You are viewing a plain text version of this content. The canonical link for it is here.
Posted to github@arrow.apache.org by "js8544 (via GitHub)" <gi...@apache.org> on 2023/06/21 15:26:23 UTC
[GitHub] [arrow] js8544 opened a new pull request, #36204: GH-36203: [C++] Support casting in both ways for is_in and index_in
js8544 opened a new pull request, #36204:
URL: https://github.com/apache/arrow/pull/36204
<!--
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
<!--
Why are you proposing this change? If this is already explained clearly in the issue then this section is not needed.
Explaining clearly why changes are proposed helps reviewers understand your changes and offer better suggestions for fixes.
-->
This is a follow up of https://github.com/apache/arrow/pull/36058#pullrequestreview-1488682384. Currently it only try to cast the value set to input type, not the other way around. This causes some valid input types to be rejected.
### What changes are included in this PR?
<!--
There is no need to duplicate the description in the issue here but it is sometimes worth providing a summary of the individual changes in this PR.
-->
The kernels will first try to case value_set to input type during preparation. If it doesn't work, it would try to cast input to value_set type before the lookup happens.
### Are these changes tested?
<!--
We typically require tests for all PRs in order to:
1. Prevent the code from being accidentally broken by subsequent changes
2. Serve as another way to document the expected behavior of the code
If tests are not included in your PR, please explain why (for example, are they covered by existing tests)?
-->
Yes.
### Are there any user-facing changes?
Some previously rejected input types will now be valid.
<!--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [arrow] js8544 commented on a diff in pull request #36204: GH-36203: [C++] Support casting in both ways for is_in and index_in
Posted by "js8544 (via GitHub)" <gi...@apache.org>.
js8544 commented on code in PR #36204:
URL: https://github.com/apache/arrow/pull/36204#discussion_r1244763406
##########
cpp/src/arrow/compute/kernels/scalar_set_lookup_test.cc:
##########
@@ -253,11 +253,11 @@ TEST_F(TestIsInKernel, TimeDuration) {
"[true, false, false, true, true]", /*skip_nulls=*/true);
}
- // Different units, invalid cast
- ASSERT_RAISES(Invalid, IsIn(ArrayFromJSON(duration(TimeUnit::SECOND), "[0, 1, 2]"),
- ArrayFromJSON(duration(TimeUnit::MILLI), "[0, 2]")));
+ // Different units, cast value_set to values
+ CheckIsIn(ArrayFromJSON(duration(TimeUnit::SECOND), "[0, 2]"),
+ ArrayFromJSON(duration(TimeUnit::MILLI), "[0, 1, 2000]"), "[true, true]");
Review Comment:
Updated the test case. Since only safe casts are allowed, it will first try casting [1ms, 2ms, 2000ms] to [0s, 2s] and fail, then cast the other way around and succeed. The result is [null, true] as expected.
