You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@arrow.apache.org by "Joris Van den Bossche (Jira)" <ji...@apache.org> on 2019/09/19 13:41:00 UTC

[jira] [Commented] (ARROW-4359) [Python] Column metadata is not saved or loaded in parquet

    [ https://issues.apache.org/jira/browse/ARROW-4359?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16933389#comment-16933389 ] 

Joris Van den Bossche commented on ARROW-4359:
----------------------------------------------

It could also be an option to use the serialized arrow schema for this. That would be sufficient for an arrow <-> parquet roundtrip, but is of course not a general parquet mechanism (so the metadata is not in the proper format to be reusable from other parquet readers)

> [Python] Column metadata is not saved or loaded in parquet
> ----------------------------------------------------------
>
>                 Key: ARROW-4359
>                 URL: https://issues.apache.org/jira/browse/ARROW-4359
>             Project: Apache Arrow
>          Issue Type: Bug
>          Components: Python
>            Reporter: Seb Fru
>            Priority: Major
>              Labels: parquet
>             Fix For: 1.0.0
>
>
> Hi all,
> a while ago I posted this issue: ARROW-3866
> While working with Pyarrow I encountered another potential bug related to column metadata: If I create a table containing columns with metadata everything is fine. But after I save the table to parquet and load it back as a table using pq.read_table, the column metadata is gone.
>  
> As of now I can not say yet whether the metadata is not saved correctly or not loaded correctly, as I have no idea how to verify it. Unfortunately I also don't have the time try a lot, but I wanted to let you know anyway. 
>  
> {code}
> field0 = pa.field('field1', pa.int64(), metadata=dict(a="A", b="B"))
> field1 = pa.field('field2', pa.int64(), nullable=False)
> columns = [
>     pa.column(field0, pa.array([1, 2])),
>     pa.column(field1, pa.array([3, 4]))
> ]
> table = pa.Table.from_arrays(columns)
> pq.write_table(tab, path)
> tab2 = pq.read_table(path)
> tab2.column(0).field.metadata
> {code}
>  



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