[feature](function) Support for aggregate function foreach combiner #31526
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Thank you for your contribution to Apache Doris. |
|
run buildall |
![github-actions[bot]](https://avatars.githubusercontent.com/in/15368?s=60&v=4){kind=small}
| const size_t nested_size_of_data; | ||
| const size_t num_arguments; | ||
|
|
||
| AggregateFunctionForEachData& ensureAggregateData(AggregateDataPtr __restrict place, |
There was a problem hiding this comment.
warning: pointer parameter 'place' can be pointer to const [readability-non-const-parameter]
Suggested change
| AggregateFunctionForEachData& ensureAggregateData(AggregateDataPtr __restrict place, | |
| AggregateFunctionForEachData& ensureAggregateData(const AggregateDataPtr __restrict place, |
| "Aggregate function {} require at least one argument", get_name()); | ||
| } | ||
| } | ||
| void set_version(const int version_) override { |
There was a problem hiding this comment.
warning: method 'set_version' can be made static [readability-convert-member-functions-to-static]
Suggested change
| void set_version(const int version_) override { | |
| static void set_version(const int version_) override { |
| return std::make_shared<DataTypeArray>(make_nullable(nested_function->get_return_type())); | ||
| } | ||
|
|
||
| void destroy(AggregateDataPtr __restrict place) const noexcept override { |
There was a problem hiding this comment.
warning: pointer parameter 'place' can be pointer to const [readability-non-const-parameter]
Suggested change
| void destroy(AggregateDataPtr __restrict place) const noexcept override { | |
| void destroy(const AggregateDataPtr __restrict place) const noexcept override { |
| @@ -78,6 +86,21 @@ class AggregateFunctionSimpleFactory { | |||
| } | |||
| } | |||
|
|
|||
| void register_foreach_function_combinator(const Creator& creator, const std::string& suffix, | |||
There was a problem hiding this comment.
warning: method 'register_foreach_function_combinator' can be made static [readability-convert-member-functions-to-static]
Suggested change
| void register_foreach_function_combinator(const Creator& creator, const std::string& suffix, | |
| static void register_foreach_function_combinator(const Creator& creator, const std::string& suffix, |
TPC-H: Total hot run time: 37760 ms |
TPC-DS: Total hot run time: 169564 ms |
ClickBench: Total hot run time: 32.49 s |
|
Load test result on machine: 'aliyun_ecs.c7a.8xlarge_32C64G' |
|
TeamCity be ut coverage result: |
|
run buildall |
morrySnow
reviewed
Feb 28, 2024
There was a problem hiding this comment.
since this class is not only process agg state function, u should rename it
morrySnow
reviewed
Feb 28, 2024
|
|
||
|
|
||
| qt_sql """ | ||
| select sum_foreach(a) , count_foreach(a) , group_concat_foreach(s) from table_with_null; |
There was a problem hiding this comment.
if all agg function could have foreach suffix. u should add all agg function with foreach suffix tests. and each function should cover its all signatures. refer to: regression-test/suites/nereids_function_p0/agg_function
|
run buildall |
|
TeamCity be ut coverage result: |
TPC-H: Total hot run time: 37999 ms |
TPC-DS: Total hot run time: 169278 ms |
ClickBench: Total hot run time: 31.44 s |
|
Load test result on machine: 'aliyun_ecs.c7a.8xlarge_32C64G' |
|
run feut |
924060929
reviewed
Feb 29, 2024
|
|
||
| @Override | ||
| public boolean nullable() { | ||
| return true; |
There was a problem hiding this comment.
why implement AlwaysNotNullable but override nullable method like AlwaysNullable
|
run buildall |
TPC-H: Total hot run time: 38868 ms |
|
TeamCity be ut coverage result: |
HappenLee
reviewed
Mar 2, 2024
| void register_aggregate_function_combinator_foreach(AggregateFunctionSimpleFactory& factory) { | ||
| AggregateFunctionCreator creator = [&](const std::string& name, const DataTypes& types, | ||
| const bool result_is_nullable) -> AggregateFunctionPtr { | ||
| const std::string suffix = AggregateFunctionForEach::AGG_FOREACH_SUFFIX; |
HappenLee
reviewed
Mar 2, 2024
| const bool result_is_nullable) -> AggregateFunctionPtr { | ||
| const std::string suffix = AggregateFunctionForEach::AGG_FOREACH_SUFFIX; | ||
| DataTypes transform_arguments; | ||
| for (auto t : types) { |
HappenLee
reviewed
Mar 2, 2024
| factory.get(nested_function_name, transform_arguments, result_is_nullable); | ||
| if (!nested_function) { | ||
| throw Exception(ErrorCode::INTERNAL_ERROR, | ||
| "The combiner did not find a corresponding function. nested function " |
HappenLee
reviewed
Mar 2, 2024
| const size_t nested_size_of_data; | ||
| const size_t num_arguments; | ||
|
|
||
| AggregateFunctionForEachData& ensureAggregateData(AggregateDataPtr __restrict place, |
Mryange
force-pushed
the
agg-foreach
branch
2 times, most recently
from
3c99923 to
7cdc52e
Compare
| const size_t nested_size_of_data; | ||
| const size_t num_arguments; | ||
|
|
||
| AggregateFunctionForEachData& ensure_aggregate_data(AggregateDataPtr __restrict place, |
There was a problem hiding this comment.
