Skip to content

[fix](parquet)Parse Parquet INT96 physical type data into a no-timezone datetime. #59514

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
hubgeter wants to merge 1 commit into
base: master
Choose a base branch
from
Open

[fix](parquet)Parse Parquet INT96 physical type data into a no-timezone datetime. #59514

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
49 changes: 49 additions & 0 deletions be/src/vec/exec/format/parquet/parquet_column_convert.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,6 +30,55 @@ namespace doris::vectorized::parquet {
#include "common/compile_check_begin.h"
const cctz::time_zone ConvertParams::utc0 = cctz::utc_time_zone();

void ConvertParams::init(const FieldSchema* field_schema_, const cctz::time_zone* ctz_) {
field_schema = field_schema_;
if (ctz_ != nullptr) {
ctz = ctz_;
}
const auto& schema = field_schema->parquet_schema;
if (field_schema->physical_type == tparquet::Type::INT96) {
ctz = &utc0;
} else if (schema.__isset.logicalType && schema.logicalType.__isset.TIMESTAMP) {
const auto& timestamp_info = schema.logicalType.TIMESTAMP;
if (!timestamp_info.isAdjustedToUTC) {
// should set timezone to utc+0
// Reference: https://github.com/apache/parquet-format/blob/master/LogicalTypes.md#instant-semantics-timestamps-normalized-to-utc
// If isAdjustedToUTC = false, the reader should display the same value no mater what local time zone is. For example:
// When a timestamp is stored as `1970-01-03 12:00:00`,
// if isAdjustedToUTC = true, UTC8 should read as `1970-01-03 20:00:00`, UTC6 should read as `1970-01-03 18:00:00`
// if isAdjustedToUTC = false, UTC8 and UTC6 should read as `1970-01-03 12:00:00`, which is the same as `1970-01-03 12:00:00` in UTC0
ctz = &utc0;
}
const auto& time_unit = timestamp_info.unit;
if (time_unit.__isset.MILLIS) {
second_mask = 1000;
scale_to_nano_factor = 1000000;
} else if (time_unit.__isset.MICROS) {
second_mask = 1000000;
scale_to_nano_factor = 1000;
} else if (time_unit.__isset.NANOS) {
second_mask = 1000000000;
scale_to_nano_factor = 1;
}
} else if (schema.__isset.converted_type) {
const auto& converted_type = schema.converted_type;
if (converted_type == tparquet::ConvertedType::TIMESTAMP_MILLIS) {
second_mask = 1000;
scale_to_nano_factor = 1000000;
} else if (converted_type == tparquet::ConvertedType::TIMESTAMP_MICROS) {
second_mask = 1000000;
scale_to_nano_factor = 1000;
}
}

if (ctz) {
VecDateTimeValue t;
t.from_unixtime(0, *ctz);
offset_days = t.day() == 31 ? -1 : 0;
}
is_type_compatibility = field_schema_->is_type_compatibility;
}

#define FOR_LOGICAL_DECIMAL_TYPES(M) \
M(TYPE_DECIMAL32) \
M(TYPE_DECIMAL64) \
Expand Down
47 changes: 1 addition & 46 deletions be/src/vec/exec/format/parquet/parquet_column_convert.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -71,52 +71,7 @@ struct ConvertParams {
}
}

void init(const FieldSchema* field_schema_, const cctz::time_zone* ctz_) {
field_schema = field_schema_;
if (ctz_ != nullptr) {
ctz = ctz_;
}
const auto& schema = field_schema->parquet_schema;
if (schema.__isset.logicalType && schema.logicalType.__isset.TIMESTAMP) {
const auto& timestamp_info = schema.logicalType.TIMESTAMP;
if (!timestamp_info.isAdjustedToUTC) {
// should set timezone to utc+0
// Reference: https://github.com/apache/parquet-format/blob/master/LogicalTypes.md#instant-semantics-timestamps-normalized-to-utc
// If isAdjustedToUTC = false, the reader should display the same value no mater what local time zone is. For example:
// When a timestamp is stored as `1970-01-03 12:00:00`,
// if isAdjustedToUTC = true, UTC8 should read as `1970-01-03 20:00:00`, UTC6 should read as `1970-01-03 18:00:00`
// if isAdjustedToUTC = false, UTC8 and UTC6 should read as `1970-01-03 12:00:00`, which is the same as `1970-01-03 12:00:00` in UTC0
ctz = &utc0;
}
const auto& time_unit = timestamp_info.unit;
if (time_unit.__isset.MILLIS) {
second_mask = 1000;
scale_to_nano_factor = 1000000;
} else if (time_unit.__isset.MICROS) {
second_mask = 1000000;
scale_to_nano_factor = 1000;
} else if (time_unit.__isset.NANOS) {
second_mask = 1000000000;
scale_to_nano_factor = 1;
}
} else if (schema.__isset.converted_type) {
const auto& converted_type = schema.converted_type;
if (converted_type == tparquet::ConvertedType::TIMESTAMP_MILLIS) {
second_mask = 1000;
scale_to_nano_factor = 1000000;
} else if (converted_type == tparquet::ConvertedType::TIMESTAMP_MICROS) {
second_mask = 1000000;
scale_to_nano_factor = 1000;
}
}

if (ctz) {
VecDateTimeValue t;
t.from_unixtime(0, *ctz);
offset_days = t.day() == 31 ? -1 : 0;
}
is_type_compatibility = field_schema_->is_type_compatibility;
}
void init(const FieldSchema* field_schema_, const cctz::time_zone* ctz_);
};

