doc: add example of RowFilter usage#9115
Conversation
37be4e1 to
f286dfd
Compare
f286dfd to
bc8e06f
Compare
|
@alamb would you mind reviewing this? Thanks! |
| use parquet::errors::Result; | ||
| use std::fs::File; | ||
|
|
||
| // RowFilter / with_row_filter usage. For background and more |
There was a problem hiding this comment.
Are we better off removing this and keeping only the doctest to reduce duplication?
There was a problem hiding this comment.
Yes, I agree -- I think the doc examples are easier to find so I recommend removing this example file
Actually, looking at the existing examples I think many of them are redundant / would be easier to find if we moved them into the documentation:
https://github.com/apache/arrow-rs/tree/main/parquet/examples
There was a problem hiding this comment.
| /// more efficient skipping over data pages. See [`ArrowReaderOptions::with_page_index`]. | ||
| /// | ||
| /// For a running example see `parquet/examples/read_with_row_filter.rs`. | ||
| /// See <https://arrow.apache.org/blog/2025/12/11/parquet-late-materialization-deep-dive/> |
There was a problem hiding this comment.
/// See the [blog post on late materialization] for a more technical explanation.
///
/// ...
///
/// [blog post on late materialization]: https://arrow.apache.org/blog/2025/12/11/parquet-late-materialization-deep-dive/
Slightly nice formatting this way
| use parquet::errors::Result; | ||
| use std::fs::File; | ||
|
|
||
| // RowFilter / with_row_filter usage. For background and more |
There was a problem hiding this comment.
Yes, I agree -- I think the doc examples are easier to find so I recommend removing this example file
Actually, looking at the existing examples I think many of them are redundant / would be easier to find if we moved them into the documentation:
https://github.com/apache/arrow-rs/tree/main/parquet/examples
| /// more efficient skipping over data pages. See [`ArrowReaderOptions::with_page_index`]. | ||
| /// | ||
| /// For a running example see `parquet/examples/read_with_row_filter.rs`. | ||
| /// See <https://arrow.apache.org/blog/2025/12/11/parquet-late-materialization-deep-dive/> |
| /// let builder = ParquetRecordBatchReaderBuilder::try_new(file)?; | ||
| /// let schema_desc = builder.metadata().file_metadata().schema_descr_ptr(); | ||
| /// | ||
| /// // Create predicate: column id > 4. This col has index 0. |
There was a problem hiding this comment.
| /// // Create predicate: column id > 4. This col has index 0. | |
| /// // Create predicate that evaluates `id > 4`. The `id` column has index 0. |
| /// // Create predicate: column id > 4. This col has index 0. | ||
| /// let projection = ProjectionMask::leaves(&schema_desc, [0]); | ||
| /// let predicate = ArrowPredicateFn::new(projection, |batch| { | ||
| /// let id_col = batch.column(0); |
There was a problem hiding this comment.
As a minor suggestion, I think it would make a nicer example if you picked a different column from the file other than 0 so that it is clear the batch passed to the predicate only contains the selected projection column
For example, perhaps you could use the int_col (column index 4)
> select * from './parquet-testing/data/alltypes_plain.parquet';
+----+----------+-------------+--------------+---------+------------+-----------+------------+------------------+------------+---------------------+
| id | bool_col | tinyint_col | smallint_col | int_col | bigint_col | float_col | double_col | date_string_col | string_col | timestamp_col |
+----+----------+-------------+--------------+---------+------------+-----------+------------+------------------+------------+---------------------+
| 4 | true | 0 | 0 | 0 | 0 | 0.0 | 0.0 | 30332f30312f3039 | 30 | 2009-03-01T00:00:00 |
| 5 | false | 1 | 1 | 1 | 10 | 1.1 | 10.1 | 30332f30312f3039 | 31 | 2009-03-01T00:01:00 |
| 6 | true | 0 | 0 | 0 | 0 | 0.0 | 0.0 | 30342f30312f3039 | 30 | 2009-04-01T00:00:00 |
| 7 | false | 1 | 1 | 1 | 10 | 1.1 | 10.1 | 30342f30312f3039 | 31 | 2009-04-01T00:01:00 |
| 2 | true | 0 | 0 | 0 | 0 | 0.0 | 0.0 | 30322f30312f3039 | 30 | 2009-02-01T00:00:00 |
| 3 | false | 1 | 1 | 1 | 10 | 1.1 | 10.1 | 30322f30312f3039 | 31 | 2009-02-01T00:01:00 |
| 0 | true | 0 | 0 | 0 | 0 | 0.0 | 0.0 | 30312f30312f3039 | 30 | 2009-01-01T00:00:00 |
| 1 | false | 1 | 1 | 1 | 10 | 1.1 | 10.1 | 30312f30312f3039 | 31 | 2009-01-01T00:01:00 |
+----+----------+-------------+--------------+---------+------------+-----------+------------+------------------+------------+---------------------+
8 row(s) fetched.
