mirror of
https://github.com/ClickHouse/ClickHouse.git
synced 2024-09-20 00:30:49 +00:00
Merge branch 'master' into named-collections-in-clickhouse-local
This commit is contained in:
Alexey Milovidov
commit
c458a504bf
@ -86,6 +86,7 @@ RUN curl -L --no-verbose -O 'https://archive.apache.org/dist/hadoop/common/hadoo
|
||||
ENV MINIO_ROOT_USER="clickhouse"
|
||||
ENV MINIO_ROOT_PASSWORD="clickhouse"
|
||||
ENV EXPORT_S3_STORAGE_POLICIES=1
|
||||
ENV CLICKHOUSE_GRPC_CLIENT="/usr/share/clickhouse-utils/grpc-client/clickhouse-grpc-client.py"
|
||||
|
||||
RUN npm install -g azurite@3.30.0 \
|
||||
&& npm install -g tslib && npm install -g node
|
||||
|
||||
@ -8,6 +8,7 @@ cryptography==3.4.8
|
||||
dbus-python==1.2.18
|
||||
distro==1.7.0
|
||||
docutils==0.17.1
|
||||
grpcio==1.47.0
|
||||
gyp==0.1
|
||||
httplib2==0.20.2
|
||||
idna==3.3
|
||||
@ -28,6 +29,7 @@ packaging==24.1
|
||||
pandas==1.5.3
|
||||
pip==24.1.1
|
||||
pipdeptree==2.23.0
|
||||
protobuf==4.25.3
|
||||
pyarrow==15.0.0
|
||||
pyasn1==0.4.8
|
||||
PyJWT==2.3.0
|
||||
|
||||
@ -6,8 +6,8 @@ source /setup_export_logs.sh
|
||||
# fail on errors, verbose and export all env variables
|
||||
set -e -x -a
|
||||
|
||||
MAX_RUN_TIME=${MAX_RUN_TIME:-10800}
|
||||
MAX_RUN_TIME=$((MAX_RUN_TIME == 0 ? 10800 : MAX_RUN_TIME))
|
||||
MAX_RUN_TIME=${MAX_RUN_TIME:-7200}
|
||||
MAX_RUN_TIME=$((MAX_RUN_TIME == 0 ? 7200 : MAX_RUN_TIME))
|
||||
|
||||
USE_DATABASE_REPLICATED=${USE_DATABASE_REPLICATED:=0}
|
||||
USE_SHARED_CATALOG=${USE_SHARED_CATALOG:=0}
|
||||
@ -320,7 +320,7 @@ export -f run_tests
|
||||
|
||||
|
||||
# This should be enough to setup job and collect artifacts
|
||||
TIMEOUT=$((MAX_RUN_TIME - 600))
|
||||
TIMEOUT=$((MAX_RUN_TIME - 700))
|
||||
if [ "$NUM_TRIES" -gt "1" ]; then
|
||||
# We don't run tests with Ordinary database in PRs, only in master.
|
||||
# So run new/changed tests with Ordinary at least once in flaky check.
|
||||
|
||||
@ -11,6 +11,7 @@ TIMEOUT_SIGN = "[ Timeout! "
|
||||
UNKNOWN_SIGN = "[ UNKNOWN "
|
||||
SKIPPED_SIGN = "[ SKIPPED "
|
||||
HUNG_SIGN = "Found hung queries in processlist"
|
||||
SERVER_DIED_SIGN = "Server died, terminating all processes"
|
||||
DATABASE_SIGN = "Database: "
|
||||
|
||||
SUCCESS_FINISH_SIGNS = ["All tests have finished", "No tests were run"]
|
||||
@ -25,6 +26,7 @@ def process_test_log(log_path, broken_tests):
|
||||
failed = 0
|
||||
success = 0
|
||||
hung = False
|
||||
server_died = False
|
||||
retries = False
|
||||
success_finish = False
|
||||
test_results = []
|
||||
@ -41,6 +43,8 @@ def process_test_log(log_path, broken_tests):
|
||||
if HUNG_SIGN in line:
|
||||
hung = True
|
||||
break
|
||||
if SERVER_DIED_SIGN in line:
|
||||
server_died = True
|
||||
if RETRIES_SIGN in line:
|
||||
retries = True
|
||||
if any(
|
||||
@ -123,6 +127,7 @@ def process_test_log(log_path, broken_tests):
|
||||
failed,
|
||||
success,
|
||||
hung,
|
||||
server_died,
|
||||
success_finish,
|
||||
retries,
|
||||
test_results,
|
||||
@ -150,6 +155,7 @@ def process_result(result_path, broken_tests):
|
||||
failed,
|
||||
success,
|
||||
hung,
|
||||
server_died,
|
||||
success_finish,
|
||||
retries,
|
||||
test_results,
|
||||
@ -165,6 +171,10 @@ def process_result(result_path, broken_tests):
|
||||
description = "Some queries hung, "
|
||||
state = "failure"
|
||||
test_results.append(("Some queries hung", "FAIL", "0", ""))
|
||||
elif server_died:
|
||||
description = "Server died, "
|
||||
state = "failure"
|
||||
test_results.append(("Server died", "FAIL", "0", ""))
|
||||
elif not success_finish:
|
||||
description = "Tests are not finished, "
|
||||
state = "failure"
|
||||
@ -218,5 +228,20 @@ if __name__ == "__main__":
|
||||
state, description, test_results = process_result(args.in_results_dir, broken_tests)
|
||||
logging.info("Result parsed")
|
||||
status = (state, description)
|
||||
|
||||
def test_result_comparator(item):
|
||||
# sort by status then by check name
|
||||
order = {
|
||||
"FAIL": 0,
|
||||
"Timeout": 1,
|
||||
"NOT_FAILED": 2,
|
||||
"BROKEN": 3,
|
||||
"OK": 4,
|
||||
"SKIPPED": 5,
|
||||
}
|
||||
return order.get(item[1], 10), str(item[0]), item[1]
|
||||
|
||||
test_results.sort(key=test_result_comparator)
|
||||
|
||||
write_results(args.out_results_file, args.out_status_file, test_results, status)
|
||||
logging.info("Result written")
|
||||
|
||||
@ -16,7 +16,7 @@ sidebar_label: clickhouse-local
|
||||
|
||||
While `clickhouse-local` is a great tool for development and testing purposes, and for processing files, it is not suitable for serving end users or applications. In these scenarios, it is recommended to use the open-source [ClickHouse](https://clickhouse.com/docs/en/install). ClickHouse is a powerful OLAP database that is designed to handle large-scale analytical workloads. It provides fast and efficient processing of complex queries on large datasets, making it ideal for use in production environments where high-performance is critical. Additionally, ClickHouse offers a wide range of features such as replication, sharding, and high availability, which are essential for scaling up to handle large datasets and serving applications. If you need to handle larger datasets or serve end users or applications, we recommend using open-source ClickHouse instead of `clickhouse-local`.
|
||||
|
||||
Please read the docs below that show example use cases for `clickhouse-local`, such as [querying local CSVs](#query-data-in-a-csv-file-using-sql) or [reading a parquet file in S3](#query-data-in-a-parquet-file-in-aws-s3).
|
||||
Please read the docs below that show example use cases for `clickhouse-local`, such as [querying local file](#query_data_in_file) or [reading a parquet file in S3](#query-data-in-a-parquet-file-in-aws-s3).
|
||||
|
||||
## Download clickhouse-local
|
||||
|
||||
|
||||
@ -18,7 +18,7 @@ ClickHouse also supports:
|
||||
|
||||
During aggregation, all `NULL` arguments are skipped. If the aggregation has several arguments it will ignore any row in which one or more of them are NULL.
|
||||
|
||||
There is an exception to this rule, which are the functions [`first_value`](../../sql-reference/aggregate-functions/reference/first_value.md), [`last_value`](../../sql-reference/aggregate-functions/reference/last_value.md) and their aliases when followed by the modifier `RESPECT NULLS`: `FIRST_VALUE(b) RESPECT NULLS`.
|
||||
There is an exception to this rule, which are the functions [`first_value`](../../sql-reference/aggregate-functions/reference/first_value.md), [`last_value`](../../sql-reference/aggregate-functions/reference/last_value.md) and their aliases (`any` and `anyLast` respectively) when followed by the modifier `RESPECT NULLS`. For example, `FIRST_VALUE(b) RESPECT NULLS`.
|
||||
|
||||
**Examples:**
|
||||
|
||||
|
||||
@ -5,12 +5,12 @@ sidebar_position: 102
|
||||
|
||||
# any
|
||||
|
||||
Selects the first encountered value of a column.
|
||||
Selects the first encountered value of a column, ignoring any `NULL` values.
|
||||
|
||||
**Syntax**
|
||||
|
||||
```sql
|
||||
any(column)
|
||||
any(column) [RESPECT NULLS]
|
||||
```
|
||||
|
||||
Aliases: `any_value`, [`first_value`](../reference/first_value.md).
|
||||
@ -20,7 +20,9 @@ Aliases: `any_value`, [`first_value`](../reference/first_value.md).
|
||||
|
||||
**Returned value**
|
||||
|
||||
By default, it ignores NULL values and returns the first NOT NULL value found in the column. Like [`first_value`](../../../sql-reference/aggregate-functions/reference/first_value.md) it supports `RESPECT NULLS`, in which case it will select the first value passed, independently on whether it's NULL or not.
|
||||
:::note
|
||||
Supports the `RESPECT NULLS` modifier after the function name. Using this modifier will ensure the function selects the first value passed, regardless of whether it is `NULL` or not.
|
||||
:::
|
||||
|
||||
:::note
|
||||
The return type of the function is the same as the input, except for LowCardinality which is discarded. This means that given no rows as input it will return the default value of that type (0 for integers, or Null for a Nullable() column). You might use the `-OrNull` [combinator](../../../sql-reference/aggregate-functions/combinators.md) ) to modify this behaviour.
|
||||
|
||||
@ -1,44 +0,0 @@
|
||||
---
|
||||
slug: /en/sql-reference/aggregate-functions/reference/any_respect_nulls
|
||||
sidebar_position: 103
|
||||
---
|
||||
|
||||
# any_respect_nulls
|
||||
|
||||
Selects the first encountered value of a column, irregardless of whether it is a `NULL` value or not.
|
||||
|
||||
Alias: `any_value_respect_nulls`, `first_value_repect_nulls`.
|
||||
|
||||
**Syntax**
|
||||
|
||||
```sql
|
||||
any_respect_nulls(column)
|
||||
```
|
||||
|
||||
**Parameters**
|
||||
- `column`: The column name.
|
||||
|
||||
**Returned value**
|
||||
|
||||
- The last value encountered, irregardless of whether it is a `NULL` value or not.
|
||||
|
||||
**Example**
|
||||
|
||||
Query:
|
||||
|
||||
```sql
|
||||
CREATE TABLE any_nulls (city Nullable(String)) ENGINE=Log;
|
||||
|
||||
INSERT INTO any_nulls (city) VALUES (NULL), ('Amsterdam'), ('New York'), ('Tokyo'), ('Valencia'), (NULL);
|
||||
|
||||
SELECT any(city), any_respect_nulls(city) FROM any_nulls;
|
||||
```
|
||||
|
||||
```response
|
||||
┌─any(city)─┬─any_respect_nulls(city)─┐
|
||||
│ Amsterdam │ ᴺᵁᴸᴸ │
|
||||
└───────────┴─────────────────────────┘
|
||||
```
|
||||
|
||||
**See Also**
|
||||
- [any](../reference/any.md)
|
||||
@ -5,17 +5,21 @@ sidebar_position: 105
|
||||
|
||||
# anyLast
|
||||
|
||||
Selects the last value encountered. The result is just as indeterminate as for the [any](../../../sql-reference/aggregate-functions/reference/any.md) function.
|
||||
Selects the last value encountered, ignoring any `NULL` values by default. The result is just as indeterminate as for the [any](../../../sql-reference/aggregate-functions/reference/any.md) function.
|
||||
|
||||
**Syntax**
|
||||
|
||||
```sql
|
||||
anyLast(column)
|
||||
anyLast(column) [RESPECT NULLS]
|
||||
```
|
||||
|
||||
**Parameters**
|
||||
- `column`: The column name.
|
||||
|
||||
:::note
|
||||
Supports the `RESPECT NULLS` modifier after the function name. Using this modifier will ensure the function selects the first value passed, regardless of whether it is `NULL` or not.
|
||||
:::
|
||||
|
||||
**Returned value**
|
||||
|
||||
- The last value encountered.
|
||||
|
||||
@ -1,39 +0,0 @@
|
||||
---
|
||||
slug: /en/sql-reference/aggregate-functions/reference/anylast_respect_nulls
|
||||
sidebar_position: 106
|
||||
---
|
||||
|
||||
# anyLast_respect_nulls
|
||||
|
||||
Selects the last value encountered, irregardless of whether it is `NULL` or not.
|
||||
|
||||
**Syntax**
|
||||
|
||||
```sql
|
||||
anyLast_respect_nulls(column)
|
||||
```
|
||||
|
||||
**Parameters**
|
||||
- `column`: The column name.
|
||||
|
||||
**Returned value**
|
||||
|
||||
- The last value encountered, irregardless of whether it is `NULL` or not.