warning: pointer parameter 'place' can be pointer to const [readability-non-const-parameter]
Suggested change
| AggregateFunctionForEachData& ensure_aggregate_data(AggregateDataPtr __restrict place, | |
| AggregateFunctionForEachData& ensure_aggregate_data(const AggregateDataPtr __restrict place, |
HappenLee
reviewed
Mar 2, 2024
| } | ||
|
|
||
| for (i = 0; i < old_size; ++i) { | ||
| nested_function->merge(&new_state[i * nested_size_of_data], |
There was a problem hiding this comment.
destory the old_state mem
HappenLee
reviewed
Mar 2, 2024
| const auto& first_array_column = assert_cast<const ColumnArray&>(*columns[0]); | ||
| const auto& offsets = first_array_column.get_offsets(); | ||
|
|
||
| size_t begin = offsets[row_num - 1]; |
There was a problem hiding this comment.
what happen if row_num == 0? maybe overflow access mem offset
There was a problem hiding this comment.
The index of offsets can be -1.
// [[2,1,5,9,1], [1,2,4]] --> data column [2,1,5,9,1,1,2,4], offset[-1] = 0, offset[0] = 5, offset[1] = 8
There was a problem hiding this comment.
Warning ！！！row_num is size_t -1 = 18446744073709551615
|
run buildall |
|
TeamCity be ut coverage result: |
|
run buildall |
|
TeamCity be ut coverage result: |
| @@ -65,7 +65,7 @@ class AggregateFunctionCount final | |||
|
|
|||
| DataTypePtr get_return_type() const override { return std::make_shared<DataTypeInt64>(); } | |||
|
|
|||
| void add(AggregateDataPtr __restrict place, const IColumn**, size_t, Arena*) const override { | |||
| void add(AggregateDataPtr __restrict place, const IColumn**, ssize_t, Arena*) const override { | |||
There was a problem hiding this comment.
warning: method 'add' can be made static [readability-convert-member-functions-to-static]
Suggested change
| void add(AggregateDataPtr __restrict place, const IColumn**, ssize_t, Arena*) const override { | |
| static void add(AggregateDataPtr __restrict place, const IColumn**, ssize_t, Arena*) override { |
| @@ -65,7 +65,7 @@ | |||
|
|
|||
| DataTypePtr get_return_type() const override { return std::make_shared<DataTypeInt64>(); } | |||
|
|
|||
| void add(AggregateDataPtr __restrict place, const IColumn**, size_t, Arena*) const override { | |||
| void add(AggregateDataPtr __restrict place, const IColumn**, ssize_t, Arena*) const override { | |||
There was a problem hiding this comment.
warning: pointer parameter 'place' can be pointer to const [readability-non-const-parameter]
Suggested change
| void add(AggregateDataPtr __restrict place, const IColumn**, ssize_t, Arena*) const override { | |
| void add(const AggregateDataPtr __restrict place, const IColumn**, ssize_t, Arena*) const override { |
Comment on lines
+197
to
198
| void add(AggregateDataPtr __restrict place, const IColumn** columns, ssize_t row_num, | ||
| Arena*) const override { |
There was a problem hiding this comment.
warning: method 'add' can be made static [readability-convert-member-functions-to-static]
Suggested change
| void add(AggregateDataPtr __restrict place, const IColumn** columns, ssize_t row_num, | |
| Arena*) const override { | |
| static void add(AggregateDataPtr __restrict place, const IColumn** columns, ssize_t row_num, | |
| Arena*) override { |
| @@ -194,7 +194,7 @@ | |||
|
|
|||
| DataTypePtr get_return_type() const override { return std::make_shared<DataTypeInt64>(); } | |||
|
|
|||
| void add(AggregateDataPtr __restrict place, const IColumn** columns, size_t row_num, | |||
| void add(AggregateDataPtr __restrict place, const IColumn** columns, ssize_t row_num, | |||
There was a problem hiding this comment.
warning: pointer parameter 'place' can be pointer to const [readability-non-const-parameter]
Suggested change
| void add(AggregateDataPtr __restrict place, const IColumn** columns, ssize_t row_num, | |
| void add(const AggregateDataPtr __restrict place, const IColumn** columns, ssize_t row_num, |
Comment on lines
+298
to
299
| void add(AggregateDataPtr __restrict place, const IColumn** columns, ssize_t row_num, | ||
| Arena*) const override { |
There was a problem hiding this comment.