Elapsed 0.039 seconds.
> describe './parquet-testing/data/alltypes_plain.parquet';
+-----------------+---------------+-------------+
| column_name | data_type | is_nullable |
+-----------------+---------------+-------------+
| id | Int32 | YES |
| bool_col | Boolean | YES |
| tinyint_col | Int32 | YES |
| smallint_col | Int32 | YES |
| int_col | Int32 | YES |
| bigint_col | Int64 | YES |
| float_col | Float32 | YES |
| double_col | Float64 | YES |
| date_string_col | BinaryView | YES |
| string_col | BinaryView | YES |
| timestamp_col | Timestamp(ns) | YES |
+-----------------+---------------+-------------+
11 row(s) fetched.
Elapsed 0.005 seconds.|
I made a follow on PR with small improvements |
# Which issue does this PR close? - part of #9096 - Follow on to #9115 # Rationale for this change @sonhmai started us off with #9115 @Jefffrey and I had some suggestions on the PR and I found some more while going through it again, so I figured I would make anew PR # What changes are included in this PR? 1. Improve the documentation 2. Improve the doc comment example 3. Remove redundant example in parquet/examples/read_with_row_filter.rs # Are these changes tested? By CI # Are there any user-facing changes? <!-- If there are user-facing changes then we may require documentation to be updated before approving the PR. If there are any breaking changes to public APIs, please call them out. --> --------- Co-authored-by: Ed Seidl <etseidl@users.noreply.github.com>
# Which issue does this PR close? - Closes apache#9096. # Rationale for this change The RowFilter API does exist and can evaluate predicates during evaluation, but it has no examples. # What changes are included in this PR? - Added a rustdoc example and blog link to `ParquetRecordBatchReaderBuilder::with_row_filter`. - Added a running example in `parquet/examples/read_with_row_filter.rs` # Are these changes tested? Yes ``` cargo run -p parquet --example read_with_row_filter cargo test -p parquet --doc ``` # Are there any user-facing changes? Yes, doc only. No API changes.
) # Which issue does this PR close? - part of apache#9096 - Follow on to apache#9115 # Rationale for this change @sonhmai started us off with apache#9115 @Jefffrey and I had some suggestions on the PR and I found some more while going through it again, so I figured I would make anew PR # What changes are included in this PR? 1. Improve the documentation 2. Improve the doc comment example 3. Remove redundant example in parquet/examples/read_with_row_filter.rs # Are these changes tested? By CI # Are there any user-facing changes? <!-- If there are user-facing changes then we may require documentation to be updated before approving the PR. If there are any breaking changes to public APIs, please call them out. --> --------- Co-authored-by: Ed Seidl <etseidl@users.noreply.github.com>
Which issue does this PR close?
Rationale for this change
The RowFilter API does exist and can evaluate predicates during evaluation, but it has no examples.
What changes are included in this PR?
ParquetRecordBatchReaderBuilder::with_row_filter.parquet/examples/read_with_row_filter.rsAre these changes tested?
Yes
Are there any user-facing changes?
Yes, doc only. No API changes.