This is an automated email from the ASF dual-hosted git repository.
gershinsky pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/parquet-format.git
The following commit(s) were added to refs/heads/master by this push:
new f4e3042 GH-452: Clarify use of RowGroup.ordinal field (#453)
f4e3042 is described below
commit f4e3042eaae5606b280d2d09a44ac54570518567
Author: ggershinsky <[email protected]>
AuthorDate: Wed Sep 25 08:52:30 2024 +0300
GH-452: Clarify use of RowGroup.ordinal field (#453)
* clarify rg ordinal in readers
* nit: sentence structure
---
Encryption.md | 5 +++++
1 file changed, 5 insertions(+)
diff --git a/Encryption.md b/Encryption.md
index 008519b..180b9aa 100644
--- a/Encryption.md
+++ b/Encryption.md
@@ -356,6 +356,11 @@ struct RowGroup {
}
```
+The integrity of this field is protected by authenticated encryption of the
footer (FileMetaData). Therefore,
+the reader implementations can use either a local row group counter (ordinal)
or the `RowGroup.ordinal`
+field as an input to AAD suffix calculation. The latter option can be helpful
when different reader
+threads process different row groups in the same parquet file.
+
A `crypto_metadata` field is set in each ColumnChunk in the encrypted columns.
ColumnCryptoMetaData
is a union - the actual structure is chosen depending on whether the column is
encrypted with the
footer encryption key, or with a column-specific key. For the latter, a key
metadata can be specified.
