You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@arrow.apache.org by "Antoine Pitrou (JIRA)" <ji...@apache.org> on 2018/06/01 08:47:00 UTC

[jira] [Updated] (ARROW-2660) [Python] Experiment with zero-copy pickling

     [ https://issues.apache.org/jira/browse/ARROW-2660?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Antoine Pitrou updated ARROW-2660:
----------------------------------
    Description: 
PEP 574 has an implementation ready and a PyPI-available backport (at [https://pypi.org/project/pickle5/] ). Adding experimental support for it would allow for zero-copy pickling of Arrow arrays, columns, etc.

I think it mainly involves implementing {{reduce_ex}} on the {{Buffer}} class, as described in [https://www.python.org/dev/peps/pep-0574/#producer-api]

In addition, the consumer API added by PEP 574 could be used in Arrow's serialization layer, to avoid or minimize copies when serializing foreign objects.

  was:
PEP 574 has an implementation ready and a PyPI-available backport (at [https://pypi.org/project/pickle5/] ). Adding experimental support for it would allow for zero-copy pickling of Arrow arrays, columns, etc.

I think it mainly involves implementing {{reduce_ex}} on the {{Buffer}} class, as described in [https://www.python.org/dev/peps/pep-0574/#producer-api]

In addition, the consumer API added by PEP 574 could be used in Arrow's serialization array, to avoid or minimize copies when serializing foreign objects.


> [Python] Experiment with zero-copy pickling
> -------------------------------------------
>
>                 Key: ARROW-2660
>                 URL: https://issues.apache.org/jira/browse/ARROW-2660
>             Project: Apache Arrow
>          Issue Type: Wish
>          Components: Python
>    Affects Versions: 0.9.0
>            Reporter: Antoine Pitrou
>            Priority: Major
>
> PEP 574 has an implementation ready and a PyPI-available backport (at [https://pypi.org/project/pickle5/] ). Adding experimental support for it would allow for zero-copy pickling of Arrow arrays, columns, etc.
> I think it mainly involves implementing {{reduce_ex}} on the {{Buffer}} class, as described in [https://www.python.org/dev/peps/pep-0574/#producer-api]
> In addition, the consumer API added by PEP 574 could be used in Arrow's serialization layer, to avoid or minimize copies when serializing foreign objects.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)