ARROW-11317: [Rust] Include the prettyprint feature in CI Coverage #9262

alamb · 2021-01-19T12:49:01Z

This PR changes the Rust CI job to include the prettyprint feature. (so we have CI coverage of that feature)

As @paddyhoran noted, passing in feature flags to the root workspace doesn't do anything, and as it turns out we already invoke cargo test for the arrow crate separately. Prior to this PR I think that just ran the same tests again with the same features. After this PR, the second run of the arrow tests are run with the prettyprint feature

Here is evidence that the tests are running twice. I picked a PR (randomly)
https://github.com/apache/arrow/pull/9258/checks?check_run_id=1725967358

If you look at the logs for the run-tests action

You can see that the same test is run twice:

2021-01-19T07:02:49.9476111Z test compute::kernels::boolean::tests::test_nonnull_array_is_not_null ... ok
2021-01-19T07:02:49.9476854Z test compute::kernels::boolean::tests::test_nonnull_array_is_null ... ok
2021-01-19T07:02:49.9477616Z test compute::kernels::boolean::tests::test_nonnull_array_with_offset_is_not_null ... ok
2021-01-19T07:02:49.9478427Z test compute::kernels::boolean::tests::test_nonnull_array_with_offset_is_null ... ok
2021-01-19T07:02:49.9479207Z test compute::kernels::boolean::tests::test_nullable_array_is_null ... ok
2021-01-19T07:02:49.9479948Z test compute::kernels::boolean::tests::test_nullable_array_is_not_null ... ok
2021-01-19T07:02:49.9480744Z test compute::kernels::boolean::tests::test_nullable_array_with_offset_is_not_null ... ok
2021-01-19T07:02:49.9481640Z test compute::kernels::boolean::tests::test_nullable_array_with_offset_is_null ... ok
2021-01-19T07:02:49.9487019Z test compute::kernels::boolean::tests::test_nullif_int_array_offset ... ok
2021-01-19T07:02:49.9487773Z test compute::kernels::boolean::tests::test_nullif_int_array ... ok
...
2021-01-19T07:07:23.4568865Z test compute::kernels::boolean::tests::test_nonnull_array_is_not_null ... ok
2021-01-19T07:07:23.4569576Z test compute::kernels::boolean::tests::test_nonnull_array_is_null ... ok
2021-01-19T07:07:23.4570337Z test compute::kernels::boolean::tests::test_nonnull_array_with_offset_is_not_null ... ok
2021-01-19T07:07:23.4571133Z test compute::kernels::boolean::tests::test_nonnull_array_with_offset_is_null ... ok
2021-01-19T07:07:23.4571885Z test compute::kernels::boolean::tests::test_nullable_array_is_not_null ... ok
2021-01-19T07:07:23.4572614Z test compute::kernels::boolean::tests::test_nullable_array_is_null ... ok
2021-01-19T07:07:23.4573383Z test compute::kernels::boolean::tests::test_nullable_array_with_offset_is_not_null ... ok
2021-01-19T07:07:23.4574183Z test compute::kernels::boolean::tests::test_nullable_array_with_offset_is_null ... ok
2021-01-19T07:07:23.4574929Z test compute::kernels::boolean::tests::test_nullif_int_array ... ok
2021-01-19T07:07:23.4575636Z test compute::kernels::boolean::tests::test_nullif_int_array_offset ... ok

alamb · 2021-01-19T12:53:48Z

.github/workflows/rust.yml

          # test datafusion examples
          cd datafusion
          cargo run --example csv_sql
          cargo run --example parquet_sql
          cd ..
          cd arrow
-          cargo test


this appears to run all the arrow tests twice -- once in the main workspace and once in the arrow crate again. I am not sure that is adding any value

This was introduced when I was trying to test the SIMD and non-SIMD versions of the code. Feature options were not being respected within the sub-crates when supplied at the workspace level, see here. So the solution was to move into the sub-crate to enable/disable the feature.

Quickly reading through the cargo issues, it's not clear that this has been resolved yet.

