etseidl commented on code in PR #8764:
URL: https://github.com/apache/arrow-rs/pull/8764#discussion_r2487402282
##########
parquet/src/file/serialized_reader.rs:
##########
@@ -398,6 +396,8 @@ pub(crate) fn decode_page(
let decompressed_size = uncompressed_page_size - offset;
let mut decompressed = Vec::with_capacity(uncompressed_page_size);
decompressed.extend_from_slice(&buffer[..offset]);
+ // decompressed size of zero corresponds to a page with only null
values
+ // see
https://github.com/apache/parquet-format/blob/master/README.md#data-pages
Review Comment:
> Would page without values or some word better here?
I think the current wording is correct. In this case `num_values` is
non-zero, so there *are* values, but they all just happen to be `null` (and
thus are not encoded in the data).
I'd think the ways `decompressed_size` can be zero are 1) v2 page with all
null values, ~2) v1 page with no nesting and all nulls at the top level (so D
is always 0)~. Second case would have definition level data.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