|
||||
|
||||
**Example**
|
||||
|
||||
Query:
|
||||
|
||||
```sql
|
||||
CREATE TABLE any_last_nulls (city Nullable(String)) ENGINE=Log;
|
||||
|
||||
INSERT INTO any_last_nulls (city) VALUES ('Amsterdam'),(NULL),('New York'),('Tokyo'),('Valencia'),(NULL);
|
||||
|
||||
SELECT anyLast(city), anyLast_respect_nulls(city) FROM any_last_nulls;
|
||||
```
|
||||
|
||||
```response
|
||||
┌─anyLast(city)─┬─anyLast_respect_nulls(city)─┐
|
||||
│ Valencia │ ᴺᵁᴸᴸ │
|
||||
└───────────────┴─────────────────────────────┘
|
||||
```
|
||||
@ -45,10 +45,9 @@ ClickHouse-specific aggregate functions:
|
||||
|
||||
- [aggThrow](../reference/aggthrow.md)
|
||||
- [analysisOfVariance](../reference/analysis_of_variance.md)
|
||||
- [any](../reference/any_respect_nulls.md)
|
||||
- [any](../reference/any.md)
|
||||
- [anyHeavy](../reference/anyheavy.md)
|
||||
- [anyLast](../reference/anylast.md)
|
||||
- [anyLast](../reference/anylast_respect_nulls.md)
|
||||
- [boundingRatio](../reference/boundrat.md)
|
||||
- [first_value](../reference/first_value.md)
|
||||
- [last_value](../reference/last_value.md)
|
||||
|
||||
@ -1,6 +1,6 @@
|
||||
---
|
||||
slug: /en/sql-reference/data-types/dynamic
|
||||
sidebar_position: 56
|
||||
sidebar_position: 62
|
||||
sidebar_label: Dynamic
|
||||
---
|
||||
|
||||
@ -494,13 +494,43 @@ SELECT count(), dynamicType(d), _part FROM test GROUP BY _part, dynamicType(d) O
|
||||
|
||||
As we can see, ClickHouse kept the most frequent types `UInt64` and `Array(UInt64)` and casted all other types to `String`.
|
||||
|
||||
## JSONExtract functions with Dynamic
|
||||
|
||||
All `JSONExtract*` functions support `Dynamic` type:
|
||||
|
||||
```sql
|
||||
SELECT JSONExtract('{"a" : [1, 2, 3]}', 'a', 'Dynamic') AS dynamic, dynamicType(dynamic) AS dynamic_type;
|
||||
```
|
||||
|
||||
```text
|
||||
┌─dynamic─┬─dynamic_type───────────┐
|
||||
│ [1,2,3] │ Array(Nullable(Int64)) │
|
||||
└─────────┴────────────────────────┘
|
||||
```
|
||||
|
||||
```sql
|
||||
SELECT JSONExtract('{"obj" : {"a" : 42, "b" : "Hello", "c" : [1,2,3]}}', 'obj', 'Map(String, Variant(UInt32, String, Array(UInt32)))') AS map_of_dynamics, mapApply((k, v) -> (k, variantType(v)), map_of_dynamics) AS map_of_dynamic_types```
|
||||
|
||||
```text
|
||||
┌─map_of_dynamics──────────────────┬─map_of_dynamic_types────────────────────────────┐
|
||||
│ {'a':42,'b':'Hello','c':[1,2,3]} │ {'a':'UInt32','b':'String','c':'Array(UInt32)'} │
|
||||
└──────────────────────────────────┴─────────────────────────────────────────────────┘
|
||||
```
|
||||
|
||||
```sql
|
||||
SELECT JSONExtractKeysAndValues('{"a" : 42, "b" : "Hello", "c" : [1,2,3]}', 'Variant(UInt32, String, Array(UInt32))') AS dynamics, arrayMap(x -> (x.1, variantType(x.2)), dynamics) AS dynamic_types```
|
||||
```
|
||||
|
||||
```text
|
||||
┌─dynamics───────────────────────────────┬─dynamic_types─────────────────────────────────────────┐
|
||||
│ [('a',42),('b','Hello'),('c',[1,2,3])] │ [('a','UInt32'),('b','String'),('c','Array(UInt32)')] │
|
||||
└────────────────────────────────────────┴───────────────────────────────────────────────────────┘
|
||||
```
|
||||
|
||||
### Binary output format
|
||||
|
||||
In [RowBinary](../../interfaces/formats.md#rowbinary-rowbinary) format values of `Dynamic` type are serialized in the following format:
|
||||
In RowBinary format values of `Dynamic` type are serialized in the following format:
|
||||
|
||||
```text
|
||||
<binary_encoded_data_type><value_in_binary_format_according_to_the_data_type>
|
||||
```
|
||||
|
||||
See the [data types binary encoding specification](../../sql-reference/data-types/data-types-binary-encoding.md)
|
||||
|
||||
@ -3080,4 +3080,4 @@ Result:
|
||||
|
||||
## Distance functions
|
||||
|
||||
All supported functions are described in [distance functions documentation](../../sql-reference/functions/distance-functions.md).
|
||||
All supported functions are described in [distance functions documentation](../../sql-reference/functions/distance-functions.md).
|
||||
@ -314,10 +314,71 @@ SELECT groupBitXor(cityHash64(*)) FROM table
|
||||
Calculates a 32-bit hash code from any type of integer.
|
||||
This is a relatively fast non-cryptographic hash function of average quality for numbers.
|
||||
|
||||
**Syntax**
|
||||
|
||||
```sql
|
||||
intHash32(int)
|
||||
```
|
||||
|
||||
**Arguments**
|
||||
|
||||
- `int` — Integer to hash. [(U)Int*](../data-types/int-uint.md).
|
||||
|
||||
**Returned value**
|
||||
|
||||
- 32-bit hash code. [UInt32](../data-types/int-uint.md).
|
||||
|
||||
**Example**
|
||||
|
||||
Query:
|
||||
|
||||
```sql
|
||||
SELECT intHash32(42);
|
||||
```
|
||||
|
||||
Result:
|
||||
|
||||
```response
|
||||
┌─intHash32(42)─┐
|
||||
│ 1228623923 │
|
||||
└───────────────┘
|
||||
```
|
||||
|
||||
## intHash64
|
||||
|
||||
Calculates a 64-bit hash code from any type of integer.
|
||||
It works faster than intHash32. Average quality.
|
||||
This is a relatively fast non-cryptographic hash function of average quality for numbers.
|
||||
It works faster than [intHash32](#inthash32).
|
||||
|
||||
**Syntax**
|
||||
|
||||
```sql
|
||||
intHash64(int)
|
||||
```
|
||||
|
||||
**Arguments**
|
||||
|
||||
- `int` — Integer to hash. [(U)Int*](../data-types/int-uint.md).
|
||||
|
||||
**Returned value**
|
||||
|
||||
- 64-bit hash code. [UInt64](../data-types/int-uint.md).
|
||||
|
||||
**Example**
|
||||
|
||||
Query:
|
||||
|
||||
```sql
|
||||
SELECT intHash64(42);
|
||||
```
|
||||
|
||||
Result:
|
||||
|
||||
```response
|
||||
┌────────intHash64(42)─┐
|
||||
│ 11490350930367293593 │
|
||||
└──────────────────────┘
|
||||
```
|
||||
|
||||
## SHA1, SHA224, SHA256, SHA512, SHA512_256
|
||||
|
||||
|
||||
73
docs/en/sql-reference/window-functions/dense_rank.md
Normal file
73
docs/en/sql-reference/window-functions/dense_rank.md
Normal file
@ -0,0 +1,73 @@
|
||||
---
|
||||
slug: /en/sql-reference/window-functions/dense_rank
|
||||
sidebar_label: dense_rank
|
||||
sidebar_position: 7
|
||||
---
|
||||
|
||||
# dense_rank
|
||||
|
||||
Ranks the current row within its partition without gaps. In other words, if the value of any new row encountered is equal to the value of one of the previous rows then it will receive the next successive rank without any gaps in ranking.
|
||||
|
||||
The [rank](./rank.md) function provides the same behaviour, but with gaps in ranking.
|
||||
|
||||
**Syntax**
|
||||
|
||||
```sql
|
||||
dense_rank (column_name)
|
||||
OVER ([[PARTITION BY grouping_column] [ORDER BY sorting_column]
|
||||
[ROWS or RANGE expression_to_bound_rows_withing_the_group]] | [window_name])
|
||||
FROM table_name
|
||||
WINDOW window_name as ([[PARTITION BY grouping_column] [ORDER BY sorting_column])
|
||||
```
|
||||
|
||||
For more detail on window function syntax see: [Window Functions - Syntax](./index.md/#syntax).
|
||||
|
||||
**Returned value**
|
||||
|
||||
- A number for the current row within its partition, without gaps in ranking. [UInt64](../data-types/int-uint.md).
|
||||
|
||||
**Example**
|
||||
|
||||
The following example is based on the example provided in the video instructional [Ranking window functions in ClickHouse](https://youtu.be/Yku9mmBYm_4?si=XIMu1jpYucCQEoXA).
|
||||
|
||||
Query:
|
||||
|
||||
```sql
|
||||
CREATE TABLE salaries
|
||||
(
|
||||
`team` String,
|
||||
`player` String,
|
||||
`salary` UInt32,
|
||||
`position` String
|
||||
)
|
||||
Engine = Memory;
|
||||
|
||||
INSERT INTO salaries FORMAT Values
|
||||
('Port Elizabeth Barbarians', 'Gary Chen', 195000, 'F'),
|
||||
('New Coreystad Archdukes', 'Charles Juarez', 190000, 'F'),
|
||||
('Port Elizabeth Barbarians', 'Michael Stanley', 150000, 'D'),
|
||||
('New Coreystad Archdukes', 'Scott Harrison', 150000, 'D'),
|
||||
('Port Elizabeth Barbarians', 'Robert George', 195000, 'M'),
|
||||
('South Hampton Seagulls', 'Douglas Benson', 150000, 'M'),
|
||||
('South Hampton Seagulls', 'James Henderson', 140000, 'M');
|
||||
```
|
||||
|
||||
```sql
|
||||
SELECT player, salary,
|
||||
dense_rank() OVER (ORDER BY salary DESC) AS dense_rank
|
||||
FROM salaries;
|
||||
```
|
||||
|
||||
Result:
|
||||
|
||||
```response
|
||||
┌─player──────────┬─salary─┬─dense_rank─┐
|
||||
1. │ Gary Chen │ 195000 │ 1 │
|
||||
2. │ Robert George │ 195000 │ 1 │
|
||||
3. │ Charles Juarez │ 190000 │ 2 │
|
||||
4. │ Michael Stanley │ 150000 │ 3 │
|
||||
5. │ Douglas Benson │ 150000 │ 3 │
|
||||
6. │ Scott Harrison │ 150000 │ 3 │
|
||||
7. │ James Henderson │ 140000 │ 4 │
|
||||
└─────────────────┴────────┴────────────┘
|
||||
```
|
||||
79
docs/en/sql-reference/window-functions/first_value.md
Normal file
79
docs/en/sql-reference/window-functions/first_value.md
Normal file
@ -0,0 +1,79 @@
|
||||
---
|
||||
slug: /en/sql-reference/window-functions/first_value
|
||||
sidebar_label: first_value
|
||||
sidebar_position: 3
|
||||
---
|
||||
|
||||
# first_value
|
||||
|
||||
Returns the first value evaluated within its ordered frame. By default, NULL arguments are skipped, however the `RESPECT NULLS` modifier can be used to override this behaviour.
|
||||
|
||||
**Syntax**
|
||||
|
||||
```sql
|
||||
first_value (column_name) [[RESPECT NULLS] | [IGNORE NULLS]]
|
||||
OVER ([[PARTITION BY grouping_column] [ORDER BY sorting_column]
|
||||
[ROWS or RANGE expression_to_bound_rows_withing_the_group]] | [window_name])
|
||||
FROM table_name
|
||||
WINDOW window_name as ([[PARTITION BY grouping_column] [ORDER BY sorting_column])
|
||||
```
|
||||
|
||||
Alias: `any`.
|
||||
|
||||
:::note
|
||||
Using the optional modifier `RESPECT NULLS` after `first_value(column_name)` will ensure that `NULL` arguments are not skipped.
|
||||
See [NULL processing](../aggregate-functions/index.md/#null-processing) for more information.
|
||||
:::
|
||||
|
||||
For more detail on window function syntax see: [Window Functions - Syntax](./index.md/#syntax).
|
||||
|
||||
**Returned value**
|
||||
|
||||
- The first value evaluated within its ordered frame.
|
||||
|
||||
**Example**
|
||||
|
||||
In this example the `first_value` function is used to find the highest paid footballer from a fictional dataset of salaries of Premier League football players.