AH, that makes sense @paddyhoran -- I think the feature flag test for simd got moved to an entire different job, namely the linux-test-simd job below. I think I need to re-instante the second CI test but run the tests with the pretty print option. I'll try and do that tomorrow

In fact the CI test failed, exactly as you predicted https://github.com/apache/arrow/runs/1727917047 -- namely that you can't specify --features in the root of the workspace:

Run export CARGO_HOME="/github/home/.cargo" error: --features is not allowed in the root of a virtual workspace note: while this was previously accepted, it didn't actually do anything help: change the current directory to the package directory, or use the --manifest-path flag to the path of the package Error: Process completed with exit code 101.

A solution could be to exclude arrow from the initial cargo test, then test it after cd arrow, with the prettyprint (and other future) feature.

Or has this been resolved?

It has been resolved (though the resolution was to keep running the tests for arrow twice)

To be fair, running the arrow tests is a second or so, so the harm is pretty small. Re-compiling DataFusion's dependencies would have been a different beast ^_^

Is everyone ok if I merge this in I think it will have minimal effect on overall test times (will require arrow to get recompiled one additional time I think)

github-actions · 2021-01-19T13:03:14Z

https://issues.apache.org/jira/browse/ARROW-11317

codecov-io · 2021-01-19T13:37:36Z

Codecov Report

Merging #9262 (063a7c8) into master (e7c69e6) will increase coverage by 0.00%.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master    #9262   +/-   ##
=======================================
  Coverage   81.62%   81.62%           
=======================================
  Files         215      215           
  Lines       52520    52520           
=======================================
+ Hits        42870    42871    +1     
+ Misses       9650     9649    -1

Impacted Files	Coverage Δ
rust/parquet/src/encodings/encoding.rs	`95.05% <0.00%> (+0.19%)`	⬆️

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update e7c69e6...063a7c8. Read the comment docs.

… feature

alamb · 2021-01-20T13:12:37Z

You can now see that the pretty print tests are running:

@paddyhoran

This PR changes the Rust CI job to include the prettyprint feature. (so we have CI coverage of that feature) As @paddyhoran noted, passing in feature flags to the root workspace doesn't do anything, and as it turns out we already invoke `cargo test` for the `arrow` crate separately. Prior to this PR I think that just ran the same tests again with the same features. After this PR, the second run of the arrow tests are run with the `prettyprint` feature Here is evidence that the tests are running twice. I picked a PR (randomly) https://github.com/apache/arrow/pull/9258/checks?check_run_id=1725967358 If you look at the logs for the run-tests action ![Screen Shot 2021-01-19 at 7 42 13 AM](https://user-images.githubusercontent.com/490673/105036642-c4188680-5a2a-11eb-8968-c570d855724e.png) You can see that the same test is run twice: ``` 2021-01-19T07:02:49.9476111Z test compute::kernels::boolean::tests::test_nonnull_array_is_not_null ... ok 2021-01-19T07:02:49.9476854Z test compute::kernels::boolean::tests::test_nonnull_array_is_null ... ok 2021-01-19T07:02:49.9477616Z test compute::kernels::boolean::tests::test_nonnull_array_with_offset_is_not_null ... ok 2021-01-19T07:02:49.9478427Z test compute::kernels::boolean::tests::test_nonnull_array_with_offset_is_null ... ok 2021-01-19T07:02:49.9479207Z test compute::kernels::boolean::tests::test_nullable_array_is_null ... ok 2021-01-19T07:02:49.9479948Z test compute::kernels::boolean::tests::test_nullable_array_is_not_null ... ok 2021-01-19T07:02:49.9480744Z test compute::kernels::boolean::tests::test_nullable_array_with_offset_is_not_null ... ok 2021-01-19T07:02:49.9481640Z test compute::kernels::boolean::tests::test_nullable_array_with_offset_is_null ... ok 2021-01-19T07:02:49.9487019Z test compute::kernels::boolean::tests::test_nullif_int_array_offset ... ok 2021-01-19T07:02:49.9487773Z test compute::kernels::boolean::tests::test_nullif_int_array ... ok ... 2021-01-19T07:07:23.4568865Z test compute::kernels::boolean::tests::test_nonnull_array_is_not_null ... ok 2021-01-19T07:07:23.4569576Z test compute::kernels::boolean::tests::test_nonnull_array_is_null ... ok 2021-01-19T07:07:23.4570337Z test compute::kernels::boolean::tests::test_nonnull_array_with_offset_is_not_null ... ok 2021-01-19T07:07:23.4571133Z test compute::kernels::boolean::tests::test_nonnull_array_with_offset_is_null ... ok 2021-01-19T07:07:23.4571885Z test compute::kernels::boolean::tests::test_nullable_array_is_not_null ... ok 2021-01-19T07:07:23.4572614Z test compute::kernels::boolean::tests::test_nullable_array_is_null ... ok 2021-01-19T07:07:23.4573383Z test compute::kernels::boolean::tests::test_nullable_array_with_offset_is_not_null ... ok 2021-01-19T07:07:23.4574183Z test compute::kernels::boolean::tests::test_nullable_array_with_offset_is_null ... ok 2021-01-19T07:07:23.4574929Z test compute::kernels::boolean::tests::test_nullif_int_array ... ok 2021-01-19T07:07:23.4575636Z test compute::kernels::boolean::tests::test_nullif_int_array_offset ... ok ``` Closes #9262 from alamb/alamb/run_prettyprint_ci Authored-by: Andrew Lamb <andrew@nerdnetworks.org> Signed-off-by: Jorge C. Leitao <jorgecarleitao@gmail.com>

