way annotations are mapped, that could in many
cases make IDL annotations harder to understand by complicating the mechanism.
> Improve documentation for logical type annotations in IDL
> -
>
> Key: AVRO-2019
> URL: https://issues.apache.org/jira/browse/AVRO-2019
> Project: Avro
> Issue Type: Improvement
> Components: doc, logical types
>Reporter: Andrew Rosca
>Assignee: Andrew Rosca
>Priority: Minor
> Attachments: AVRO-2019.patch
>
>
> The IDL documentation lacks information for how annotations can be specified
> for logical types, like in the following example:
> {code}
> protocol test {
> record test {
> @logicalType("timestamp-millis")
> long time;
> }
> }
> {code}
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (AVRO-2019) Improve documentation for logical type annotations in IDL
[
https://issues.apache.org/jira/browse/AVRO-2019?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16167884#comment-16167884
]
Sean Busbey commented on AVRO-2019:
---
it looks like the timestamp related logical types are only in 1.8+. Folks have
any preference between getting them added back into the 1.7 line vs updating
this doc to use decimal (which is the only logical type defined in 1.7 AFAICT)
> Improve documentation for logical type annotations in IDL
> -
>
> Key: AVRO-2019
> URL: https://issues.apache.org/jira/browse/AVRO-2019
> Project: Avro
> Issue Type: Improvement
> Components: doc, logical types
>Reporter: Andrew Rosca
>Assignee: Andrew Rosca
>Priority: Minor
> Attachments: AVRO-2019.patch
>
>
> The IDL documentation lacks information for how annotations can be specified
> for logical types, like in the following example:
> {code}
> protocol test {
> record test {
> @logicalType("timestamp-millis")
> long time;
> }
> }
> {code}
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Assigned] (AVRO-2019) Improve documentation for logical type annotations in IDL
[
https://issues.apache.org/jira/browse/AVRO-2019?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Sean Busbey reassigned AVRO-2019:
-
Assignee: Andrew Rosca
> Improve documentation for logical type annotations in IDL
> -
>
> Key: AVRO-2019
> URL: https://issues.apache.org/jira/browse/AVRO-2019
> Project: Avro
> Issue Type: Improvement
> Components: doc, logical types
>Reporter: Andrew Rosca
>Assignee: Andrew Rosca
>Priority: Minor
> Attachments: AVRO-2019.patch
>
>
> The IDL documentation lacks information for how annotations can be specified
> for logical types, like in the following example:
> {code}
> protocol test {
> record test {
> @logicalType("timestamp-millis")
> long time;
> }
> }
> {code}
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (AVRO-2019) Improve documentation for logical type annotations in IDL
[
https://issues.apache.org/jira/browse/AVRO-2019?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16167839#comment-16167839
]
Sean Busbey commented on AVRO-2019:
---
Given the docs on logical types elsewhere, I think it's worth calling out this
particular use of annotations in IDL. Folks who rely on IDL to use Avro
shouldn't have to know the implementation details of how logical types are
implemented via properties so that they can reason out how to use them in IDL.
> Improve documentation for logical type annotations in IDL
> -
>
> Key: AVRO-2019
> URL: https://issues.apache.org/jira/browse/AVRO-2019
> Project: Avro
> Issue Type: Improvement
> Components: doc, logical types
>Reporter: Andrew Rosca
>Priority: Minor
> Attachments: AVRO-2019.patch
>
>
> The IDL documentation lacks information for how annotations can be specified
> for logical types, like in the following example:
> {code}
> protocol test {
> record test {
> @logicalType("timestamp-millis")
> long time;
> }
> }
> {code}
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (AVRO-2069) Use primitive fields in generated getters & setters in Java code
[ https://issues.apache.org/jira/browse/AVRO-2069?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16167822#comment-16167822 ] Sean Busbey commented on AVRO-2069: --- I like this as the approach going forward, but this will be an incompatible change. Should we make a follow on jira to make a backwards compatible version for 1.8/1.7 (something like getUnboxedFieldName in addition to getFieldName). > Use primitive fields in generated getters & setters in Java code > > > Key: AVRO-2069 > URL: https://issues.apache.org/jira/browse/AVRO-2069 > Project: Avro > Issue Type: Improvement >Affects Versions: 1.8.2 >Reporter: Daniil Gitelson >Assignee: Daniil Gitelson > > Currently, for primitive types (such as int, long, etc) generated getters and > setters return and accept java.lang.* boxed (while fields actually holds > primitive values). This is inefeccient and produces code boilerplate. > Changed this behaviour in pull request: > https://github.com/apache/avro/pull/243 -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Commented] (AVRO-2076) Combine already serialized Avro records to an Avro file
[ https://issues.apache.org/jira/browse/AVRO-2076?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16167422#comment-16167422 ] Erik van Oosten commented on AVRO-2076: --- Awesome! Thanks Doug. Somehow I missed that method. > Combine already serialized Avro records to an Avro file > --- > > Key: AVRO-2076 > URL: https://issues.apache.org/jira/browse/AVRO-2076 > Project: Avro > Issue Type: Wish >Reporter: Erik van Oosten > > In some use cases Avro events arrive already serialized (e.g. when listening > to a Kafka topic). It would be great if there would an API that allows > writing an Avro file without the need for deserializing and serializing these > Avro records. > Providing such an API allows for very efficient creation of Avro files: given > that these Avro records are written with the same schema, an Avro file would > write will the exact same bytes anyway (before block compression). -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Resolved] (AVRO-2076) Combine already serialized Avro records to an Avro file
[ https://issues.apache.org/jira/browse/AVRO-2076?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Erik van Oosten resolved AVRO-2076. --- Resolution: Not A Problem > Combine already serialized Avro records to an Avro file > --- > > Key: AVRO-2076 > URL: https://issues.apache.org/jira/browse/AVRO-2076 > Project: Avro > Issue Type: Wish >Reporter: Erik van Oosten > > In some use cases Avro events arrive already serialized (e.g. when listening > to a Kafka topic). It would be great if there would an API that allows > writing an Avro file without the need for deserializing and serializing these > Avro records. > Providing such an API allows for very efficient creation of Avro files: given > that these Avro records are written with the same schema, an Avro file would > write will the exact same bytes anyway (before block compression). -- This message was sent by Atlassian JIRA (v6.4.14#64029)
