[jira] [Commented] (AVRO-2891) Expose last sync offset written on DataFileWriter

["ASF subversion and git services (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/AVRO-2891?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17204759#comment-17204759 ] 

ASF subversion and git services commented on AVRO-2891:
-------------------------------------------------------

Commit 7d1e63b219e6d0778bc57195152477adee97fcab in avro's branch refs/heads/master from Krishnan Sundaram
[ https://gitbox.apache.org/repos/asf?p=avro.git;h=7d1e63b ]

AVRO-2891: Expose last sync offset written on DataFileWriter (#954)

* AVRO-2891: Add code and tests for lastSync API

* AVRO-2891: simplify tests

Tests now only validate equivalence of sync markers between DataFileWriter and Reader. Remove additional test validating that a specific row can be read using the sync returned from DataFileWriter since it is somewhat redundant.

* AVRO-2891: revert change to cmakelists.txt

The original change was needed to cmakelists.txt due to linker errors
faced when using the standard build instructions.

* add another test to highlight efficient read of avro record using the
new API

* add few more comments to the new test

* Change API name as per PR feedback

Co-authored-by: Krishnan Sundaram <kr...@microsoft.com>

> Expose last sync offset written on DataFileWriter
> -------------------------------------------------
>
>                 Key: AVRO-2891
>                 URL: https://issues.apache.org/jira/browse/AVRO-2891
>             Project: Apache Avro
>          Issue Type: Improvement
>          Components: c++
>    Affects Versions: 1.9.2
>            Reporter: Krishnan Sundaram
>            Assignee: Krishnan Sundaram
>            Priority: Minor
>             Fix For: 1.9.3, 1.11.0, 1.10.1
>
>   Original Estimate: 168h
>  Remaining Estimate: 168h
>
> Requesting an API lastSync() on DataFileWriter. It is useful to expose sync offsets during the process of writing an Avro file - Given sync info, it makes it easier to fetch a specific block and read if so desired, rather than having to read the whole file to locate contents of a record in a specific block. 



--
This message was sent by Atlassian Jira
(v8.3.4#803005)