##########
cpp/src/arrow/compute/kernels/scalar_set_lookup.cc:
##########
@@ -215,16 +221,24 @@ struct InitStateVisitor {
return Status::Invalid("Array type didn't match type of values set: ", *arg_type,
" vs ", *options.value_set.type());
}
+
if (!options.value_set.is_arraylike()) {
return Status::Invalid("Set lookup value set must be Array or ChunkedArray");
} else if (!options.value_set.type()->Equals(*arg_type)) {
- ARROW_ASSIGN_OR_RAISE(
- options.value_set,
+ auto cast_result =
Cast(options.value_set, CastOptions::Safe(arg_type.GetSharedPtr()),
- ctx->exec_context()));
+ ctx->exec_context());
+ if (cast_result.ok()) {
+ options.value_set = *cast_result;
+ } else if (CanCast(*arg_type.type, *options.value_set.type())) {
+ // Will try to cast input array to value set type during kernel exec
+ } else {
+ return Status::Invalid("Input type doesn't match type of values set: ", *arg_type,
Review Comment:
Created #36345
##########
cpp/src/arrow/compute/kernels/scalar_set_lookup.cc:
##########
@@ -359,13 +386,11 @@ struct IsInVisitor {
}
template <typename Type>
- Status ProcessIsIn() {
+ Status ProcessIsIn(const SetLookupState<Type>& state, const ArraySpan& data) {
Review Comment:
Changed `data` to `input`
##########
cpp/src/arrow/compute/kernels/scalar_set_lookup_test.cc:
##########
@@ -822,6 +822,37 @@ TEST_F(TestIndexInKernel, Boolean) {
CheckIndexIn(boolean(), "[null, null, null, null]", "[null]", "[0, 0, 0, 0]");
}
+TEST_F(TestIndexInKernel, ImplicitlyCastValueSet) {
+ auto input = ArrayFromJSON(int8(), "[0, 1, 2, 3, 4, 5, 6, 7, 8]");
+
+ SetLookupOptions opts{ArrayFromJSON(int32(), "[2, 3, 5, 7]")};
+ ASSERT_OK_AND_ASSIGN(Datum out, CallFunction("index_in", {input}, &opts));
+
+ auto expected = ArrayFromJSON(int32(), ("[null, null, 0, 1, null,"
+ "2, null, 3, null]"));
+ AssertArraysEqual(*expected, *out.make_array());
+
+ // Although value_set cannot be cast to int8, but int8 is castable to float
+ CheckIndexIn(input, ArrayFromJSON(float32(), "[1.0, 2.5, 3.1, 5.0]"),
+ "[null, 0, null, null, null, 3, null, null, null]");
+
+ // Allow implicit casts between binary types...
+ CheckIndexIn(ArrayFromJSON(binary(), R"(["aaa", "bbb", "ccc", null, "bbb"])"),
+ ArrayFromJSON(fixed_size_binary(3), R"(["aaa", "bbb"])"),
+ "[0, 1, null, null, 1]");
+ CheckIndexIn(ArrayFromJSON(utf8(), R"(["aaa", "bbb", "ccc", null, "bbb"])"),
+ ArrayFromJSON(large_utf8(), R"(["aaa", "bbb"])"), "[0, 1, null, null, 1]");
Review Comment:
Done. I also explicitly forbid casting non-binary to binary, same as the tests down below.
##########
cpp/src/arrow/compute/kernels/scalar_set_lookup.cc:
##########
@@ -263,11 +277,8 @@ struct IndexInVisitor {
}
template <typename Type>
- Status ProcessIndexIn() {
+ Status ProcessIndexIn(const SetLookupState<Type>& state, const ArraySpan& data) {
Review Comment:
Changed `data` to `input`
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [arrow] pitrou commented on a diff in pull request #36204: GH-36203: [C++] Support casting in both ways for is_in and index_in
Posted by "pitrou (via GitHub)" <gi...@apache.org>.
pitrou commented on code in PR #36204:
URL: https://github.com/apache/arrow/pull/36204#discussion_r1243799859
##########
cpp/src/arrow/compute/kernels/scalar_set_lookup_test.cc:
##########
@@ -126,9 +126,9 @@ TEST_F(TestIsInKernel, ImplicitlyCastValueSet) {
"true, false, true, false]"));
AssertArraysEqual(*expected, *out.make_array());
- // fails; value_set cannot be cast to int8
- opts = SetLookupOptions{ArrayFromJSON(float32(), "[2.5, 3.1, 5.0]")};
- ASSERT_RAISES(Invalid, CallFunction("is_in", {input}, &opts));
+ // Although value_set cannot be cast to int8, but int8 is castable to float
Review Comment:
```suggestion
// value_set cannot be cast to int8, but int8 is castable to float
```
##########
cpp/src/arrow/compute/kernels/scalar_set_lookup_test.cc:
##########
@@ -822,6 +822,37 @@ TEST_F(TestIndexInKernel, Boolean) {
CheckIndexIn(boolean(), "[null, null, null, null]", "[null]", "[0, 0, 0, 0]");
}
+TEST_F(TestIndexInKernel, ImplicitlyCastValueSet) {
+ auto input = ArrayFromJSON(int8(), "[0, 1, 2, 3, 4, 5, 6, 7, 8]");
+
+ SetLookupOptions opts{ArrayFromJSON(int32(), "[2, 3, 5, 7]")};
+ ASSERT_OK_AND_ASSIGN(Datum out, CallFunction("index_in", {input}, &opts));
+
+ auto expected = ArrayFromJSON(int32(), ("[null, null, 0, 1, null,"
+ "2, null, 3, null]"));
+ AssertArraysEqual(*expected, *out.make_array());
+
+ // Although value_set cannot be cast to int8, but int8 is castable to float
+ CheckIndexIn(input, ArrayFromJSON(float32(), "[1.0, 2.5, 3.1, 5.0]"),
+ "[null, 0, null, null, null, 3, null, null, null]");
+
+ // Allow implicit casts between binary types...