|
||||
|
||||
Query:
|
||||
|
||||
```sql
|
||||
DROP TABLE IF EXISTS salaries;
|
||||
CREATE TABLE salaries
|
||||
(
|
||||
`team` String,
|
||||
`player` String,
|
||||
`salary` UInt32,
|
||||
`position` String
|
||||
)
|
||||
Engine = Memory;
|
||||
|
||||
INSERT INTO salaries FORMAT Values
|
||||
('Port Elizabeth Barbarians', 'Gary Chen', 196000, 'F'),
|
||||
('New Coreystad Archdukes', 'Charles Juarez', 190000, 'F'),
|
||||
('Port Elizabeth Barbarians', 'Michael Stanley', 100000, 'D'),
|
||||
('New Coreystad Archdukes', 'Scott Harrison', 180000, 'D'),
|
||||
('Port Elizabeth Barbarians', 'Robert George', 195000, 'M'),
|
||||
('South Hampton Seagulls', 'Douglas Benson', 150000, 'M'),
|
||||
('South Hampton Seagulls', 'James Henderson', 140000, 'M');
|
||||
```
|
||||
|
||||
```sql
|
||||
SELECT player, salary,
|
||||
first_value(player) OVER (ORDER BY salary DESC) AS highest_paid_player
|
||||
FROM salaries;
|
||||
```
|
||||
|
||||
Result:
|
||||
|
||||
```response
|
||||
┌─player──────────┬─salary─┬─highest_paid_player─┐
|
||||
1. │ Gary Chen │ 196000 │ Gary Chen │
|
||||
2. │ Robert George │ 195000 │ Gary Chen │
|
||||
3. │ Charles Juarez │ 190000 │ Gary Chen │
|
||||
4. │ Scott Harrison │ 180000 │ Gary Chen │
|
||||
5. │ Douglas Benson │ 150000 │ Gary Chen │
|
||||
6. │ James Henderson │ 140000 │ Gary Chen │
|
||||
7. │ Michael Stanley │ 100000 │ Gary Chen │
|
||||
└─────────────────┴────────┴─────────────────────┘
|
||||
```
|
||||
@ -1,10 +1,11 @@
|
||||
---
|
||||
slug: /en/sql-reference/window-functions/
|
||||
sidebar_position: 62
|
||||
sidebar_label: Window Functions
|
||||
title: Window Functions
|
||||
sidebar_position: 1
|
||||
---
|
||||
|
||||
# Window Functions
|
||||
|
||||
Windows functions let you perform calculations across a set of rows that are related to the current row.
|
||||
Some of the calculations that you can do are similar to those that can be done with an aggregate function, but a window function doesn't cause rows to be grouped into a single output - the individual rows are still returned.
|
||||
|
||||
@ -12,8 +13,8 @@ Some of the calculations that you can do are similar to those that can be done w
|
||||
|
||||
ClickHouse supports the standard grammar for defining windows and window functions. The table below indicates whether a feature is currently supported.
|
||||
|
||||
| Feature | Supported? |
|
||||
|------------------------------------------------------------------------------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
|
||||
| Feature | Supported? |
|
||||
|--------------------------------------------------------------------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
|
||||
| ad hoc window specification (`count(*) over (partition by id order by time desc)`) | ✅ |
|
||||
| expressions involving window functions, e.g. `(count(*) over ()) / 2)` | ✅ |
|
||||
| `WINDOW` clause (`select ... from table window w as (partition by id)`) | ✅ |
|
||||
@ -75,14 +76,14 @@ WINDOW window_name as ([[PARTITION BY grouping_column] [ORDER BY sorting_column]
|
||||
|
||||
These functions can be used only as a window function.
|
||||
|
||||
- `row_number()` - Number the current row within its partition starting from 1.
|
||||
- `first_value(x)` - Return the first non-NULL value evaluated within its ordered frame.
|
||||
- `last_value(x)` - Return the last non-NULL value evaluated within its ordered frame.
|
||||
- `nth_value(x, offset)` - Return the first non-NULL value evaluated against the nth row (offset) in its ordered frame.
|
||||
- `rank()` - Rank the current row within its partition with gaps.
|
||||
- `dense_rank()` - Rank the current row within its partition without gaps.
|
||||
- `lagInFrame(x[, offset[, default]])` - Return a value evaluated at the row that is at a specified physical offset row before the current row within the ordered frame. The offset parameter, if not specified, defaults to 1, meaning it will fetch the value from the next row. If the calculated row exceeds the boundaries of the window frame, the specified default value is returned.
|
||||
- `leadInFrame(x[, offset[, default]])` - Return a value evaluated at the row that is offset rows after the current row within the ordered frame. If offset is not provided, it defaults to 1. If the offset leads to a position outside the window frame, the specified default value is used.
|
||||
- [`row_number()`](./row_number.md) - Number the current row within its partition starting from 1.
|
||||
- [`first_value(x)`](./first_value.md) - Return the first value evaluated within its ordered frame.
|
||||
- [`last_value(x)`](./last_value.md) - Return the last value evaluated within its ordered frame.
|
||||
- [`nth_value(x, offset)`](./nth_value.md) - Return the first non-NULL value evaluated against the nth row (offset) in its ordered frame.
|
||||
- [`rank()`](./rank.md) - Rank the current row within its partition with gaps.
|
||||
- [`dense_rank()`](./dense_rank.md) - Rank the current row within its partition without gaps.
|
||||
- [`lagInFrame(x)`](./lagInFrame.md) - Return a value evaluated at the row that is at a specified physical offset row before the current row within the ordered frame.
|
||||
- [`leadInFrame(x)`](./leadInFrame.md) - Return a value evaluated at the row that is offset rows after the current row within the ordered frame.
|
||||
|
||||
## Examples
|
||||
|
||||
|
||||
79
docs/en/sql-reference/window-functions/lagInFrame.md
Normal file
79
docs/en/sql-reference/window-functions/lagInFrame.md
Normal file
@ -0,0 +1,79 @@
|
||||
---
|
||||
slug: /en/sql-reference/window-functions/lagInFrame
|
||||
sidebar_label: lagInFrame
|
||||
sidebar_position: 8
|
||||
---
|
||||
|
||||
# lagInFrame
|
||||
|
||||
Returns a value evaluated at the row that is at a specified physical offset row before the current row within the ordered frame.
|
||||
|
||||
**Syntax**
|
||||
|
||||
```sql
|
||||
lagInFrame(x[, offset[, default]])
|
||||
OVER ([[PARTITION BY grouping_column] [ORDER BY sorting_column]
|
||||
[ROWS or RANGE expression_to_bound_rows_withing_the_group]] | [window_name])
|
||||
FROM table_name
|
||||
WINDOW window_name as ([[PARTITION BY grouping_column] [ORDER BY sorting_column])
|
||||
```
|
||||
|
||||
For more detail on window function syntax see: [Window Functions - Syntax](./index.md/#syntax).
|
||||
|
||||
**Parameters**
|
||||
- `x` — Column name.
|
||||
- `offset` — Offset to apply. [(U)Int*](../data-types/int-uint.md). (Optional - `1` by default).
|
||||
- `default` — Value to return if calculated row exceeds the boundaries of the window frame. (Optional - `null` by default).
|
||||
|
||||
**Returned value**
|
||||
|
||||
- Value evaluated at the row that is at a specified physical offset before the current row within the ordered frame.
|
||||
|
||||
**Example**
|
||||
|
||||
This example looks at historical data for a specific stock and uses the `lagInFrame` function to calculate a day-to-day delta and percentage change in the closing price of the stock.
|
||||
|
||||
Query:
|
||||
|
||||
```sql
|
||||
CREATE TABLE stock_prices
|
||||
(
|
||||
`date` Date,
|
||||
`open` Float32, -- opening price
|
||||
`high` Float32, -- daily high
|
||||
`low` Float32, -- daily low
|
||||
`close` Float32, -- closing price
|
||||
`volume` UInt32 -- trade volume
|
||||
)
|
||||
Engine = Memory;
|
||||
|
||||
INSERT INTO stock_prices FORMAT Values
|
||||
('2024-06-03', 113.62, 115.00, 112.00, 115.00, 438392000),
|
||||
('2024-06-04', 115.72, 116.60, 114.04, 116.44, 403324000),
|
||||
('2024-06-05', 118.37, 122.45, 117.47, 122.44, 528402000),
|
||||
('2024-06-06', 124.05, 125.59, 118.32, 121.00, 664696000),
|
||||
('2024-06-07', 119.77, 121.69, 118.02, 120.89, 412386000);
|
||||
```
|
||||
|
||||
```sql
|
||||
SELECT
|
||||
date,
|
||||
close,
|
||||
lagInFrame(close, 1, close) OVER (ORDER BY date ASC) AS previous_day_close,
|
||||
COALESCE(ROUND(close - previous_day_close, 2)) AS delta,
|
||||
COALESCE(ROUND((delta / previous_day_close) * 100, 2)) AS percent_change
|
||||
FROM stock_prices
|
||||
ORDER BY date DESC;
|
||||
```
|
||||
|
||||
Result:
|
||||
|
||||
```response
|
||||
┌───────date─┬──close─┬─previous_day_close─┬─delta─┬─percent_change─┐
|
||||
1. │ 2024-06-07 │ 120.89 │ 121 │ -0.11 │ -0.09 │
|
||||
2. │ 2024-06-06 │ 121 │ 122.44 │ -1.44 │ -1.18 │
|
||||
3. │ 2024-06-05 │ 122.44 │ 116.44 │ 6 │ 5.15 │
|
||||
4. │ 2024-06-04 │ 116.44 │ 115 │ 1.44 │ 1.25 │
|
||||
5. │ 2024-06-03 │ 115 │ 115 │ 0 │ 0 │
|
||||
└────────────┴────────┴────────────────────┴───────┴────────────────┘
|
||||
```
|
||||
79
docs/en/sql-reference/window-functions/last_value.md
Normal file
79
docs/en/sql-reference/window-functions/last_value.md
Normal file
@ -0,0 +1,79 @@
|
||||
---
|
||||
slug: /en/sql-reference/window-functions/last_value
|
||||
sidebar_label: last_value
|
||||
sidebar_position: 4
|
||||
---
|
||||
|
||||
# last_value
|
||||
|
||||
Returns the last value evaluated within its ordered frame. By default, NULL arguments are skipped, however the `RESPECT NULLS` modifier can be used to override this behaviour.
|
||||
|
||||
**Syntax**
|
||||
|
||||
```sql
|
||||
last_value (column_name) [[RESPECT NULLS] | [IGNORE NULLS]]
|
||||
OVER ([[PARTITION BY grouping_column] [ORDER BY sorting_column]
|
||||
[ROWS or RANGE expression_to_bound_rows_withing_the_group]] | [window_name])
|
||||
FROM table_name
|
||||
WINDOW window_name as ([[PARTITION BY grouping_column] [ORDER BY sorting_column])
|
||||
```
|
||||
|
||||
Alias: `anyLast`.
|
||||
|
||||
:::note
|
||||
Using the optional modifier `RESPECT NULLS` after `first_value(column_name)` will ensure that `NULL` arguments are not skipped.
|
||||
See [NULL processing](../aggregate-functions/index.md/#null-processing) for more information.
|
||||
:::
|
||||
|
||||
For more detail on window function syntax see: [Window Functions - Syntax](./index.md/#syntax).
|
||||
|
||||
**Returned value**
|
||||
|
||||
- The last value evaluated within its ordered frame.
|
||||
|
||||
**Example**
|
||||
|
||||
In this example the `last_value` function is used to find the highest paid footballer from a fictional dataset of salaries of Premier League football players.
|
||||
|
||||
Query:
|
||||
|
||||
```sql
|
||||
DROP TABLE IF EXISTS salaries;
|
||||
CREATE TABLE salaries
|
||||
(
|
||||
`team` String,
|
||||
`player` String,
|
||||
`salary` UInt32,
|
||||
`position` String
|
||||
)
|
||||
Engine = Memory;
|
||||
|
||||
INSERT INTO salaries FORMAT Values
|
||||
('Port Elizabeth Barbarians', 'Gary Chen', 196000, 'F'),
|
||||
('New Coreystad Archdukes', 'Charles Juarez', 190000, 'F'),
|
||||
('Port Elizabeth Barbarians', 'Michael Stanley', 100000, 'D'),
|
||||
('New Coreystad Archdukes', 'Scott Harrison', 180000, 'D'),
|
||||
('Port Elizabeth Barbarians', 'Robert George', 195000, 'M'),
|
||||
('South Hampton Seagulls', 'Douglas Benson', 150000, 'M'),
|
||||
('South Hampton Seagulls', 'James Henderson', 140000, 'M');
|
||||
```
|
||||
|
||||
```sql
|
||||
SELECT player, salary,
|
||||
last_value(player) OVER (ORDER BY salary DESC RANGE BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING) AS lowest_paid_player
|
||||
FROM salaries;
|
||||
```
|
||||
|
||||
Result:
|
||||
|
||||
```response
|
||||
┌─player──────────┬─salary─┬─lowest_paid_player─┐
|
||||
1. │ Gary Chen │ 196000 │ Michael Stanley │
|
||||
2. │ Robert George │ 195000 │ Michael Stanley │
|
||||
3. │ Charles Juarez │ 190000 │ Michael Stanley │
|
||||
4. │ Scott Harrison │ 180000 │ Michael Stanley │
|
||||
5. │ Douglas Benson │ 150000 │ Michael Stanley │
|
||||
6. │ James Henderson │ 140000 │ Michael Stanley │
|
||||
7. │ Michael Stanley │ 100000 │ Michael Stanley │
|
||||
└─────────────────┴────────┴────────────────────┘
|
||||
```
|
||||
60
docs/en/sql-reference/window-functions/leadInFrame.md
Normal file
60
docs/en/sql-reference/window-functions/leadInFrame.md
Normal file
@ -0,0 +1,60 @@
|
||||
---
|
||||
slug: /en/sql-reference/window-functions/leadInFrame
|
||||
sidebar_label: leadInFrame
|
||||
sidebar_position: 9
|
||||
---
|
||||
|
||||
# leadInFrame
|
||||
|
||||
Returns a value evaluated at the row that is offset rows after the current row within the ordered frame.
