sgilmore10 opened a new pull request, #37579:
URL: https://github.com/apache/arrow/pull/37579
<!--
Thanks for opening a pull request!
If this is your first pull request you can find detailed information on how
to contribute here:
* [New Contributor's
Guide](https://arrow.apache.org/docs/dev/developers/guide/step_by_step/pr_lifecycle.html#reviews-and-merge-of-the-pull-request)
* [Contributing
Overview](https://arrow.apache.org/docs/dev/developers/overview.html)
If this is not a [minor
PR](https://github.com/apache/arrow/blob/main/CONTRIBUTING.md#Minor-Fixes).
Could you open an issue for this pull request on GitHub?
https://github.com/apache/arrow/issues/new/choose
Opening GitHub issues ahead of time contributes to the
[Openness](http://theapacheway.com/open/#:~:text=Openness%20allows%20new%20users%20the,must%20happen%20in%20the%20open.)
of the Apache Arrow project.
Then could you also rename the pull request title in the following format?
GH-${GITHUB_ISSUE_ID}: [${COMPONENT}] ${SUMMARY}
or
MINOR: [${COMPONENT}] ${SUMMARY}
In the case of PARQUET issues on JIRA the title also supports:
PARQUET-${JIRA_ISSUE_ID}: [${COMPONENT}] ${SUMMARY}
-->
### Rationale for this change
`ChunkedArray::Equals(const std::shared_ptr<ChunkedArray>& other)` assumes
that if the two `ChunkedArray`s share the same memory address, then they must
be equal. However, this optimization doesn't take into account that `NaN`
values are not considered equal by default. Consequently, this can lead to
surprising, inconsistent results from a user's perspective. For example,
`ChunkedArray::Equals(const std::shared_ptr<ChunkedArray>& other)` and
`ChunkedArray::Equals(const ChunkedArray& other)` may return different results.
The program below illustrates this inconsistency:
```c++
#include <arrow/api.h>
#include <arrow/type.h>
#include <iostream>
#include <math.h>
#include <sstream>
arrow::Result<std::shared_ptr<arrow::ChunkedArray>> make_chunked_array() {
arrow::NumericBuilder<arrow::DoubleType> builder;
std::shared_ptr<arrow::Array> array;
ARROW_RETURN_NOT_OK(builder.AppendValues({0, 1, NAN, 2, 4}));
ARROW_RETURN_NOT_OK(builder.Finish(&array));
return arrow::ChunkedArray::Make({array});
}
int main(int argc, char *argv[])
{
auto maybe_chunked_array = make_chunked_array();
if (!maybe_chunked_array.ok()) {
return -1;
}
auto chunked_array = std::move(maybe_chunked_array).ValueUnsafe();
auto array = chunked_array->chunk(0);
std::stringstream stream;
stream << "chunked_array contents: ";
stream << "\n\n";
stream << chunked_array->ToString();
stream << "\n\n";
stream << "chunked_array->Equals(chunked_array): ";
stream << chunked_array->Equals(chunked_array);
stream << "chunked_array->Equals(*chunked_array): ";
stream << chunked_array->Equals(*chunked_array);
std::cout << stream.str() << std::endl;
}
```
Here is the output of this program:
```shell
chunked_array contents:
[
[
0,
1,
nan,
2,
4
]
]
chunked_array->Equals(chunked_array): 1
chunked_array->Equals(*chunked_array): 0
```
### What changes are included in this PR?
Updated `ChunkedArray::Equals(const std::shared_ptr<ChunkedArray>& other)`
to only return `true` early IF:
- The two share the same address AND
- They cannot have `NaN` values
If both of those conditions are not satisfied, `ChunkedArray::Equals(const
std::shared_ptr<ChunkedArray>& other)` will do the element-by-element
comparison.
### Are these changes tested?
Yes. I added a new test case called `EqualsSameAddressWithNaNs` to
`chunked_array_test.cc`.
### Are there any user-facing changes?
Yes. `ChunkedArray::Equals(const std::shared_ptr<arrow::ChunkedArray>&
other)` may return `false` even if the two `ChunkedArray`s have the same memory
address. This will only occur if the `ChunkedArray`'s contain `NaN` values.
<!--
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 uncomment the line
below and explain which changes are breaking.
-->
<!-- **This PR includes breaking changes to public APIs.** -->
<!--
Please uncomment the line below (and provide explanation) if the changes fix
either (a) a security vulnerability, (b) a bug that caused incorrect or invalid
data to be produced, or (c) a bug that causes a crash (even when the API
contract is upheld). We use this to highlight fixes to issues that may affect
users without their knowledge. For this reason, fixing bugs that cause errors
don't count, since those are usually obvious.
-->
<!-- **This PR contains a "Critical Fix".** -->
--
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]