+ CheckIndexIn(ArrayFromJSON(binary(), R"(["aaa", "bbb", "ccc", null, "bbb"])"),
+ ArrayFromJSON(fixed_size_binary(3), R"(["aaa", "bbb"])"),
+ "[0, 1, null, null, 1]");
+ CheckIndexIn(ArrayFromJSON(utf8(), R"(["aaa", "bbb", "ccc", null, "bbb"])"),
+ ArrayFromJSON(large_utf8(), R"(["aaa", "bbb"])"), "[0, 1, null, null, 1]");
Review Comment:
Can you check the other way round as well for both checks above?
##########
cpp/src/arrow/compute/kernels/scalar_set_lookup.cc:
##########
@@ -359,13 +386,11 @@ struct IsInVisitor {
}
template <typename Type>
- Status ProcessIsIn() {
+ Status ProcessIsIn(const SetLookupState<Type>& state, const ArraySpan& data) {
Review Comment:
Same remark.
##########
cpp/src/arrow/compute/kernels/scalar_set_lookup.cc:
##########
@@ -263,11 +277,8 @@ struct IndexInVisitor {
}
template <typename Type>
- Status ProcessIndexIn() {
+ Status ProcessIndexIn(const SetLookupState<Type>& state, const ArraySpan& data) {
Review Comment:
The `data` argument here shadows the `data` member variable, which is confusing and error-prone. Do you need it?
##########
cpp/src/arrow/compute/kernels/scalar_set_lookup.cc:
##########
@@ -215,16 +221,24 @@ struct InitStateVisitor {
return Status::Invalid("Array type didn't match type of values set: ", *arg_type,
" vs ", *options.value_set.type());
}
+
if (!options.value_set.is_arraylike()) {
return Status::Invalid("Set lookup value set must be Array or ChunkedArray");
} else if (!options.value_set.type()->Equals(*arg_type)) {
- ARROW_ASSIGN_OR_RAISE(
- options.value_set,
+ auto cast_result =
Cast(options.value_set, CastOptions::Safe(arg_type.GetSharedPtr()),
- ctx->exec_context()));
+ ctx->exec_context());
+ if (cast_result.ok()) {
+ options.value_set = *cast_result;
+ } else if (CanCast(*arg_type.type, *options.value_set.type())) {
+ // Will try to cast input array to value set type during kernel exec
+ } else {
+ return Status::Invalid("Input type doesn't match type of values set: ", *arg_type,
Review Comment:
Ideally, this should have been `TypeError` instead of `Invalid` here, but let's keep consistent for now.
Perhaps open an issue to change this later?
##########
cpp/src/arrow/compute/kernels/scalar_set_lookup_test.cc:
##########
@@ -253,11 +253,11 @@ TEST_F(TestIsInKernel, TimeDuration) {
"[true, false, false, true, true]", /*skip_nulls=*/true);
}
- // Different units, invalid cast
- ASSERT_RAISES(Invalid, IsIn(ArrayFromJSON(duration(TimeUnit::SECOND), "[0, 1, 2]"),
- ArrayFromJSON(duration(TimeUnit::MILLI), "[0, 2]")));
+ // Different units, cast value_set to values
+ CheckIsIn(ArrayFromJSON(duration(TimeUnit::SECOND), "[0, 2]"),
+ ArrayFromJSON(duration(TimeUnit::MILLI), "[0, 1, 2000]"), "[true, true]");
Review Comment:
Ok, but you're choosing an easy case here.