|
||||
|
||||
**Syntax**
|
||||
|
||||
```sql
|
||||
leadInFrame(x[, offset[, default]])
|
||||
OVER ([[PARTITION BY grouping_column] [ORDER BY sorting_column]
|
||||
[ROWS or RANGE expression_to_bound_rows_withing_the_group]] | [window_name])
|
||||
FROM table_name
|
||||
WINDOW window_name as ([[PARTITION BY grouping_column] [ORDER BY sorting_column])
|
||||
```
|
||||
|
||||
For more detail on window function syntax see: [Window Functions - Syntax](./index.md/#syntax).
|
||||
|
||||
**Parameters**
|
||||
- `x` — Column name.
|
||||
- `offset` — Offset to apply. [(U)Int*](../data-types/int-uint.md). (Optional - `1` by default).
|
||||
- `default` — Value to return if calculated row exceeds the boundaries of the window frame. (Optional - `null` by default).
|
||||
|
||||
**Returned value**
|
||||
|
||||
- value evaluated at the row that is offset rows after the current row within the ordered frame.
|
||||
|
||||
**Example**
|
||||
|
||||
This example looks at [historical data](https://www.kaggle.com/datasets/sazidthe1/nobel-prize-data) for Nobel Prize winners and uses the `leadInFrame` function to return a list of successive winners in the physics category.
|
||||
|
||||
Query:
|
||||
|
||||
```sql
|
||||
CREATE OR REPLACE VIEW nobel_prize_laureates AS FROM file('nobel_laureates_data.csv') SELECT *;
|
||||
```
|
||||
|
||||
```sql
|
||||
FROM nobel_prize_laureates SELECT fullName, leadInFrame(year, 1, year) OVER (PARTITION BY category ORDER BY year) AS year, category, motivation WHERE category == 'physics' ORDER BY year DESC LIMIT 9;
|
||||
```
|
||||
|
||||
Result:
|
||||
|
||||
```response
|
||||
┌─fullName─────────┬─year─┬─category─┬─motivation─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┐
|
||||
1. │ Pierre Agostini │ 2023 │ physics │ for experimental methods that generate attosecond pulses of light for the study of electron dynamics in matter │
|
||||
2. │ Ferenc Krausz │ 2023 │ physics │ for experimental methods that generate attosecond pulses of light for the study of electron dynamics in matter │
|
||||
3. │ Anne L Huillier │ 2023 │ physics │ for experimental methods that generate attosecond pulses of light for the study of electron dynamics in matter │
|
||||
4. │ Alain Aspect │ 2022 │ physics │ for experiments with entangled photons establishing the violation of Bell inequalities and pioneering quantum information science │
|
||||
5. │ Anton Zeilinger │ 2022 │ physics │ for experiments with entangled photons establishing the violation of Bell inequalities and pioneering quantum information science │
|
||||
6. │ John Clauser │ 2022 │ physics │ for experiments with entangled photons establishing the violation of Bell inequalities and pioneering quantum information science │
|
||||
7. │ Syukuro Manabe │ 2021 │ physics │ for the physical modelling of Earths climate quantifying variability and reliably predicting global warming │
|
||||
8. │ Klaus Hasselmann │ 2021 │ physics │ for the physical modelling of Earths climate quantifying variability and reliably predicting global warming │
|
||||
9. │ Giorgio Parisi │ 2021 │ physics │ for the discovery of the interplay of disorder and fluctuations in physical systems from atomic to planetary scales │
|
||||
└──────────────────┴──────┴──────────┴────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┘
|
||||
```
|
||||
75
docs/en/sql-reference/window-functions/nth_value.md
Normal file
75
docs/en/sql-reference/window-functions/nth_value.md
Normal file
@ -0,0 +1,75 @@
|
||||
---
|
||||
slug: /en/sql-reference/window-functions/nth_value
|
||||
sidebar_label: nth_value
|
||||
sidebar_position: 5
|
||||
---
|
||||
|
||||
# nth_value
|
||||
|
||||
Returns the first non-NULL value evaluated against the nth row (offset) in its ordered frame.
|
||||
|
||||
**Syntax**
|
||||
|
||||
```sql
|
||||
nth_value (x, offset)
|
||||
OVER ([[PARTITION BY grouping_column] [ORDER BY sorting_column]
|
||||
[ROWS or RANGE expression_to_bound_rows_withing_the_group]] | [window_name])
|
||||
FROM table_name
|
||||
WINDOW window_name as ([[PARTITION BY grouping_column] [ORDER BY sorting_column])
|
||||
```
|
||||
|
||||
For more detail on window function syntax see: [Window Functions - Syntax](./index.md/#syntax).
|
||||
|
||||
**Parameters**
|
||||
|
||||
- `x` — Column name.
|
||||
- `offset` — nth row to evaluate current row against.
|
||||
|
||||
**Returned value**
|
||||
|
||||
- The first non-NULL value evaluated against the nth row (offset) in its ordered frame.
|
||||
|
||||
**Example**
|
||||
|
||||
In this example the `nth-value` function is used to find the third-highest salary from a fictional dataset of salaries of Premier League football players.
|
||||
|
||||
Query:
|
||||
|
||||
```sql
|
||||
DROP TABLE IF EXISTS salaries;
|
||||
CREATE TABLE salaries
|
||||
(
|
||||
`team` String,
|
||||
`player` String,
|
||||
`salary` UInt32,
|
||||
`position` String
|
||||
)
|
||||
Engine = Memory;
|
||||
|
||||
INSERT INTO salaries FORMAT Values
|
||||
('Port Elizabeth Barbarians', 'Gary Chen', 195000, 'F'),
|
||||
('New Coreystad Archdukes', 'Charles Juarez', 190000, 'F'),
|
||||
('Port Elizabeth Barbarians', 'Michael Stanley', 100000, 'D'),
|
||||
('New Coreystad Archdukes', 'Scott Harrison', 180000, 'D'),
|
||||
('Port Elizabeth Barbarians', 'Robert George', 195000, 'M'),
|
||||
('South Hampton Seagulls', 'Douglas Benson', 150000, 'M'),
|
||||
('South Hampton Seagulls', 'James Henderson', 140000, 'M');
|
||||
```
|
||||
|
||||
```sql
|
||||
SELECT player, salary, nth_value(player,3) OVER(ORDER BY salary DESC) AS third_highest_salary FROM salaries;
|
||||
```
|
||||
|
||||
Result:
|
||||
|
||||
```response
|
||||
┌─player──────────┬─salary─┬─third_highest_salary─┐
|
||||
1. │ Gary Chen │ 195000 │ │
|
||||
2. │ Robert George │ 195000 │ │
|
||||
3. │ Charles Juarez │ 190000 │ Charles Juarez │
|
||||
4. │ Scott Harrison │ 180000 │ Charles Juarez │
|
||||
5. │ Douglas Benson │ 150000 │ Charles Juarez │
|
||||
6. │ James Henderson │ 140000 │ Charles Juarez │
|
||||
7. │ Michael Stanley │ 100000 │ Charles Juarez │
|
||||
└─────────────────┴────────┴──────────────────────┘
|
||||
```
|
||||
74
docs/en/sql-reference/window-functions/rank.md
Normal file
74
docs/en/sql-reference/window-functions/rank.md
Normal file
@ -0,0 +1,74 @@
|
||||
---
|
||||
slug: /en/sql-reference/window-functions/rank
|
||||
sidebar_label: rank
|
||||
sidebar_position: 6
|
||||
---
|
||||
|
||||
# rank
|
||||
|
||||
Ranks the current row within its partition with gaps. In other words, if the value of any row it encounters is equal to the value of a previous row then it will receive the same rank as that previous row.
|
||||
The rank of the next row is then equal to the rank of the previous row plus a gap equal to the number of times the previous rank was given.
|
||||
|
||||
The [dense_rank](./dense_rank.md) function provides the same behaviour but without gaps in ranking.
|
||||
|
||||
**Syntax**
|
||||
|
||||
```sql
|
||||
rank (column_name)
|
||||
OVER ([[PARTITION BY grouping_column] [ORDER BY sorting_column]
|
||||
[ROWS or RANGE expression_to_bound_rows_withing_the_group]] | [window_name])
|
||||
FROM table_name
|
||||
WINDOW window_name as ([[PARTITION BY grouping_column] [ORDER BY sorting_column])
|
||||
```
|
||||
|
||||
For more detail on window function syntax see: [Window Functions - Syntax](./index.md/#syntax).
|
||||
|
||||
**Returned value**
|
||||
|
||||
- A number for the current row within its partition, including gaps. [UInt64](../data-types/int-uint.md).
|
||||
|
||||
**Example**
|
||||
|
||||
The following example is based on the example provided in the video instructional [Ranking window functions in ClickHouse](https://youtu.be/Yku9mmBYm_4?si=XIMu1jpYucCQEoXA).
|
||||
|
||||
Query:
|
||||
|
||||
```sql
|
||||
CREATE TABLE salaries
|
||||
(
|
||||
`team` String,
|
||||
`player` String,
|
||||
`salary` UInt32,
|
||||
`position` String
|
||||
)
|
||||
Engine = Memory;
|
||||
|
||||
INSERT INTO salaries FORMAT Values
|
||||
('Port Elizabeth Barbarians', 'Gary Chen', 195000, 'F'),
|
||||
('New Coreystad Archdukes', 'Charles Juarez', 190000, 'F'),
|
||||
('Port Elizabeth Barbarians', 'Michael Stanley', 150000, 'D'),
|
||||
('New Coreystad Archdukes', 'Scott Harrison', 150000, 'D'),
|
||||
('Port Elizabeth Barbarians', 'Robert George', 195000, 'M'),
|
||||
('South Hampton Seagulls', 'Douglas Benson', 150000, 'M'),
|
||||
('South Hampton Seagulls', 'James Henderson', 140000, 'M');
|
||||
```
|
||||
|
||||
```sql
|
||||
SELECT player, salary,
|
||||
rank() OVER (ORDER BY salary DESC) AS rank
|
||||
FROM salaries;
|
||||
```
|
||||
|
||||
Result:
|
||||
|
||||
```response
|
||||
┌─player──────────┬─salary─┬─rank─┐
|
||||
1. │ Gary Chen │ 195000 │ 1 │
|
||||
2. │ Robert George │ 195000 │ 1 │
|
||||
3. │ Charles Juarez │ 190000 │ 3 │
|
||||
4. │ Douglas Benson │ 150000 │ 4 │
|
||||
5. │ Michael Stanley │ 150000 │ 4 │
|
||||
6. │ Scott Harrison │ 150000 │ 4 │
|
||||
7. │ James Henderson │ 140000 │ 7 │
|
||||
└─────────────────┴────────┴──────┘
|
||||
```
|
||||
67
docs/en/sql-reference/window-functions/row_number.md
Normal file
67
docs/en/sql-reference/window-functions/row_number.md
Normal file
@ -0,0 +1,67 @@
|
||||
---
|
||||
slug: /en/sql-reference/window-functions/row_number
|
||||
sidebar_label: row_number
|
||||
sidebar_position: 2
|
||||
---
|
||||
|
||||
# row_number
|
||||
|
||||
Numbers the current row within its partition starting from 1.
|
||||
|
||||
**Syntax**
|
||||
|
||||
```sql
|
||||
row_number (column_name)
|
||||
OVER ([[PARTITION BY grouping_column] [ORDER BY sorting_column]
|
||||
[ROWS or RANGE expression_to_bound_rows_withing_the_group]] | [window_name])
|
||||
FROM table_name
|
||||
WINDOW window_name as ([[PARTITION BY grouping_column] [ORDER BY sorting_column])
|
||||
```
|
||||
|
||||
For more detail on window function syntax see: [Window Functions - Syntax](./index.md/#syntax).
|
||||
|
||||
**Returned value**
|
||||
|
||||
- A number for the current row within its partition. [UInt64](../data-types/int-uint.md).
|
||||
|
||||
**Example**
|
||||
|
||||
The following example is based on the example provided in the video instructional [Ranking window functions in ClickHouse](https://youtu.be/Yku9mmBYm_4?si=XIMu1jpYucCQEoXA).