@paddyhoran

This PR changes the Rust CI job to include the prettyprint feature. (so we have CI coverage of that feature) As @paddyhoran noted, passing in feature flags to the root workspace doesn't do anything, and as it turns out we already invoke `cargo test` for the `arrow` crate separately. Prior to this PR I think that just ran the same tests again with the same features. After this PR, the second run of the arrow tests are run with the `prettyprint` feature Here is evidence that the tests are running twice. I picked a PR (randomly) https://github.com/apache/arrow/pull/9258/checks?check_run_id=1725967358 If you look at the logs for the run-tests action ![Screen Shot 2021-01-19 at 7 42 13 AM](https://user-images.githubusercontent.com/490673/105036642-c4188680-5a2a-11eb-8968-c570d855724e.png) You can see that the same test is run twice: ``` 2021-01-19T07:02:49.9476111Z test compute::kernels::boolean::tests::test_nonnull_array_is_not_null ... ok 2021-01-19T07:02:49.9476854Z test compute::kernels::boolean::tests::test_nonnull_array_is_null ... ok 2021-01-19T07:02:49.9477616Z test compute::kernels::boolean::tests::test_nonnull_array_with_offset_is_not_null ... ok 2021-01-19T07:02:49.9478427Z test compute::kernels::boolean::tests::test_nonnull_array_with_offset_is_null ... ok 2021-01-19T07:02:49.9479207Z test compute::kernels::boolean::tests::test_nullable_array_is_null ... ok 2021-01-19T07:02:49.9479948Z test compute::kernels::boolean::tests::test_nullable_array_is_not_null ... ok 2021-01-19T07:02:49.9480744Z test compute::kernels::boolean::tests::test_nullable_array_with_offset_is_not_null ... ok 2021-01-19T07:02:49.9481640Z test compute::kernels::boolean::tests::test_nullable_array_with_offset_is_null ... ok 2021-01-19T07:02:49.9487019Z test compute::kernels::boolean::tests::test_nullif_int_array_offset ... ok 2021-01-19T07:02:49.9487773Z test compute::kernels::boolean::tests::test_nullif_int_array ... ok ... 2021-01-19T07:07:23.4568865Z test compute::kernels::boolean::tests::test_nonnull_array_is_not_null ... ok 2021-01-19T07:07:23.4569576Z test compute::kernels::boolean::tests::test_nonnull_array_is_null ... ok 2021-01-19T07:07:23.4570337Z test compute::kernels::boolean::tests::test_nonnull_array_with_offset_is_not_null ... ok 2021-01-19T07:07:23.4571133Z test compute::kernels::boolean::tests::test_nonnull_array_with_offset_is_null ... ok 2021-01-19T07:07:23.4571885Z test compute::kernels::boolean::tests::test_nullable_array_is_not_null ... ok 2021-01-19T07:07:23.4572614Z test compute::kernels::boolean::tests::test_nullable_array_is_null ... ok 2021-01-19T07:07:23.4573383Z test compute::kernels::boolean::tests::test_nullable_array_with_offset_is_not_null ... ok 2021-01-19T07:07:23.4574183Z test compute::kernels::boolean::tests::test_nullable_array_with_offset_is_null ... ok 2021-01-19T07:07:23.4574929Z test compute::kernels::boolean::tests::test_nullif_int_array ... ok 2021-01-19T07:07:23.4575636Z test compute::kernels::boolean::tests::test_nullif_int_array_offset ... ok ``` Closes apache#9262 from alamb/alamb/run_prettyprint_ci Authored-by: Andrew Lamb <andrew@nerdnetworks.org> Signed-off-by: Jorge C. Leitao <jorgecarleitao@gmail.com>