What if values = `[0s, 2s]` and value_set = `[1ms, 2ms, 2000ms]`? Will it fail because of truncation during casting? Will it return an incorrect value? Will it try casting the other way round?
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [arrow] js8544 commented on a diff in pull request #36204: GH-36203: [C++] Support casting in both ways for is_in and index_in
Posted by "js8544 (via GitHub)" <gi...@apache.org>.
js8544 commented on code in PR #36204:
URL: https://github.com/apache/arrow/pull/36204#discussion_r1245037561
##########
cpp/src/arrow/compute/kernels/scalar_set_lookup_test.cc:
##########
@@ -779,11 +794,11 @@ TEST_F(TestIndexInKernel, TimeDuration) {
CheckIndexIn(duration(TimeUnit::SECOND), "[null, null, null, null]", "[null]",
"[0, 0, 0, 0]");
- // Different units, invalid cast
- ASSERT_RAISES(Invalid, IndexIn(ArrayFromJSON(duration(TimeUnit::SECOND), "[0, 1, 2]"),
- ArrayFromJSON(duration(TimeUnit::MILLI), "[0, 2]")));
+ // Different units, cast value_set to values
Review Comment:
Done
##########
cpp/src/arrow/compute/kernels/scalar_set_lookup_test.cc:
##########
@@ -822,6 +837,50 @@ TEST_F(TestIndexInKernel, Boolean) {
CheckIndexIn(boolean(), "[null, null, null, null]", "[null]", "[0, 0, 0, 0]");
}
+TEST_F(TestIndexInKernel, ImplicitlyCastValueSet) {
+ auto input = ArrayFromJSON(int8(), "[0, 1, 2, 3, 4, 5, 6, 7, 8]");
+
+ SetLookupOptions opts{ArrayFromJSON(int32(), "[2, 3, 5, 7]")};
+ ASSERT_OK_AND_ASSIGN(Datum out, CallFunction("index_in", {input}, &opts));
+
+ auto expected = ArrayFromJSON(int32(), ("[null, null, 0, 1, null,"
+ "2, null, 3, null]"));
+ AssertArraysEqual(*expected, *out.make_array());
+
+ // Although value_set cannot be cast to int8, but int8 is castable to float
+ CheckIndexIn(input, ArrayFromJSON(float32(), "[1.0, 2.5, 3.1, 5.0]"),
+ "[null, 0, null, null, null, 3, null, null, null]");
+
+ // Allow implicit casts between binary types...
+ CheckIndexIn(ArrayFromJSON(binary(), R"(["aaa", "bbb", "ccc", null, "bbb"])"),
Review Comment:
Done
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [arrow] js8544 commented on a diff in pull request #36204: GH-36203: [C++] Support casting in both ways for is_in and index_in
Posted by "js8544 (via GitHub)" <gi...@apache.org>.
js8544 commented on code in PR #36204:
URL: https://github.com/apache/arrow/pull/36204#discussion_r1245037302
##########
cpp/src/arrow/compute/kernels/scalar_set_lookup_test.cc:
##########
@@ -253,11 +268,11 @@ TEST_F(TestIsInKernel, TimeDuration) {
"[true, false, false, true, true]", /*skip_nulls=*/true);
}
- // Different units, invalid cast
- ASSERT_RAISES(Invalid, IsIn(ArrayFromJSON(duration(TimeUnit::SECOND), "[0, 1, 2]"),
- ArrayFromJSON(duration(TimeUnit::MILLI), "[0, 2]")));
+ // Different units, cast value_set to values
+ CheckIsIn(ArrayFromJSON(duration(TimeUnit::SECOND), "[0, 2]"),
+ ArrayFromJSON(duration(TimeUnit::MILLI), "[1, 2, 2000]"), "[false, true]");
Review Comment:
Done
##########
cpp/src/arrow/compute/kernels/scalar_set_lookup_test.cc:
##########
@@ -253,11 +268,11 @@ TEST_F(TestIsInKernel, TimeDuration) {
"[true, false, false, true, true]", /*skip_nulls=*/true);
}
- // Different units, invalid cast
- ASSERT_RAISES(Invalid, IsIn(ArrayFromJSON(duration(TimeUnit::SECOND), "[0, 1, 2]"),
- ArrayFromJSON(duration(TimeUnit::MILLI), "[0, 2]")));
+ // Different units, cast value_set to values
Review Comment:
Comment updated.