|
||||
|
||||
Query:
|
||||
|
||||
```sql
|
||||
CREATE TABLE salaries
|
||||
(
|
||||
`team` String,
|
||||
`player` String,
|
||||
`salary` UInt32,
|
||||
`position` String
|
||||
)
|
||||
Engine = Memory;
|
||||
|
||||
INSERT INTO salaries FORMAT Values
|
||||
('Port Elizabeth Barbarians', 'Gary Chen', 195000, 'F'),
|
||||
('New Coreystad Archdukes', 'Charles Juarez', 190000, 'F'),
|
||||
('Port Elizabeth Barbarians', 'Michael Stanley', 150000, 'D'),
|
||||
('New Coreystad Archdukes', 'Scott Harrison', 150000, 'D'),
|
||||
('Port Elizabeth Barbarians', 'Robert George', 195000, 'M');
|
||||
```
|
||||
|
||||
```sql
|
||||
SELECT player, salary,
|
||||
row_number() OVER (ORDER BY salary DESC) AS row_number
|
||||
FROM salaries;
|
||||
```
|
||||
|
||||
Result:
|
||||
|
||||
```response
|
||||
┌─player──────────┬─salary─┬─row_number─┐
|
||||
1. │ Gary Chen │ 195000 │ 1 │
|
||||
2. │ Robert George │ 195000 │ 2 │
|
||||
3. │ Charles Juarez │ 190000 │ 3 │
|
||||
4. │ Scott Harrison │ 150000 │ 4 │
|
||||
5. │ Michael Stanley │ 150000 │ 5 │
|
||||
└─────────────────┴────────┴────────────┘
|
||||
```
|
||||
@ -35,10 +35,9 @@ disable = '''
|
||||
broad-except,
|
||||
bare-except,
|
||||
no-else-return,
|
||||
global-statement
|
||||
global-statement,
|
||||
'''
|
||||
|
||||
[tool.pylint.SIMILARITIES]
|
||||
# due to SQL
|
||||
min-similarity-lines=1000
|
||||
|
||||
|
||||
@ -38,10 +38,19 @@ namespace ErrorCodes
|
||||
extern const int CANNOT_MREMAP;
|
||||
}
|
||||
|
||||
void abortOnFailedAssertion(const String & description, void * const * trace, size_t trace_offset, size_t trace_size)
|
||||
{
|
||||
auto & logger = Poco::Logger::root();
|
||||
LOG_FATAL(&logger, "Logical error: '{}'.", description);
|
||||
if (trace)
|
||||
LOG_FATAL(&logger, "Stack trace (when copying this message, always include the lines below):\n\n{}", StackTrace::toString(trace, trace_offset, trace_size));
|
||||
abort();
|
||||
}
|
||||
|
||||
void abortOnFailedAssertion(const String & description)
|
||||
{
|
||||
LOG_FATAL(&Poco::Logger::root(), "Logical error: '{}'.", description);
|
||||
abort();
|
||||
StackTrace st;
|
||||
abortOnFailedAssertion(description, st.getFramePointers().data(), st.getOffset(), st.getSize());
|
||||
}
|
||||
|
||||
bool terminate_on_any_exception = false;
|
||||
@ -58,7 +67,7 @@ void handle_error_code(const std::string & msg, int code, bool remote, const Exc
|
||||
#ifdef ABORT_ON_LOGICAL_ERROR
|
||||
if (code == ErrorCodes::LOGICAL_ERROR)
|
||||
{
|
||||
abortOnFailedAssertion(msg);
|
||||
abortOnFailedAssertion(msg, trace.data(), 0, trace.size());
|
||||
}
|
||||
#endif
|
||||
|
||||
|
||||
@ -25,8 +25,6 @@ namespace DB
|
||||
|
||||
class AtomicLogger;
|
||||
|
||||
[[noreturn]] void abortOnFailedAssertion(const String & description);
|
||||
|
||||
/// This flag can be set for testing purposes - to check that no exceptions are thrown.
|
||||
extern bool terminate_on_any_exception;
|
||||
|
||||
@ -167,6 +165,8 @@ protected:
|
||||
mutable std::vector<StackTrace::FramePointers> capture_thread_frame_pointers;
|
||||
};
|
||||
|
||||
[[noreturn]] void abortOnFailedAssertion(const String & description, void * const * trace, size_t trace_offset, size_t trace_size);
|
||||
[[noreturn]] void abortOnFailedAssertion(const String & description);
|
||||
|
||||
std::string getExceptionStackTraceString(const std::exception & e);
|
||||
std::string getExceptionStackTraceString(std::exception_ptr e);
|
||||
|
||||
@ -14,6 +14,7 @@
|
||||
|
||||
namespace DB
|
||||
{
|
||||
|
||||
namespace ErrorCodes
|
||||
{
|
||||
extern const int CANNOT_ALLOCATE_MEMORY;
|
||||
|
||||
@ -545,7 +545,7 @@ std::string StackTrace::toString() const
|
||||
return toStringCached(frame_pointers, offset, size);
|
||||
}
|
||||
|
||||
std::string StackTrace::toString(void ** frame_pointers_raw, size_t offset, size_t size)
|
||||
std::string StackTrace::toString(void * const * frame_pointers_raw, size_t offset, size_t size)
|
||||
{
|
||||
__msan_unpoison(frame_pointers_raw, size * sizeof(*frame_pointers_raw));
|
||||
|
||||
|
||||
@ -59,7 +59,7 @@ public:
|
||||
const FramePointers & getFramePointers() const { return frame_pointers; }
|
||||
std::string toString() const;
|
||||
|
||||
static std::string toString(void ** frame_pointers, size_t offset, size_t size);
|
||||
static std::string toString(void * const * frame_pointers, size_t offset, size_t size);
|
||||
static void dropCache();
|
||||
|
||||
/// @param fatal - if true, will process inline frames (slower)
|
||||
|
||||
@ -996,6 +996,10 @@ void ZooKeeper::receiveEvent()
|
||||
|
||||
if (request_info.callback)
|
||||
request_info.callback(*response);
|
||||
|
||||
/// Finalize current session if we receive a hardware error from ZooKeeper
|
||||
if (err != Error::ZOK && isHardwareError(err))
|
||||
finalize(/*error_send*/ false, /*error_receive*/ true, fmt::format("Hardware error: {}", err));
|
||||
}
|
||||
|
||||
|
||||
|
||||
@ -25,7 +25,7 @@ namespace DB
|
||||
template <typename To, typename From>
|
||||
inline To assert_cast(From && from)
|
||||
{
|
||||
#ifndef NDEBUG
|
||||
#ifdef ABORT_ON_LOGICAL_ERROR
|
||||
try
|
||||
{
|
||||
if constexpr (std::is_pointer_v<To>)
|
||||
|
||||
@ -228,7 +228,6 @@ Pool::Entry Pool::tryGet()
|
||||
for (auto connection_it = connections.cbegin(); connection_it != connections.cend();)
|
||||
{
|
||||
Connection * connection_ptr = *connection_it;
|
||||
/// Fixme: There is a race condition here b/c we do not synchronize with Pool::Entry's copy-assignment operator
|
||||
if (connection_ptr->ref_count == 0)
|
||||
{
|
||||
{
|
||||
|
||||
@ -64,17 +64,6 @@ public:
|
||||
decrementRefCount();
|
||||
}
|
||||
|
||||
Entry & operator= (const Entry & src) /// NOLINT
|
||||
{
|
||||
pool = src.pool;
|
||||
if (data)
|
||||
decrementRefCount();
|
||||
data = src.data;
|
||||
if (data)
|
||||
incrementRefCount();
|
||||
return * this;
|
||||
}
|
||||
|
||||
bool isNull() const
|
||||
{
|
||||
return data == nullptr;
|
||||
|
||||
@ -13,13 +13,11 @@ mysqlxx::Pool::Entry getWithFailover(mysqlxx::Pool & connections_pool)
|
||||
|
||||
constexpr size_t max_tries = 3;
|
||||
|
||||
mysqlxx::Pool::Entry worker_connection;
|
||||
|
||||
for (size_t try_no = 1; try_no <= max_tries; ++try_no)
|
||||
{
|
||||
try
|
||||
{
|
||||
worker_connection = connections_pool.tryGet();
|
||||
mysqlxx::Pool::Entry worker_connection = connections_pool.tryGet();
|
||||
|
||||
if (!worker_connection.isNull())
|
||||
{
|
||||
|
||||
@ -11,6 +11,7 @@
|
||||
#include <Parsers/ASTFunction.h>
|
||||
#include <Parsers/ASTIdentifier.h>
|
||||
#include <Parsers/ASTLiteral.h>
|
||||
#include <Parsers/ASTSubquery.h>
|
||||
#include <Parsers/ASTSelectWithUnionQuery.h>
|
||||
#include <Parsers/ASTTTLElement.h>
|
||||
#include <Poco/String.h>
|
||||
@ -211,6 +212,13 @@ void DDLLoadingDependencyVisitor::extractTableNameFromArgument(const ASTFunction
|
||||
qualified_name.database = table_identifier->getDatabaseName();
|
||||
qualified_name.table = table_identifier->shortName();
|
||||
}
|
||||
else if (arg->as<ASTSubquery>())
|
||||
{
|
||||
/// Allow IN subquery.
|
||||
/// Do not add tables from the subquery into dependencies,
|
||||
/// because CREATE will succeed anyway.
|
||||
return;
|
||||
}
|
||||
else
|
||||
{
|
||||
assert(false);
|
||||
|
||||
@ -107,12 +107,24 @@ void DatabaseAtomic::attachTable(ContextPtr /* context_ */, const String & name,
|
||||
|
||||
StoragePtr DatabaseAtomic::detachTable(ContextPtr /* context */, const String & name)
|
||||
{
|
||||
// it is important to call the destructors of not_in_use without
|
||||
// locked mutex to avoid potential deadlock.
|
||||
DetachedTables not_in_use;
|
||||
std::lock_guard lock(mutex);
|
||||
auto table = DatabaseOrdinary::detachTableUnlocked(name);
|
||||
table_name_to_path.erase(name);
|
||||
detached_tables.emplace(table->getStorageID().uuid, table);
|
||||
not_in_use = cleanupDetachedTables();
|
||||
StoragePtr table;
|
||||
{
|
||||
std::lock_guard lock(mutex);
|
||||
table = DatabaseOrdinary::detachTableUnlocked(name);
|
||||
table_name_to_path.erase(name);
|
||||
detached_tables.emplace(table->getStorageID().uuid, table);
|
||||
not_in_use = cleanupDetachedTables();
|
||||
}
|
||||
|
||||
if (!not_in_use.empty())
|
||||
{
|
||||
not_in_use.clear();
|
||||
LOG_DEBUG(log, "Finished removing not used detached tables");
|
||||
}
|
||||
|
||||
return table;
|
||||
}
|
||||
|
||||
|
||||
@ -29,6 +29,7 @@
|
||||
#include <Common/randomNumber.h>
|
||||
#include <Common/setThreadName.h>
|
||||
#include <base/sleep.h>
|
||||
#include <base/scope_guard.h>
|
||||
#include <boost/algorithm/string/split.hpp>
|
||||
#include <boost/algorithm/string/trim.hpp>
|
||||
#include <Parsers/CommonParsers.h>
|
||||
@ -532,13 +533,17 @@ static inline void dumpDataForTables(
|
||||
bool MaterializedMySQLSyncThread::prepareSynchronized(MaterializeMetadata & metadata)
|
||||
{
|
||||
bool opened_transaction = false;
|
||||
mysqlxx::PoolWithFailover::Entry connection;
|
||||
|
||||
while (!isCancelled())
|
||||
{
|
||||
try
|
||||
{
|
||||
connection = pool.tryGet();
|
||||
mysqlxx::PoolWithFailover::Entry connection = pool.tryGet();
|
||||
SCOPE_EXIT({
|
||||
if (opened_transaction)
|
||||
connection->query("ROLLBACK").execute();
|
||||
});
|
||||
|
||||
if (connection.isNull())
|
||||
{
|
||||
if (settings->max_wait_time_when_mysql_unavailable < 0)
|
||||
@ -602,9 +607,6 @@ bool MaterializedMySQLSyncThread::prepareSynchronized(MaterializeMetadata & meta
|
||||
{
|
||||
tryLogCurrentException(log);
|
||||
|
||||
if (opened_transaction)
|
||||
connection->query("ROLLBACK").execute();
|
||||
|
||||
if (settings->max_wait_time_when_mysql_unavailable < 0)
|
||||
throw;
|
||||
|
||||
|
||||
1660
src/Formats/JSONExtractTree.cpp
Normal file
1660
src/Formats/JSONExtractTree.cpp
Normal file
File diff suppressed because it is too large
Load Diff
41
src/Formats/JSONExtractTree.h
Normal file
41
src/Formats/JSONExtractTree.h
Normal file
@ -0,0 +1,41 @@
|
||||
#pragma once
|
||||
#include <DataTypes/IDataType.h>
|
||||
#include <Columns/IColumn.h>
|
||||
#include <Formats/FormatSettings.h>
|
||||
|
||||
|
||||
namespace DB
|
||||
{
|
||||
|
||||
struct JSONExtractInsertSettings
|
||||
{
|
||||
/// If false, JSON boolean values won't be inserted into columns with integer types
|
||||
/// It's used in JSONExtractInt64/JSONExtractUInt64/... functions.
|
||||
bool convert_bool_to_integer = true;
|
||||
/// If true, when complex type like Array/Map has both valid and invalid elements,
|
||||
/// the default value will be inserted on invalid elements.
|
||||
/// For example, if we have [1, "hello", 2] and type Array(UInt32),
|
||||
/// we will insert [1, 0, 2] in the column. Used in all JSONExtract functions.
|
||||
bool insert_default_on_invalid_elements_in_complex_types = false;
|
||||
};
|
||||
|
||||
template <typename JSONParser>
|
||||
class JSONExtractTreeNode
|
||||
{
|
||||
public:
|
||||
JSONExtractTreeNode() = default;
|
||||
virtual ~JSONExtractTreeNode() = default;
|
||||
virtual bool insertResultToColumn(IColumn &, const typename JSONParser::Element &, const JSONExtractInsertSettings & insert_setting, const FormatSettings & format_settings, String & error) const = 0;
|
||||
};
|
||||
|
||||
/// Build a tree for insertion JSON element into a column with provided data type.