@paddyhoran

This PR changes the Rust CI job to include the prettyprint feature. (so we have CI coverage of that feature) As @paddyhoran noted, passing in feature flags to the root workspace doesn't do anything, and as it turns out we already invoke `cargo test` for the `arrow` crate separately. Prior to this PR I think that just ran the same tests again with the same features. After this PR, the second run of the arrow tests are run with the `prettyprint` feature Here is evidence that the tests are running twice. I picked a PR (randomly) https://github.com/apache/arrow/pull/9258/checks?check_run_id=1725967358 If you look at the logs for the run-tests action ![Screen Shot 2021-01-19 at 7 42 13 AM](https://user-images.githubusercontent.com/490673/105036642-c4188680-5a2a-11eb-8968-c570d855724e.png) You can see that the same test is run twice: ``` 2021-01-19T07:02:49.9476111Z test compute::kernels::boolean::tests::test_nonnull_array_is_not_null ... ok 2021-01-19T07:02:49.9476854Z test compute::kernels::boolean::tests::test_nonnull_array_is_null ... ok 2021-01-19T07:02:49.9477616Z test compute::kernels::boolean::tests::test_nonnull_array_with_offset_is_not_null ... ok 2021-01-19T07:02:49.9478427Z test compute::kernels::boolean::tests::test_nonnull_array_with_offset_is_null ... ok 2021-01-19T07:02:49.9479207Z test compute::kernels::boolean::tests::test_nullable_array_is_null ... ok 2021-01-19T07:02:49.9479948Z test compute::kernels::boolean::tests::test_nullable_array_is_not_null ... ok 2021-01-19T07:02:49.9480744Z test compute::kernels::boolean::tests::test_nullable_array_with_offset_is_not_null ... ok 2021-01-19T07:02:49.9481640Z test compute::kernels::boolean::tests::test_nullable_array_with_offset_is_null ... ok 2021-01-19T07:02:49.9487019Z test compute::kernels::boolean::tests::test_nullif_int_array_offset ... ok 2021-01-19T07:02:49.9487773Z test compute::kernels::boolean::tests::test_nullif_int_array ... ok ... 2021-01-19T07:07:23.4568865Z test compute::kernels::boolean::tests::test_nonnull_array_is_not_null ... ok 2021-01-19T07:07:23.4569576Z test compute::kernels::boolean::tests::test_nonnull_array_is_null ... ok 2021-01-19T07:07:23.4570337Z test compute::kernels::boolean::tests::test_nonnull_array_with_offset_is_not_null ... ok 2021-01-19T07:07:23.4571133Z test compute::kernels::boolean::tests::test_nonnull_array_with_offset_is_null ... ok 2021-01-19T07:07:23.4571885Z test compute::kernels::boolean::tests::test_nullable_array_is_not_null ... ok 2021-01-19T07:07:23.4572614Z test compute::kernels::boolean::tests::test_nullable_array_is_null ... ok 2021-01-19T07:07:23.4573383Z test compute::kernels::boolean::tests::test_nullable_array_with_offset_is_not_null ... ok 2021-01-19T07:07:23.4574183Z test compute::kernels::boolean::tests::test_nullable_array_with_offset_is_null ... ok 2021-01-19T07:07:23.4574929Z test compute::kernels::boolean::tests::test_nullif_int_array ... ok 2021-01-19T07:07:23.4575636Z test compute::kernels::boolean::tests::test_nullif_int_array_offset ... ok ``` Closes apache#9262 from alamb/alamb/run_prettyprint_ci Authored-by: Andrew Lamb <andrew@nerdnetworks.org> Signed-off-by: Jorge C. Leitao <jorgecarleitao@gmail.com>

codecov-commenter · 2024-08-20T10:57:50Z

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 81.62%. Comparing base (e7c69e6) to head (063a7c8).