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [arrow] pitrou commented on a diff in pull request #36204: GH-36203: [C++] Support casting in both ways for is_in and index_in
Posted by "pitrou (via GitHub)" <gi...@apache.org>.
pitrou commented on code in PR #36204:
URL: https://github.com/apache/arrow/pull/36204#discussion_r1244998090
##########
cpp/src/arrow/compute/kernels/scalar_set_lookup_test.cc:
##########
@@ -253,11 +268,11 @@ TEST_F(TestIsInKernel, TimeDuration) {
"[true, false, false, true, true]", /*skip_nulls=*/true);
}
- // Different units, invalid cast
- ASSERT_RAISES(Invalid, IsIn(ArrayFromJSON(duration(TimeUnit::SECOND), "[0, 1, 2]"),
- ArrayFromJSON(duration(TimeUnit::MILLI), "[0, 2]")));
+ // Different units, cast value_set to values
+ CheckIsIn(ArrayFromJSON(duration(TimeUnit::SECOND), "[0, 2]"),
+ ArrayFromJSON(duration(TimeUnit::MILLI), "[1, 2, 2000]"), "[false, true]");
Review Comment:
More explicitly ensure that the cast actually happens (either 2 or 2000 could be the successful match...)
```suggestion
CheckIsIn(ArrayFromJSON(duration(TimeUnit::SECOND), "[0, 1, 2]"),
ArrayFromJSON(duration(TimeUnit::MILLI), "[1, 2, 2000]"), "[false, false, true]");
```
##########
cpp/src/arrow/compute/kernels/scalar_set_lookup_test.cc:
##########
@@ -126,25 +126,40 @@ TEST_F(TestIsInKernel, ImplicitlyCastValueSet) {
"true, false, true, false]"));
AssertArraysEqual(*expected, *out.make_array());
- // fails; value_set cannot be cast to int8
- opts = SetLookupOptions{ArrayFromJSON(float32(), "[2.5, 3.1, 5.0]")};
- ASSERT_RAISES(Invalid, CallFunction("is_in", {input}, &opts));
+ // value_set cannot be casted to int8, but int8 is castable to float
+ CheckIsIn(input, ArrayFromJSON(float32(), "[1.0, 2.5, 3.1, 5.0]"),
+ "[false, true, false, false, false, true, false, false, false]");
// Allow implicit casts between binary types...
CheckIsIn(ArrayFromJSON(binary(), R"(["aaa", "bbb", "ccc", null, "bbb"])"),
ArrayFromJSON(fixed_size_binary(3), R"(["aaa", "bbb"])"),
"[true, true, false, false, true]");
+ CheckIsIn(ArrayFromJSON(fixed_size_binary(3), R"(["aaa", "bbb", "ccc", null, "bbb"])"),
+ ArrayFromJSON(binary(), R"(["aaa", "bbb"])"),
+ "[true, true, false, false, true]");
Review Comment:
Can you perhaps make things more interesting by making entries that are not 3-bytes long?