|
||||
template <typename JSONParser>
|
||||
std::unique_ptr<JSONExtractTreeNode<JSONParser>> buildJSONExtractTree(const DataTypePtr & type, const char * source_for_exception_message);
|
||||
|
||||
template <typename JSONParser>
|
||||
void jsonElementToString(const typename JSONParser::Element & element, WriteBuffer & buf, const FormatSettings & format_settings);
|
||||
|
||||
template <typename JSONParser, typename NumberType>
|
||||
bool tryGetNumericValueFromJSONElement(NumberType & value, const typename JSONParser::Element & element, bool convert_bool_to_integer, String & error);
|
||||
|
||||
}
|
||||
@ -225,19 +225,6 @@ namespace
|
||||
Paths paths;
|
||||
};
|
||||
|
||||
bool checkIfTypesAreEqual(const DataTypes & types)
|
||||
{
|
||||
if (types.empty())
|
||||
return true;
|
||||
|
||||
for (size_t i = 1; i < types.size(); ++i)
|
||||
{
|
||||
if (!types[0]->equals(*types[i]))
|
||||
return false;
|
||||
}
|
||||
return true;
|
||||
}
|
||||
|
||||
void updateTypeIndexes(DataTypes & data_types, TypeIndexesSet & type_indexes)
|
||||
{
|
||||
type_indexes.clear();
|
||||
@ -272,24 +259,31 @@ namespace
|
||||
type_indexes.erase(TypeIndex::Nothing);
|
||||
}
|
||||
|
||||
/// If we have both Int64 and UInt64, convert all Int64 to UInt64,
|
||||
/// If we have both Int64 and UInt64, convert all not-negative Int64 to UInt64,
|
||||
/// because UInt64 is inferred only in case of Int64 overflow.
|
||||
void transformIntegers(DataTypes & data_types, TypeIndexesSet & type_indexes)
|
||||
void transformIntegers(DataTypes & data_types, TypeIndexesSet & type_indexes, JSONInferenceInfo * json_info)
|
||||
{
|
||||
if (!type_indexes.contains(TypeIndex::Int64) || !type_indexes.contains(TypeIndex::UInt64))
|
||||
return;
|
||||
|
||||
bool have_negative_integers = false;
|
||||
for (auto & type : data_types)
|
||||
{
|
||||
if (WhichDataType(type).isInt64())
|
||||
type = std::make_shared<DataTypeUInt64>();
|
||||
{
|
||||
bool is_negative = json_info && json_info->negative_integers.contains(type.get());
|
||||
have_negative_integers |= is_negative;
|
||||
if (!is_negative)
|
||||
type = std::make_shared<DataTypeUInt64>();
|
||||
}
|
||||
}
|
||||
|
||||
type_indexes.erase(TypeIndex::Int64);
|
||||
if (!have_negative_integers)
|
||||
type_indexes.erase(TypeIndex::Int64);
|
||||
}
|
||||
|
||||
/// If we have both Int64 and Float64 types, convert all Int64 to Float64.
|
||||
void transformIntegersAndFloatsToFloats(DataTypes & data_types, TypeIndexesSet & type_indexes)
|
||||
void transformIntegersAndFloatsToFloats(DataTypes & data_types, TypeIndexesSet & type_indexes, JSONInferenceInfo * json_info)
|
||||
{
|
||||
bool have_floats = type_indexes.contains(TypeIndex::Float64);
|
||||
bool have_integers = type_indexes.contains(TypeIndex::Int64) || type_indexes.contains(TypeIndex::UInt64);
|
||||
@ -300,7 +294,12 @@ namespace
|
||||
{
|
||||
WhichDataType which(type);
|
||||
if (which.isInt64() || which.isUInt64())
|
||||
type = std::make_shared<DataTypeFloat64>();
|
||||
{
|
||||
auto new_type = std::make_shared<DataTypeFloat64>();
|
||||
if (json_info && json_info->numbers_parsed_from_json_strings.erase(type.get()))
|
||||
json_info->numbers_parsed_from_json_strings.insert(new_type.get());
|
||||
type = new_type;
|
||||
}
|
||||
}
|
||||
|
||||
type_indexes.erase(TypeIndex::Int64);
|
||||
@ -635,9 +634,9 @@ namespace
|
||||
if (settings.try_infer_integers)
|
||||
{
|
||||
/// Transform Int64 to UInt64 if needed.
|
||||
transformIntegers(data_types, type_indexes);
|
||||
transformIntegers(data_types, type_indexes, json_info);
|
||||
/// Transform integers to floats if needed.
|
||||
transformIntegersAndFloatsToFloats(data_types, type_indexes);
|
||||
transformIntegersAndFloatsToFloats(data_types, type_indexes, json_info);
|
||||
}
|
||||
|
||||
/// Transform Date to DateTime or both to String if needed.
|
||||
@ -887,7 +886,7 @@ namespace
|
||||
}
|
||||
|
||||
template <bool is_json>
|
||||
DataTypePtr tryInferNumber(ReadBuffer & buf, const FormatSettings & settings)
|
||||
DataTypePtr tryInferNumber(ReadBuffer & buf, const FormatSettings & settings, JSONInferenceInfo * json_info)
|
||||
{
|
||||
if (buf.eof())
|
||||
return nullptr;
|
||||
@ -911,7 +910,12 @@ namespace
|
||||
Int64 tmp_int;
|
||||
buf.position() = number_start;
|
||||
if (tryReadIntText(tmp_int, buf))
|
||||
return std::make_shared<DataTypeInt64>();
|
||||
{
|
||||
auto type = std::make_shared<DataTypeInt64>();
|
||||
if (json_info && tmp_int < 0)
|
||||
json_info->negative_integers.insert(type.get());
|
||||
return type;
|
||||
}
|
||||
|
||||
/// In case of Int64 overflow we can try to infer UInt64.
|
||||
UInt64 tmp_uint;
|
||||
@ -934,7 +938,12 @@ namespace
|
||||
|
||||
Int64 tmp_int;
|
||||
if (tryReadIntText(tmp_int, peekable_buf))
|
||||
return std::make_shared<DataTypeInt64>();
|
||||
{
|
||||
auto type = std::make_shared<DataTypeInt64>();
|
||||
if (json_info && tmp_int < 0)
|
||||
json_info->negative_integers.insert(type.get());
|
||||
return type;
|
||||
}
|
||||
peekable_buf.rollbackToCheckpoint(/* drop= */ true);
|
||||
|
||||
/// In case of Int64 overflow we can try to infer UInt64.
|
||||
@ -952,7 +961,7 @@ namespace
|
||||
}
|
||||
|
||||
template <bool is_json>
|
||||
DataTypePtr tryInferNumberFromStringImpl(std::string_view field, const FormatSettings & settings)
|
||||
DataTypePtr tryInferNumberFromStringImpl(std::string_view field, const FormatSettings & settings, JSONInferenceInfo * json_inference_info = nullptr)
|
||||
{
|
||||
ReadBufferFromString buf(field);
|
||||
|
||||
@ -960,7 +969,12 @@ namespace
|
||||
{
|
||||
Int64 tmp_int;
|
||||
if (tryReadIntText(tmp_int, buf) && buf.eof())
|
||||
return std::make_shared<DataTypeInt64>();
|
||||
{
|
||||
auto type = std::make_shared<DataTypeInt64>();
|
||||
if (json_inference_info && tmp_int < 0)
|
||||
json_inference_info->negative_integers.insert(type.get());
|
||||
return type;
|
||||
}
|
||||
|
||||
/// We can safely get back to the start of buffer, because we read from a string and we didn't reach eof.
|
||||
buf.position() = buf.buffer().begin();
|
||||
@ -1011,7 +1025,7 @@ namespace
|
||||
{
|
||||
if (settings.json.try_infer_numbers_from_strings)
|
||||
{
|
||||
if (auto number_type = tryInferNumberFromStringImpl<true>(field, settings))
|
||||
if (auto number_type = tryInferNumberFromStringImpl<true>(field, settings, json_info))
|
||||
{
|
||||
json_info->numbers_parsed_from_json_strings.insert(number_type.get());
|
||||
return number_type;
|
||||
@ -1254,10 +1268,23 @@ namespace
|
||||
}
|
||||
|
||||
/// Number
|
||||
return tryInferNumber<is_json>(buf, settings);
|
||||
return tryInferNumber<is_json>(buf, settings, json_info);
|
||||
}
|
||||
}
|
||||
|
||||
bool checkIfTypesAreEqual(const DataTypes & types)
|
||||
{
|
||||
if (types.empty())
|
||||
return true;
|
||||
|
||||
for (size_t i = 1; i < types.size(); ++i)
|
||||
{
|
||||
if (!types[0]->equals(*types[i]))
|
||||
return false;
|
||||
}
|
||||
return true;
|
||||
}
|
||||
|
||||
void transformInferredTypesIfNeeded(DataTypePtr & first, DataTypePtr & second, const FormatSettings & settings)
|
||||
{
|
||||
DataTypes types = {first, second};
|
||||
@ -1275,6 +1302,11 @@ void transformInferredJSONTypesIfNeeded(
|
||||
second = std::move(types[1]);
|
||||
}
|
||||
|
||||
void transformInferredJSONTypesIfNeeded(DataTypes & types, const FormatSettings & settings, JSONInferenceInfo * json_info)
|
||||
{
|
||||
transformInferredTypesIfNeededImpl<true>(types, settings, json_info);
|
||||
}
|
||||
|
||||
void transformInferredJSONTypesFromDifferentFilesIfNeeded(DataTypePtr & first, DataTypePtr & second, const FormatSettings & settings)
|
||||
{
|
||||
JSONInferenceInfo json_info;
|
||||
@ -1396,6 +1428,12 @@ DataTypePtr tryInferNumberFromString(std::string_view field, const FormatSetting
|
||||
return tryInferNumberFromStringImpl<false>(field, settings);
|
||||
}
|
||||
|
||||
DataTypePtr tryInferJSONNumberFromString(std::string_view field, const FormatSettings & settings, JSONInferenceInfo * json_info)
|
||||
{
|
||||
return tryInferNumberFromStringImpl<false>(field, settings, json_info);
|
||||
|
||||
}
|
||||
|
||||
DataTypePtr tryInferDateOrDateTimeFromString(std::string_view field, const FormatSettings & settings)
|
||||
{
|
||||
if (settings.try_infer_dates && tryInferDate(field))
|
||||
|
||||
@ -2,6 +2,7 @@
|
||||
|
||||
#include <DataTypes/IDataType.h>
|
||||
#include <IO/ReadBuffer.h>
|
||||
#include <Formats/FormatSettings.h>
|
||||
|
||||
#include <vector>
|
||||
|
||||
@ -18,6 +19,11 @@ struct JSONInferenceInfo
|
||||
/// We store numbers that were parsed from strings.
|
||||
/// It's used in types transformation to change such numbers back to string if needed.
|
||||
std::unordered_set<const IDataType *> numbers_parsed_from_json_strings;
|
||||
/// Store integer types that were inferred from negative numbers.
|
||||
/// It's used to determine common type for Int64 and UInt64
|
||||
/// TODO: check it not only in JSON formats.
|
||||
std::unordered_set<const IDataType *> negative_integers;
|
||||
|
||||
/// Indicates if currently we are inferring type for Map/Object key.
|
||||
bool is_object_key = false;
|
||||
/// When we transform types for the same column from different files
|
||||
@ -48,6 +54,7 @@ DataTypePtr tryInferDateOrDateTimeFromString(std::string_view field, const Forma
|
||||
/// Try to parse a number value from a string. By default, it tries to parse Float64,
|
||||
/// but if setting try_infer_integers is enabled, it also tries to parse Int64.
|
||||
DataTypePtr tryInferNumberFromString(std::string_view field, const FormatSettings & settings);
|
||||
DataTypePtr tryInferJSONNumberFromString(std::string_view field, const FormatSettings & settings, JSONInferenceInfo * json_info);
|
||||
|
||||
/// It takes two types inferred for the same column and tries to transform them to a common type if possible.
|
||||
/// It's also used when we try to infer some not ordinary types from another types.
|
||||
@ -77,6 +84,7 @@ void transformInferredTypesIfNeeded(DataTypePtr & first, DataTypePtr & second, c
|
||||
/// Example 2:
|
||||
/// We merge DataTypeJSONPaths types to a single DataTypeJSONPaths type with union of all JSON paths.