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #9262   +/-   ##
=======================================
  Coverage   81.62%   81.62%           
=======================================
  Files         215      215           
  Lines       52520    52520           
=======================================
+ Hits        42870    42871    +1     
+ Misses       9650     9649    -1

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

alamb force-pushed the alamb/run_prettyprint_ci branch from 1e5dc72 to f0b002f Compare January 19, 2021 12:52

alamb commented Jan 19, 2021

View reviewed changes

alamb marked this pull request as ready for review January 19, 2021 12:53

vertexclique approved these changes Jan 19, 2021

View reviewed changes

alamb added the Component: Rust label Jan 19, 2021

alamb added 2 commits January 20, 2021 06:39

ARROW-11317: Run tests once in CI (not twice) and include prettyprint…

cd4d5bd

… feature

keep second CI run

063a7c8

alamb force-pushed the alamb/run_prettyprint_ci branch from f0b002f to 063a7c8 Compare January 20, 2021 11:41

alamb changed the title ~~ARROW-11317: Run tests once in CI (not twice) and include prettyprint feature~~ ARROW-11317: [Rust] Include the prettyprint feature in CI Coverage Jan 20, 2021

alamb requested review from jorgecarleitao and paddyhoran January 22, 2021 12:48

jorgecarleitao closed this in 10f4ada Jan 24, 2021

asfimport mentioned this pull request Apr 18, 2021

[Rust] Test the prettyprint feature in CI #27215

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

ARROW-11317: [Rust] Include the prettyprint feature in CI Coverage #9262

ARROW-11317: [Rust] Include the prettyprint feature in CI Coverage #9262

alamb commented Jan 19, 2021 •

edited

Loading

alamb Jan 19, 2021

paddyhoran Jan 19, 2021

alamb Jan 19, 2021 •

edited

Loading

alamb Jan 20, 2021

nevi-me Jan 22, 2021

alamb Jan 22, 2021

jorgecarleitao Jan 22, 2021

alamb Jan 23, 2021

github-actions bot commented Jan 19, 2021

codecov-io commented Jan 19, 2021 •

edited

Loading

alamb commented Jan 20, 2021 •

edited

Loading

codecov-commenter commented Aug 20, 2024

ARROW-11317: [Rust] Include the prettyprint feature in CI Coverage #9262

ARROW-11317: [Rust] Include the prettyprint feature in CI Coverage #9262

Conversation

alamb commented Jan 19, 2021 • edited Loading

alamb Jan 19, 2021

Choose a reason for hiding this comment

paddyhoran Jan 19, 2021

Choose a reason for hiding this comment

alamb Jan 19, 2021 • edited Loading

Choose a reason for hiding this comment

alamb Jan 20, 2021

Choose a reason for hiding this comment

nevi-me Jan 22, 2021

Choose a reason for hiding this comment

alamb Jan 22, 2021

Choose a reason for hiding this comment

jorgecarleitao Jan 22, 2021

Choose a reason for hiding this comment

alamb Jan 23, 2021

Choose a reason for hiding this comment

github-actions bot commented Jan 19, 2021

codecov-io commented Jan 19, 2021 • edited Loading

Codecov Report

alamb commented Jan 20, 2021 • edited Loading

codecov-commenter commented Aug 20, 2024

Codecov Report

alamb commented Jan 19, 2021 •

edited

Loading

alamb Jan 19, 2021 •

edited

Loading

codecov-io commented Jan 19, 2021 •

edited

Loading

alamb commented Jan 20, 2021 •

edited

Loading