```suggestion
CheckIsIn(ArrayFromJSON(binary(), R"(["aaa", "bb", "ccc", null, "bbb"])"),
ArrayFromJSON(fixed_size_binary(3), R"(["aaa", "bbb"])"),
"[true, false, false, false, true]");
CheckIsIn(ArrayFromJSON(fixed_size_binary(3), R"(["aaa", "bbb", "ccc", null, "bbb"])"),
ArrayFromJSON(binary(), R"(["aa", "bbb"])"),
"[false, true, false, false, true]");
```
##########
cpp/src/arrow/compute/kernels/scalar_set_lookup_test.cc:
##########
@@ -253,11 +268,11 @@ TEST_F(TestIsInKernel, TimeDuration) {
"[true, false, false, true, true]", /*skip_nulls=*/true);
}
- // Different units, invalid cast
- ASSERT_RAISES(Invalid, IsIn(ArrayFromJSON(duration(TimeUnit::SECOND), "[0, 1, 2]"),
- ArrayFromJSON(duration(TimeUnit::MILLI), "[0, 2]")));
+ // Different units, cast value_set to values
Review Comment:
Is the comment correct? It will attempt to cast value_set to values, fail because of truncation, and then cast values to value_set.
##########
cpp/src/arrow/compute/kernels/scalar_set_lookup_test.cc:
##########
@@ -779,11 +794,11 @@ TEST_F(TestIndexInKernel, TimeDuration) {
CheckIndexIn(duration(TimeUnit::SECOND), "[null, null, null, null]", "[null]",
"[0, 0, 0, 0]");
- // Different units, invalid cast
- ASSERT_RAISES(Invalid, IndexIn(ArrayFromJSON(duration(TimeUnit::SECOND), "[0, 1, 2]"),
- ArrayFromJSON(duration(TimeUnit::MILLI), "[0, 2]")));
+ // Different units, cast value_set to values
Review Comment:
Similar comments here :-)
##########
cpp/src/arrow/compute/kernels/scalar_set_lookup_test.cc:
##########
@@ -822,6 +837,50 @@ TEST_F(TestIndexInKernel, Boolean) {
CheckIndexIn(boolean(), "[null, null, null, null]", "[null]", "[0, 0, 0, 0]");
}
+TEST_F(TestIndexInKernel, ImplicitlyCastValueSet) {
+ auto input = ArrayFromJSON(int8(), "[0, 1, 2, 3, 4, 5, 6, 7, 8]");
+
+ SetLookupOptions opts{ArrayFromJSON(int32(), "[2, 3, 5, 7]")};
+ ASSERT_OK_AND_ASSIGN(Datum out, CallFunction("index_in", {input}, &opts));
+
+ auto expected = ArrayFromJSON(int32(), ("[null, null, 0, 1, null,"
+ "2, null, 3, null]"));
+ AssertArraysEqual(*expected, *out.make_array());
+
+ // Although value_set cannot be cast to int8, but int8 is castable to float
+ CheckIndexIn(input, ArrayFromJSON(float32(), "[1.0, 2.5, 3.1, 5.0]"),
+ "[null, 0, null, null, null, 3, null, null, null]");
+
+ // Allow implicit casts between binary types...
+ CheckIndexIn(ArrayFromJSON(binary(), R"(["aaa", "bbb", "ccc", null, "bbb"])"),
Review Comment:
Similar suggestions here as for `is_in`.
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [arrow] github-actions[bot] commented on pull request #36204: GH-36203: [C++] Support casting in both ways for is_in and index_in
Posted by "github-actions[bot] (via GitHub)" <gi...@apache.org>.
github-actions[bot] commented on PR #36204:
URL: https://github.com/apache/arrow/pull/36204#issuecomment-1601054800
:warning: GitHub issue #36203 **has been automatically assigned in GitHub** to PR creator.
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [arrow] conbench-apache-arrow[bot] commented on pull request #36204: GH-36203: [C++] Support casting in both ways for is_in and index_in
Posted by "conbench-apache-arrow[bot] (via GitHub)" <gi...@apache.org>.
conbench-apache-arrow[bot] commented on PR #36204:
URL: https://github.com/apache/arrow/pull/36204#issuecomment-1616436646
Conbench analyzed the 6 benchmark runs on commit `6f3bd252`.