|
||||
void transformInferredJSONTypesIfNeeded(DataTypePtr & first, DataTypePtr & second, const FormatSettings & settings, JSONInferenceInfo * json_info);
|
||||
void transformInferredJSONTypesIfNeeded(DataTypes & types, const FormatSettings & settings, JSONInferenceInfo * json_info);
|
||||
|
||||
/// Make final transform for types inferred in JSON format. It does 3 types of transformation:
|
||||
/// 1) Checks if type is unnamed Tuple(...), tries to transform nested types to find a common type for them and if all nested types
|
||||
@ -107,4 +115,6 @@ NamesAndTypesList getNamesAndRecursivelyNullableTypes(const Block & header);
|
||||
/// Check if type contains Nothing, like Array(Tuple(Nullable(Nothing), String))
|
||||
bool checkIfTypeIsComplete(const DataTypePtr & type);
|
||||
|
||||
bool checkIfTypesAreEqual(const DataTypes & types);
|
||||
|
||||
}
|
||||
|
||||
File diff suppressed because it is too large
Load Diff
File diff suppressed because it is too large
Load Diff
@ -49,7 +49,7 @@ namespace
|
||||
const String & dest_blob_,
|
||||
std::shared_ptr<const AzureBlobStorage::RequestSettings> settings_,
|
||||
ThreadPoolCallbackRunnerUnsafe<void> schedule_,
|
||||
const Poco::Logger * log_)
|
||||
LoggerPtr log_)
|
||||
: create_read_buffer(create_read_buffer_)
|
||||
, client(client_)
|
||||
, offset (offset_)
|
||||
@ -74,7 +74,7 @@ namespace
|
||||
const String & dest_blob;
|
||||
std::shared_ptr<const AzureBlobStorage::RequestSettings> settings;
|
||||
ThreadPoolCallbackRunnerUnsafe<void> schedule;
|
||||
const Poco::Logger * log;
|
||||
const LoggerPtr log;
|
||||
size_t max_single_part_upload_size;
|
||||
|
||||
struct UploadPartTask
|
||||
@ -83,7 +83,6 @@ namespace
|
||||
size_t part_size;
|
||||
std::vector<std::string> block_ids;
|
||||
bool is_finished = false;
|
||||
std::exception_ptr exception;
|
||||
};
|
||||
|
||||
size_t normal_part_size;
|
||||
@ -92,6 +91,7 @@ namespace
|
||||
std::list<UploadPartTask> TSA_GUARDED_BY(bg_tasks_mutex) bg_tasks;
|
||||
int num_added_bg_tasks TSA_GUARDED_BY(bg_tasks_mutex) = 0;
|
||||
int num_finished_bg_tasks TSA_GUARDED_BY(bg_tasks_mutex) = 0;
|
||||
std::exception_ptr bg_exception TSA_GUARDED_BY(bg_tasks_mutex);
|
||||
std::mutex bg_tasks_mutex;
|
||||
std::condition_variable bg_tasks_condvar;
|
||||
|
||||
@ -186,7 +186,7 @@ namespace
|
||||
}
|
||||
catch (...)
|
||||
{
|
||||
tryLogCurrentException(__PRETTY_FUNCTION__);
|
||||
tryLogCurrentException(log, fmt::format("While performing multipart upload of blob {} in container {}", dest_blob, dest_container_for_logging));
|
||||
waitForAllBackgroundTasks();
|
||||
throw;
|
||||
}
|
||||
@ -242,7 +242,12 @@ namespace
|
||||
}
|
||||
catch (...)
|
||||
{
|
||||
task->exception = std::current_exception();
|
||||
std::lock_guard lock(bg_tasks_mutex);
|
||||
if (!bg_exception)
|
||||
{
|
||||
tryLogCurrentException(log, "While writing part");
|
||||
bg_exception = std::current_exception(); /// The exception will be rethrown after all background tasks stop working.
|
||||
}
|
||||
}
|
||||
task_finish_notify();
|
||||
}, Priority{});
|
||||
@ -299,13 +304,13 @@ namespace
|
||||
/// Suppress warnings because bg_tasks_mutex is actually hold, but tsa annotations do not understand std::unique_lock
|
||||
bg_tasks_condvar.wait(lock, [this]() {return TSA_SUPPRESS_WARNING_FOR_READ(num_added_bg_tasks) == TSA_SUPPRESS_WARNING_FOR_READ(num_finished_bg_tasks); });
|
||||
|
||||
auto & tasks = TSA_SUPPRESS_WARNING_FOR_WRITE(bg_tasks);
|
||||
for (auto & task : tasks)
|
||||
{
|
||||
if (task.exception)
|
||||
std::rethrow_exception(task.exception);
|
||||
auto exception = TSA_SUPPRESS_WARNING_FOR_READ(bg_exception);
|
||||
if (exception)
|
||||
std::rethrow_exception(exception);
|
||||
|
||||
const auto & tasks = TSA_SUPPRESS_WARNING_FOR_READ(bg_tasks);
|
||||
for (const auto & task : tasks)
|
||||
block_ids.insert(block_ids.end(),task.block_ids.begin(), task.block_ids.end());
|
||||
}
|
||||
}
|
||||
};
|
||||
}
|
||||
@ -321,7 +326,8 @@ void copyDataToAzureBlobStorageFile(
|
||||
std::shared_ptr<const AzureBlobStorage::RequestSettings> settings,
|
||||
ThreadPoolCallbackRunnerUnsafe<void> schedule)
|
||||
{
|
||||
UploadHelper helper{create_read_buffer, dest_client, offset, size, dest_container_for_logging, dest_blob, settings, schedule, &Poco::Logger::get("copyDataToAzureBlobStorageFile")};
|
||||
auto log = getLogger("copyDataToAzureBlobStorageFile");
|
||||
UploadHelper helper{create_read_buffer, dest_client, offset, size, dest_container_for_logging, dest_blob, settings, schedule, log};
|
||||
helper.performCopy();
|
||||
}
|
||||
|
||||
@ -339,9 +345,11 @@ void copyAzureBlobStorageFile(
|
||||
const ReadSettings & read_settings,
|
||||
ThreadPoolCallbackRunnerUnsafe<void> schedule)
|
||||
{
|
||||
auto log = getLogger("copyAzureBlobStorageFile");
|
||||
|
||||
if (settings->use_native_copy)
|
||||
{
|
||||
LOG_TRACE(getLogger("copyAzureBlobStorageFile"), "Copying Blob: {} from Container: {} using native copy", src_container_for_logging, src_blob);
|
||||
LOG_TRACE(log, "Copying Blob: {} from Container: {} using native copy", src_container_for_logging, src_blob);
|
||||
ProfileEvents::increment(ProfileEvents::AzureCopyObject);
|
||||
if (dest_client->GetClickhouseOptions().IsClientForDisk)
|
||||
ProfileEvents::increment(ProfileEvents::DiskAzureCopyObject);
|
||||
@ -352,7 +360,7 @@ void copyAzureBlobStorageFile(
|
||||
|
||||
if (size < settings->max_single_part_copy_size)
|
||||
{
|
||||
LOG_TRACE(getLogger("copyAzureBlobStorageFile"), "Copy blob sync {} -> {}", src_blob, dest_blob);
|
||||
LOG_TRACE(log, "Copy blob sync {} -> {}", src_blob, dest_blob);
|
||||
block_blob_client_dest.CopyFromUri(source_uri);
|
||||
}
|
||||
else
|
||||
@ -368,7 +376,7 @@ void copyAzureBlobStorageFile(
|
||||
|
||||
if (copy_status.HasValue() && copy_status.Value() == Azure::Storage::Blobs::Models::CopyStatus::Success)
|
||||
{
|
||||
LOG_TRACE(getLogger("copyAzureBlobStorageFile"), "Copy of {} to {} finished", properties_model.CopySource.Value(), dest_blob);
|
||||
LOG_TRACE(log, "Copy of {} to {} finished", properties_model.CopySource.Value(), dest_blob);
|
||||
}
|
||||
else
|
||||
{
|
||||
@ -382,14 +390,14 @@ void copyAzureBlobStorageFile(
|
||||
}
|
||||
else
|
||||
{
|
||||
LOG_TRACE(&Poco::Logger::get("copyAzureBlobStorageFile"), "Reading from Container: {}, Blob: {}", src_container_for_logging, src_blob);
|
||||
LOG_TRACE(log, "Reading from Container: {}, Blob: {}", src_container_for_logging, src_blob);
|
||||
auto create_read_buffer = [&]
|
||||
{
|
||||
return std::make_unique<ReadBufferFromAzureBlobStorage>(
|
||||
src_client, src_blob, read_settings, settings->max_single_read_retries, settings->max_single_download_retries);
|
||||
};
|
||||
|
||||
UploadHelper helper{create_read_buffer, dest_client, offset, size, dest_container_for_logging, dest_blob, settings, schedule, &Poco::Logger::get("copyAzureBlobStorageFile")};
|
||||
UploadHelper helper{create_read_buffer, dest_client, offset, size, dest_container_for_logging, dest_blob, settings, schedule, log};
|
||||
helper.performCopy();
|
||||
}
|
||||
}
|
||||
|
||||
@ -98,7 +98,6 @@ namespace
|
||||
size_t part_size;
|
||||
String tag;
|
||||
bool is_finished = false;
|
||||
std::exception_ptr exception;
|
||||
};
|
||||
|
||||
size_t num_parts;
|
||||
@ -111,6 +110,7 @@ namespace
|
||||
size_t num_added_bg_tasks TSA_GUARDED_BY(bg_tasks_mutex) = 0;
|
||||
size_t num_finished_bg_tasks TSA_GUARDED_BY(bg_tasks_mutex) = 0;
|
||||
size_t num_finished_parts TSA_GUARDED_BY(bg_tasks_mutex) = 0;
|
||||
std::exception_ptr bg_exception TSA_GUARDED_BY(bg_tasks_mutex);
|
||||
std::mutex bg_tasks_mutex;
|
||||
std::condition_variable bg_tasks_condvar;
|
||||
|
||||
@ -273,7 +273,7 @@ namespace
|
||||
}
|
||||
catch (...)
|
||||
{
|
||||
tryLogCurrentException(__PRETTY_FUNCTION__);
|
||||
tryLogCurrentException(log, fmt::format("While performing multipart upload of {}", dest_key));
|
||||
// Multipart upload failed because it wasn't possible to schedule all the tasks.
|
||||
// To avoid execution of already scheduled tasks we abort MultipartUpload.
|
||||
abortMultipartUpload();
|
||||
@ -385,7 +385,12 @@ namespace
|
||||
}
|
||||
catch (...)
|
||||
{
|
||||
task->exception = std::current_exception();
|
||||
std::lock_guard lock(bg_tasks_mutex);
|
||||
if (!bg_exception)
|
||||
{
|
||||
tryLogCurrentException(log, fmt::format("While writing part #{}", task->part_number));
|
||||
bg_exception = std::current_exception(); /// The exception will be rethrown after all background tasks stop working.
|
||||
}
|
||||
}
|
||||
task_finish_notify();
|
||||
}, Priority{});
|
||||
@ -435,22 +440,21 @@ namespace
|
||||
/// Suppress warnings because bg_tasks_mutex is actually hold, but tsa annotations do not understand std::unique_lock
|
||||
bg_tasks_condvar.wait(lock, [this]() {return TSA_SUPPRESS_WARNING_FOR_READ(num_added_bg_tasks) == TSA_SUPPRESS_WARNING_FOR_READ(num_finished_bg_tasks); });
|
||||
|
||||
auto & tasks = TSA_SUPPRESS_WARNING_FOR_WRITE(bg_tasks);
|
||||
for (auto & task : tasks)
|
||||
auto exception = TSA_SUPPRESS_WARNING_FOR_READ(bg_exception);
|
||||
if (exception)
|
||||
{
|
||||
if (task.exception)
|
||||
{
|
||||
/// abortMultipartUpload() might be called already, see processUploadPartRequest().
|
||||
/// However if there were concurrent uploads at that time, those part uploads might or might not succeed.
|
||||
/// As a result, it might be necessary to abort a given multipart upload multiple times in order to completely free
|
||||
/// all storage consumed by all parts.
|
||||
abortMultipartUpload();
|
||||
/// abortMultipartUpload() might be called already, see processUploadPartRequest().
|
||||
/// However if there were concurrent uploads at that time, those part uploads might or might not succeed.
|
||||
/// As a result, it might be necessary to abort a given multipart upload multiple times in order to completely free
|
||||
/// all storage consumed by all parts.
|
||||
abortMultipartUpload();
|
||||
|
||||
std::rethrow_exception(task.exception);
|
||||
}
|
||||
|
||||
part_tags.push_back(task.tag);
|
||||
std::rethrow_exception(exception);
|
||||
}
|
||||
|
||||
const auto & tasks = TSA_SUPPRESS_WARNING_FOR_READ(bg_tasks);
|
||||
for (const auto & task : tasks)
|
||||
part_tags.push_back(task.tag);
|
||||
}
|
||||
};
|
||||
|
||||
|
||||
@ -73,66 +73,55 @@ static bool tryExtractConstValueFromCondition(const ASTPtr & condition, bool & v
|
||||
return false;
|
||||
}
|
||||
|
||||
void OptimizeIfWithConstantConditionVisitor::visit(ASTPtr & current_ast)
|
||||
void OptimizeIfWithConstantConditionVisitorData::visit(ASTFunction & function_node, ASTPtr & ast)
|
||||
{
|
||||
if (!current_ast)
|
||||
return;
|
||||
|
||||
checkStackSize();
|
||||
|
||||
for (ASTPtr & child : current_ast->children)
|
||||
if (function_node.name != "if")
|
||||
return;
|
||||
|
||||
if (!function_node.arguments)
|
||||
throw Exception(ErrorCodes::NUMBER_OF_ARGUMENTS_DOESNT_MATCH, "Wrong number of arguments for function 'if' (0 instead of 3)");
|
||||
|
||||
if (function_node.arguments->children.size() != 3)
|
||||
throw Exception(ErrorCodes::NUMBER_OF_ARGUMENTS_DOESNT_MATCH,
|
||||
"Wrong number of arguments for function 'if' ({} instead of 3)",
|
||||
function_node.arguments->children.size());
|
||||
|
||||
const auto * args = function_node.arguments->as<ASTExpressionList>();
|
||||
|
||||
ASTPtr condition_expr = args->children[0];
|
||||
ASTPtr then_expr = args->children[1];
|
||||
ASTPtr else_expr = args->children[2];
|
||||
|
||||
bool condition;
|
||||
if (tryExtractConstValueFromCondition(condition_expr, condition))
|
||||
{
|
||||
auto * function_node = child->as<ASTFunction>();
|
||||
if (!function_node || function_node->name != "if")
|
||||
ASTPtr replace_ast = condition ? then_expr : else_expr;
|
||||
ASTPtr child_copy = ast;
|
||||
String replace_alias = replace_ast->tryGetAlias();
|
||||
String if_alias = ast->tryGetAlias();
|
||||
|
||||
if (replace_alias.empty())
|
||||
{
|
||||
visit(child);
|
||||
continue;
|
||||
replace_ast->setAlias(if_alias);
|
||||
ast = replace_ast;
|
||||
}
|
||||
else
|
||||
{
|
||||
/// Only copy of one node is required here.
