You are viewing a plain text version of this content. The canonical link for it is here.
Posted to jira@arrow.apache.org by "Andy Salnikov (Jira)" <ji...@apache.org> on 2022/07/01 22:51:00 UTC

[jira] [Commented] (ARROW-16893) [Python] Add quoting style support for pyarrow.csv.WriteOptions

    [ https://issues.apache.org/jira/browse/ARROW-16893?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17561633#comment-17561633 ] 

Andy Salnikov commented on ARROW-16893:
---------------------------------------

Closely related to that, so I do not want to open new ticket for it. In our CSV output we need to produce non-empty string for None values (mysql likes {{{}\N{}}}). C++ definition for WriteOptions has {{null_string}} attribute which can be used to configure that, but that attribute is not exposed in pyarrow wrapper class. Same for {{eol}} attribute - it is also missing from pyarrow. Would be nice to add complete set of attributes to Python wrapper.

> [Python] Add quoting style support for pyarrow.csv.WriteOptions
> ---------------------------------------------------------------
>
>                 Key: ARROW-16893
>                 URL: https://issues.apache.org/jira/browse/ARROW-16893
>             Project: Apache Arrow
>          Issue Type: Improvement
>          Components: Python
>    Affects Versions: 8.0.0
>            Reporter: David Lee
>            Priority: Minor
>              Labels: easyfix
>
> https://issues.apache.org/jira/browse/ARROW-14905
> The quoting style option was added forĀ  C++, but is not supported in Python.
> The C++ default for quoting style is "needed"
> The pyarrow.csv writer module currently produces a CSV file where all strings are double quoted with no option to not wrap strings in double quotes.
> "portfolioID","marketValue","notionalMarketValue","weight","notionalWeight"
> "ABCXYZ12345",26260.74,0.039716113109573174,26260.74,0.039716113109573174



--
This message was sent by Atlassian Jira
(v8.20.10#820010)