There were 9 benchmark results indicating a performance regression:
- Commit Run on `arm64-t4g-linux-compute` at [2023-06-28 17:26:18Z](http://conbench.ursa.dev/compare/runs/8beeb07705694f5b81874a97d1b949b2...8edb4f8f196640aa8684de44a71d0559/)
- [params=num_cols:8/is_partial:1/real_time, source=cpp-micro, suite=arrow-ipc-read-write-benchmark](http://conbench.ursa.dev/compare/benchmarks/0649c56261ec7c49800010ec178d0ad7...0649c6d927967fd580000fafc89392da)
- Commit Run on `arm64-m6g-linux-compute` at [2023-06-28 17:11:20Z](http://conbench.ursa.dev/compare/runs/857a0eece264456e95485081709f876c...a583c3d2c83447c9be0c7869ff0985c6/)
- [params=4, source=cpp-micro, suite=arrow-compute-scalar-set-lookup-benchmark](http://conbench.ursa.dev/compare/benchmarks/0649c542e0047a718000d5f05765eb7a...0649c69d704d7fe5800030c9836662f0)
- and 7 more (see the report linked below)
The [full Conbench report](https://github.com/apache/arrow/runs/14716219290) has more details.
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [arrow] js8544 commented on pull request #36204: GH-36203: [C++] Support casting in both ways for is_in and index_in
Posted by "js8544 (via GitHub)" <gi...@apache.org>.
js8544 commented on PR #36204:
URL: https://github.com/apache/arrow/pull/36204#issuecomment-1608938811
> @js8544 Is this ready for review?
Yes, please!
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [arrow] pitrou merged pull request #36204: GH-36203: [C++] Support casting in both ways for is_in and index_in
Posted by "pitrou (via GitHub)" <gi...@apache.org>.
pitrou merged PR #36204:
URL: https://github.com/apache/arrow/pull/36204
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [arrow] js8544 commented on a diff in pull request #36204: GH-36203: [C++] Support casting in both ways for is_in and index_in
Posted by "js8544 (via GitHub)" <gi...@apache.org>.
js8544 commented on code in PR #36204:
URL: https://github.com/apache/arrow/pull/36204#discussion_r1245037013
##########
cpp/src/arrow/compute/kernels/scalar_set_lookup_test.cc:
##########
@@ -126,25 +126,40 @@ TEST_F(TestIsInKernel, ImplicitlyCastValueSet) {
"true, false, true, false]"));
AssertArraysEqual(*expected, *out.make_array());
- // fails; value_set cannot be cast to int8
- opts = SetLookupOptions{ArrayFromJSON(float32(), "[2.5, 3.1, 5.0]")};
- ASSERT_RAISES(Invalid, CallFunction("is_in", {input}, &opts));
+ // value_set cannot be casted to int8, but int8 is castable to float
+ CheckIsIn(input, ArrayFromJSON(float32(), "[1.0, 2.5, 3.1, 5.0]"),
+ "[false, true, false, false, false, true, false, false, false]");
// Allow implicit casts between binary types...
CheckIsIn(ArrayFromJSON(binary(), R"(["aaa", "bbb", "ccc", null, "bbb"])"),
ArrayFromJSON(fixed_size_binary(3), R"(["aaa", "bbb"])"),
"[true, true, false, false, true]");
+ CheckIsIn(ArrayFromJSON(fixed_size_binary(3), R"(["aaa", "bbb", "ccc", null, "bbb"])"),
+ ArrayFromJSON(binary(), R"(["aaa", "bbb"])"),
+ "[true, true, false, false, true]");
Review Comment:
Done
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [arrow] pitrou commented on pull request #36204: GH-36203: [C++] Support casting in both ways for is_in and index_in
Posted by "pitrou (via GitHub)" <gi...@apache.org>.
pitrou commented on PR #36204:
URL: https://github.com/apache/arrow/pull/36204#issuecomment-1608937149
@js8544 Is this ready for review?
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org