/**
Expand Down
Binary file modified regression-test/data/external_table_p0/tvf/test_hdfs_parquet_group0.out
View file
Open in desktop
Binary file not shown.
6 changes: 3 additions & 3 deletions regression-test/data/external_table_p0/tvf/test_hdfs_parquet_group1.out
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -90,9 +90,9 @@
[{"latitude":0, "longitude":180}, {"latitude":0, "longitude":0}]

-- !test_17 --
1 1880-01-01T15:58:41
2 1884-01-01T16:05:43
3 1990-01-01T16:00
1 1880-01-01T07:52:58
2 1884-01-01T08:00
3 1990-01-01T08:00

-- !test_18 --
[{"latitude":0, "longitude":0}, null, {"latitude":0, "longitude":180}]
Expand Down
96 changes: 48 additions & 48 deletions regression-test/data/external_table_p0/tvf/test_hdfs_parquet_group2.out
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -122,14 +122,14 @@ apple_banana_mango9
3 Cecilia 2022-11-16T02:32:09

-- !test_17 --
1001-01-07T17:07:47.172032 1001-01-07T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-08T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-09T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-10T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-11T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-12T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-13T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-14T17:07:47.172032
1001-01-07T09:02:04.172032 1001-01-07T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-08T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-09T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-10T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-11T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-12T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-13T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-14T09:02:04.172032

-- !test_18 --
0.00
Expand All @@ -151,14 +151,14 @@ apple_banana_mango9
2

-- !test_20 --
1001-01-07T17:07:47.172032 1001-01-07T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-08T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-09T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-10T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-11T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-12T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-13T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-14T17:07:47.172032
1001-01-07T09:02:04.172032 1001-01-07T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-08T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-09T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-10T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-11T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-12T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-13T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-14T09:02:04.172032

-- !test_21 --
1001-01-07 1001-01-07
Expand All @@ -181,14 +181,14 @@ apple_banana_mango9
1001-01-07T17:07:47.171 1001-01-14T17:07:47.171

-- !test_23 --
1001-01-07T17:07:47.172032 1001-01-07T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-08T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-09T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-10T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-11T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-12T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-13T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-14T17:07:47.172032
1001-01-07T09:02:04.172032 1001-01-07T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-08T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-09T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-10T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-11T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-12T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-13T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-14T09:02:04.172032

-- !test_24 --
false 1 2 3 10 1.2 val_1 val_1 HEARTS false 1 2 3 10 1.2 val_1 val_1 HEARTS ["arr_1", "arr_2", "arr_3"] [1] {1:"val_1", 2:"val_2", 3:"val_3"} {1:[{"nestedintscolumn":[1, 2, 3], "nestedstringcolumn":"val_1"}, {"nestedintscolumn":[2, 3, 4], "nestedstringcolumn":"val_2"}, {"nestedintscolumn":[3, 4, 5], "nestedstringcolumn":"val_3"}], 2:[{"nestedintscolumn":[1, 2, 3], "nestedstringcolumn":"val_1"}, {"nestedintscolumn":[2, 3, 4], "nestedstringcolumn":"val_2"}, {"nestedintscolumn":[3, 4, 5], "nestedstringcolumn":"val_3"}], 3:[{"nestedintscolumn":[1, 2, 3], "nestedstringcolumn":"val_1"}, {"nestedintscolumn":[2, 3, 4], "nestedstringcolumn":"val_2"}, {"nestedintscolumn":[3, 4, 5], "nestedstringcolumn":"val_3"}]}
Expand Down Expand Up @@ -238,14 +238,14 @@ true 8 9 10 80 8.2 val_8 val_8 SPADES true 8 9 10 80 8.2 val_8 val_8 SPADES ["ar
9.00

-- !test_29 --
1001-01-07T17:07:47.172032 1001-01-07T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-08T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-09T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-10T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-11T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-12T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-13T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-14T17:07:47.172032
1001-01-07T09:02:04.172032 1001-01-07T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-08T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-09T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-10T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-11T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-12T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-13T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-14T09:02:04.172032

-- !test_30 --
\N
Expand All @@ -268,14 +268,14 @@ true 8 9 10 80 8.2 val_8 val_8 SPADES true 8 9 10 80 8.2 val_8 val_8 SPADES ["ar
1970-01-01T08:00:00.010

-- !test_33 --
1001-01-07T17:07:47.172032 1001-01-07T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-08T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-09T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-10T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-11T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-12T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-13T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-14T17:07:47.172032
1001-01-07T09:02:04.172032 1001-01-07T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-08T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-09T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-10T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-11T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-12T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-13T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-14T09:02:04.172032

-- !test_34 --
1001-01-07 1001-01-07
Expand All @@ -298,12 +298,12 @@ true 8 9 10 80 8.2 val_8 val_8 SPADES true 8 9 10 80 8.2 val_8 val_8 SPADES ["ar
1001-01-07T17:07:47.171 1001-01-14T17:07:47.171

-- !test_36 --
1001-01-07T17:07:47.172032 1001-01-07T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-08T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-09T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-10T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-11T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-12T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-13T17:07:47.172032
1001-01-07T17:07:47.172032 1001-01-14T17:07:47.172032
1001-01-07T09:02:04.172032 1001-01-07T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-08T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-09T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-10T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-11T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-12T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-13T09:02:04.172032
1001-01-07T09:02:04.172032 1001-01-14T09:02:04.172032

Binary file modified regression-test/data/external_table_p0/tvf/test_hdfs_parquet_group4.out
View file
Open in desktop
Binary file not shown.
Binary file modified regression-test/data/external_table_p0/tvf/test_hdfs_parquet_group5.out
View file
Open in desktop
Binary file not shown.
Loading