warning: method 'add' can be made static [readability-convert-member-functions-to-static]
Suggested change
| void add(AggregateDataPtr __restrict place, const IColumn** columns, ssize_t row_num, | |
| Arena*) const override { | |
| static void add(AggregateDataPtr __restrict place, const IColumn** columns, ssize_t row_num, | |
| Arena*) override { |
| @@ -65,7 +65,7 @@ | |||
|
|
|||
| DataTypePtr get_return_type() const override { return std::make_shared<DataTypeInt64>(); } | |||
|
|
|||
| void add(AggregateDataPtr place, const IColumn**, size_t, Arena*) const override { | |||
| void add(AggregateDataPtr place, const IColumn**, ssize_t, Arena*) const override { | |||
There was a problem hiding this comment.
warning: pointer parameter 'place' can be pointer to const [readability-non-const-parameter]
Suggested change
| void add(AggregateDataPtr place, const IColumn**, ssize_t, Arena*) const override { | |
| void add(const AggregateDataPtr place, const IColumn**, ssize_t, Arena*) const override { |
| @@ -103,7 +103,7 @@ | |||
|
|
|||
| DataTypePtr get_return_type() const override { return std::make_shared<DataTypeInt64>(); } | |||
|
|
|||
| void add(AggregateDataPtr place, const IColumn**, size_t, Arena*) const override { | |||
| void add(AggregateDataPtr place, const IColumn**, ssize_t, Arena*) const override { | |||
There was a problem hiding this comment.
warning: method 'add' can be made static [readability-convert-member-functions-to-static]
Suggested change
| void add(AggregateDataPtr place, const IColumn**, ssize_t, Arena*) const override { | |
| static void add(AggregateDataPtr place, const IColumn**, ssize_t, Arena*) override { |
| @@ -103,7 +103,7 @@ | |||
|
|
|||
| DataTypePtr get_return_type() const override { return std::make_shared<DataTypeInt64>(); } | |||
|
|
|||
| void add(AggregateDataPtr place, const IColumn**, size_t, Arena*) const override { | |||
| void add(AggregateDataPtr place, const IColumn**, ssize_t, Arena*) const override { | |||
There was a problem hiding this comment.
warning: pointer parameter 'place' can be pointer to const [readability-non-const-parameter]
Suggested change
| void add(AggregateDataPtr place, const IColumn**, ssize_t, Arena*) const override { | |
| void add(const AggregateDataPtr place, const IColumn**, ssize_t, Arena*) const override { |
| @@ -148,7 +148,7 @@ | |||
|
|
|||
| DataTypePtr get_return_type() const override { return std::make_shared<DataTypeInt64>(); } | |||
|
|
|||
| void add(AggregateDataPtr place, const IColumn**, size_t, Arena*) const override { | |||
| void add(AggregateDataPtr place, const IColumn**, ssize_t, Arena*) const override { | |||
There was a problem hiding this comment.
warning: method 'add' can be made static [readability-convert-member-functions-to-static]
Suggested change
| void add(AggregateDataPtr place, const IColumn**, ssize_t, Arena*) const override { | |
| static void add(AggregateDataPtr place, const IColumn**, ssize_t, Arena*) override { |
| @@ -148,7 +148,7 @@ | |||
|
|
|||
| DataTypePtr get_return_type() const override { return std::make_shared<DataTypeInt64>(); } | |||
|
|
|||
| void add(AggregateDataPtr place, const IColumn**, size_t, Arena*) const override { | |||
| void add(AggregateDataPtr place, const IColumn**, ssize_t, Arena*) const override { | |||
There was a problem hiding this comment.
warning: pointer parameter 'place' can be pointer to const [readability-non-const-parameter]
Suggested change
| void add(AggregateDataPtr place, const IColumn**, ssize_t, Arena*) const override { | |
| void add(const AggregateDataPtr place, const IColumn**, ssize_t, Arena*) const override { |
|
run buildall |
|
TeamCity be ut coverage result: |
github-actions
bot
added
the
approved
|
PR approved by at least one committer and no changes requested. |
|
PR approved by anyone and no changes requested. |
zclllyybb
reviewed
Mar 5, 2024
Due to error with some functions' self-implemented features, there are currently some that cannot be used. Documentation will be added once all issues are addressed. |
yiguolei
pushed a commit
that referenced
this pull request
Mar 6, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Proposed changes
Currently not supported on Nereids, further follow-up is needed.
Further comments
If this is a relatively large or complex change, kick off the discussion at dev@doris.apache.org by explaining why you chose the solution you did and what alternatives you considered, etc...