Skip to content

Commit

Permalink
Add can we use this for the Arrow IPC format
Browse files Browse the repository at this point in the history
  • Loading branch information
kou committed Aug 27, 2024
1 parent 007a310 commit 06b06a8
Showing 1 changed file with 12 additions and 1 deletion.
13 changes: 12 additions & 1 deletion docs/source/format/CDataInterfaceStatistics.rst
Original file line number Diff line number Diff line change
Expand Up @@ -32,7 +32,7 @@ be read as Apache Arrow data may have statistics. For example, the
Apache Parquet C++ implementation can read an Apache Parquet file as
Apache Arrow data and the Apache Parquet file may have statistics.

One of the Arrow C data interface use cases is the following:
One of :ref:`c-data-interface` use cases is the following:

1. Module A reads Apache Parquet file as Apache Arrow data.
2. Module A passes the read Apache Arrow data to module B through the
Expand Down Expand Up @@ -61,6 +61,17 @@ For example, ADBC has `the statistics related APIs
<https://arrow.apache.org/adbc/current/format/specification.html#statistics>`__.
This specification doesn't replace them.

This specification may fit some use cases of :ref:`format-ipc` not the
Arrow data interface. But we don't recommend this specification for
the Arrow IPC format for now. Because we may be able to define better
specification for the Arrow IPC format. The Arrow IPC format has some
different features compared with the Arrow C data interface. For
example, the Arrow IPC format can have :ref:`ipc-message-format
metadata for each message`. If you're interested in the specification
for passing statistics through the Arrow IPC format, please start a
discussion on the `Arrow development mailing-list
<https://arrow.apache.org/community/>`__.

.. _c-data-interface-statistics-schema:

Schema
Expand Down

0 comments on commit 06b06a8

Please sign in to comment.