[
https://issues.apache.org/jira/browse/ARROW-8493?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Micah Kornfield updated ARROW-8493:
-----------------------------------
Description:
We need a function/class that can take the parquet schema and a proposed arrow
schema (potentially retrieved from parquet metadata) and outputs a data
structure that contains, all of the information in "SchemaField" and the
following additional options:
1. Corresponding Definition level for nullability (wouldn't be populated for
non-null arrays).
2. Correspond Repetition level for lists (wouldn't be populated for for
non-lists).
3. Definition level for "empty lists". (wouldn't be populated for legacy two
level encoded lists).
One option is to augment and populate these on the SchemaField.
was:
We need a function/class that can take the parquet schema and a proposed arrow
schema (potentially retrieved from parquet metadata) and outputs a data
structure that contains, all of the information in "SchemaField" and the
following additional options:
1. Corresponding Definition level for nullability (wouldn't be populated for
non-null arrays).
2. Correspond Repetition level for lists (wouldn't be populated for for
non-lists).
One option is to augment and populate these on the SchemaField.
> [C++] Create unified schema resolution code for Array reconstruction.
> ---------------------------------------------------------------------
>
> Key: ARROW-8493
> URL: https://issues.apache.org/jira/browse/ARROW-8493
> Project: Apache Arrow
> Issue Type: Sub-task
> Components: C++
> Reporter: Micah Kornfield
> Priority: Major
>
> We need a function/class that can take the parquet schema and a proposed
> arrow schema (potentially retrieved from parquet metadata) and outputs a data
> structure that contains, all of the information in "SchemaField" and the
> following additional options:
>
> 1. Corresponding Definition level for nullability (wouldn't be populated for
> non-null arrays).
> 2. Correspond Repetition level for lists (wouldn't be populated for for
> non-lists).
> 3. Definition level for "empty lists". (wouldn't be populated for legacy
> two level encoded lists).
>
> One option is to augment and populate these on the SchemaField.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