|
||||
/// But IAST has only method for deep copy of subtree.
|
||||
/// This can be a reason of performance degradation in case of deep queries.
|
||||
ASTPtr replace_ast_deep_copy = replace_ast->clone();
|
||||
replace_ast_deep_copy->setAlias(if_alias);
|
||||
ast = replace_ast_deep_copy;
|
||||
}
|
||||
|
||||
if (!function_node->arguments)
|
||||
throw Exception(ErrorCodes::NUMBER_OF_ARGUMENTS_DOESNT_MATCH, "Wrong number of arguments for function 'if' (0 instead of 3)");
|
||||
|
||||
if (function_node->arguments->children.size() != 3)
|
||||
throw Exception(ErrorCodes::NUMBER_OF_ARGUMENTS_DOESNT_MATCH,
|
||||
"Wrong number of arguments for function 'if' ({} instead of 3)",
|
||||
function_node->arguments->children.size());
|
||||
|
||||
visit(function_node->arguments);
|
||||
const auto * args = function_node->arguments->as<ASTExpressionList>();
|
||||
|
||||
ASTPtr condition_expr = args->children[0];
|
||||
ASTPtr then_expr = args->children[1];
|
||||
ASTPtr else_expr = args->children[2];
|
||||
|
||||
bool condition;
|
||||
if (tryExtractConstValueFromCondition(condition_expr, condition))
|
||||
if (!if_alias.empty())
|
||||
{
|
||||
ASTPtr replace_ast = condition ? then_expr : else_expr;
|
||||
ASTPtr child_copy = child;
|
||||
String replace_alias = replace_ast->tryGetAlias();
|
||||
String if_alias = child->tryGetAlias();
|
||||
|
||||
if (replace_alias.empty())
|
||||
{
|
||||
replace_ast->setAlias(if_alias);
|
||||
child = replace_ast;
|
||||
}
|
||||
else
|
||||
{
|
||||
/// Only copy of one node is required here.
|
||||
/// But IAST has only method for deep copy of subtree.
|
||||
/// This can be a reason of performance degradation in case of deep queries.
|
||||
ASTPtr replace_ast_deep_copy = replace_ast->clone();
|
||||
replace_ast_deep_copy->setAlias(if_alias);
|
||||
child = replace_ast_deep_copy;
|
||||
}
|
||||
|
||||
if (!if_alias.empty())
|
||||
{
|
||||
auto alias_it = aliases.find(if_alias);
|
||||
if (alias_it != aliases.end() && alias_it->second.get() == child_copy.get())
|
||||
alias_it->second = child;
|
||||
}
|
||||
auto alias_it = aliases.find(if_alias);
|
||||
if (alias_it != aliases.end() && alias_it->second.get() == child_copy.get())
|
||||
alias_it->second = ast;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
@ -1,23 +1,24 @@
|
||||
#pragma once
|
||||
|
||||
#include <Interpreters/Aliases.h>
|
||||
#include <Interpreters/InDepthNodeVisitor.h>
|
||||
|
||||
namespace DB
|
||||
{
|
||||
|
||||
/// It removes Function_if node from AST if condition is constant.
|
||||
/// TODO: rewrite with InDepthNodeVisitor
|
||||
class OptimizeIfWithConstantConditionVisitor
|
||||
struct OptimizeIfWithConstantConditionVisitorData
|
||||
{
|
||||
public:
|
||||
explicit OptimizeIfWithConstantConditionVisitor(Aliases & aliases_)
|
||||
using TypeToVisit = ASTFunction;
|
||||
|
||||
explicit OptimizeIfWithConstantConditionVisitorData(Aliases & aliases_)
|
||||
: aliases(aliases_)
|
||||
{}
|
||||
|
||||
void visit(ASTPtr & ast);
|
||||
|
||||
void visit(ASTFunction & function_node, ASTPtr & ast);
|
||||
private:
|
||||
Aliases & aliases;
|
||||
};
|
||||
|
||||
/// It removes Function_if node from AST if condition is constant.
|
||||
using OptimizeIfWithConstantConditionVisitor = InDepthNodeVisitor<OneTypeMatcher<OptimizeIfWithConstantConditionVisitorData>, false>;
|
||||
|
||||
}
|
||||
|
||||
@ -577,7 +577,8 @@ void TreeOptimizer::optimizeIf(ASTPtr & query, Aliases & aliases, bool if_chain_
|
||||
optimizeMultiIfToIf(query);
|
||||
|
||||
/// Optimize if with constant condition after constants was substituted instead of scalar subqueries.
|
||||
OptimizeIfWithConstantConditionVisitor(aliases).visit(query);
|
||||
OptimizeIfWithConstantConditionVisitorData visitor_data(aliases);
|
||||
OptimizeIfWithConstantConditionVisitor(visitor_data).visit(query);
|
||||
|
||||
if (if_chain_to_multiif)
|
||||
OptimizeIfChainsVisitor().visit(query);
|
||||
|
||||
@ -5,20 +5,21 @@
|
||||
|
||||
#include <base/hex.h>
|
||||
#include <base/interpolate.h>
|
||||
#include <Common/FailPoint.h>
|
||||
#include <Common/Macros.h>
|
||||
#include <Common/MemoryTracker.h>
|
||||
#include <Common/ProfileEventsScope.h>
|
||||
#include <Common/StringUtils.h>
|
||||
#include <Common/ThreadFuzzer.h>
|
||||
#include <Common/ZooKeeper/KeeperException.h>
|
||||
#include <Common/ZooKeeper/Types.h>
|
||||
#include <Common/escapeForFileName.h>
|
||||
#include <Common/formatReadable.h>
|
||||
#include <Common/logger_useful.h>
|
||||
#include <Common/noexcept_scope.h>
|
||||
#include <Common/randomDelay.h>
|
||||
#include <Common/thread_local_rng.h>
|
||||
#include <Common/typeid_cast.h>
|
||||
#include <Common/ThreadFuzzer.h>
|
||||
#include <Common/FailPoint.h>
|
||||
#include <Common/randomDelay.h>
|
||||
|
||||
#include <Core/ServerUUID.h>
|
||||
|
||||
@ -5272,6 +5273,8 @@ void StorageReplicatedMergeTree::flushAndPrepareForShutdown()
|
||||
if (shutdown_prepared_called.exchange(true))
|
||||
return;
|
||||
|
||||
LOG_TRACE(log, "Start preparing for shutdown");
|
||||
|
||||
try
|
||||
{
|
||||
auto settings_ptr = getSettings();
|
||||
@ -5282,7 +5285,11 @@ void StorageReplicatedMergeTree::flushAndPrepareForShutdown()
|
||||
stopBeingLeader();
|
||||
|
||||
if (attach_thread)
|
||||
{
|
||||
attach_thread->shutdown();
|
||||
LOG_TRACE(log, "The attach thread is shutdown");
|
||||
}
|
||||
|
||||
|
||||
restarting_thread.shutdown(/* part_of_full_shutdown */true);
|
||||
/// Explicitly set the event, because the restarting thread will not set it again
|
||||
@ -5295,6 +5302,8 @@ void StorageReplicatedMergeTree::flushAndPrepareForShutdown()
|
||||
shutdown_deadline.emplace(std::chrono::system_clock::now());
|
||||
throw;
|
||||
}
|
||||
|
||||
LOG_TRACE(log, "Finished preparing for shutdown");
|
||||
}
|
||||
|
||||
void StorageReplicatedMergeTree::partialShutdown()
|
||||
@ -5332,6 +5341,8 @@ void StorageReplicatedMergeTree::shutdown(bool)
|
||||
if (shutdown_called.exchange(true))
|
||||
return;
|
||||
|
||||
LOG_TRACE(log, "Shutdown started");
|
||||
|
||||
flushAndPrepareForShutdown();
|
||||
|
||||
if (!shutdown_deadline.has_value())
|
||||
@ -5374,6 +5385,7 @@ void StorageReplicatedMergeTree::shutdown(bool)
|
||||
/// Wait for all of them
|
||||
std::lock_guard lock(data_parts_exchange_ptr->rwlock);
|
||||
}
|
||||
LOG_TRACE(log, "Shutdown finished");
|
||||
}
|
||||
|
||||
|
||||
|
||||
@ -15,3 +15,4 @@ warn_return_any = True
|
||||
no_implicit_reexport = True
|
||||
strict_equality = True
|
||||
extra_checks = True
|
||||
ignore_missing_imports = True
|
||||
@ -15,7 +15,7 @@ import upload_result_helper
|
||||
from build_check import get_release_or_pr
|
||||
from ci_config import CI
|
||||
from ci_metadata import CiMetadata
|
||||
from ci_utils import GHActions, normalize_string
|
||||
from ci_utils import GHActions, normalize_string, Shell
|
||||
from clickhouse_helper import (
|
||||
CiLogsCredentials,
|
||||
ClickHouseHelper,
|
||||
@ -53,6 +53,7 @@ from stopwatch import Stopwatch
|
||||
from tee_popen import TeePopen
|
||||
from ci_cache import CiCache
|
||||
from ci_settings import CiSettings
|
||||
from ci_buddy import CIBuddy
|
||||
from version_helper import get_version_from_repo
|
||||
|
||||
# pylint: disable=too-many-lines
|
||||
@ -262,6 +263,8 @@ def check_missing_images_on_dockerhub(
|
||||
|
||||
|
||||
def _pre_action(s3, indata, pr_info):
|
||||
print("Clear dmesg")
|
||||
Shell.run("sudo dmesg --clear ||:")
|
||||
CommitStatusData.cleanup()
|
||||
JobReport.cleanup()
|
||||
BuildResult.cleanup()
|
||||
@ -992,6 +995,10 @@ def main() -> int:
|
||||
ci_settings,
|
||||
args.skip_jobs,
|
||||
)
|
||||
|
||||
if IS_CI and pr_info.is_pr:
|
||||
ci_cache.filter_out_not_affected_jobs()
|
||||
|
||||
ci_cache.print_status()
|
||||
|
||||
if IS_CI and not pr_info.is_merge_queue:
|
||||
@ -1118,6 +1125,14 @@ def main() -> int:
|
||||
|
||||
### POST action: start
|
||||
elif args.post:
|
||||
if Shell.check(
|
||||
"sudo dmesg -T | grep -q -e 'Out of memory: Killed process' -e 'oom_reaper: reaped process' -e 'oom-kill:constraint=CONSTRAINT_NONE'"
|
||||
):
|
||||
print("WARNING: OOM while job execution")
|
||||
CIBuddy(dry_run=not pr_info.is_release).post_error(
|
||||
"Out Of Memory", job_name=_get_ext_check_name(args.job_name)
|
||||
)
|
||||
|
||||
job_report = JobReport.load() if JobReport.exist() else None
|
||||
if job_report:
|
||||
ch_helper = ClickHouseHelper()
|
||||
|
||||
88
tests/ci/ci_buddy.py
Normal file
88
tests/ci/ci_buddy.py
Normal file
@ -0,0 +1,88 @@
|
||||
import json
|
||||
import os
|
||||
|
||||
import boto3
|
||||
import requests
|
||||
from botocore.exceptions import ClientError
|
||||
|
||||
from pr_info import PRInfo
|
||||
from ci_utils import Shell
|
||||
|
||||
|
||||
class CIBuddy:
|
||||
_HEADERS = {"Content-Type": "application/json"}
|
||||
|
||||
def __init__(self, dry_run=False):
|
||||
self.repo = os.getenv("GITHUB_REPOSITORY", "")
|
||||
self.dry_run = dry_run
|
||||
res = self._get_webhooks()
|
||||
self.test_channel = ""
|
||||
self.dev_ci_channel = ""
|
||||
if res:
|
||||
self.test_channel = json.loads(res)["test_channel"]
|
||||
self.dev_ci_channel = json.loads(res)["ci_channel"]
|
||||
self.job_name = os.getenv("CHECK_NAME", "unknown")
|
||||
pr_info = PRInfo()
|
||||
self.pr_number = pr_info.number
|
||||
self.head_ref = pr_info.head_ref
|
||||
self.commit_url = pr_info.commit_html_url
|
||||
|
||||
@staticmethod
|
||||
def _get_webhooks():
|
||||
name = "ci_buddy_web_hooks"
|
||||
|
||||
session = boto3.Session(region_name="us-east-1") # Replace with your region
|
||||
ssm_client = session.client("ssm")
|
||||
json_string = None
|
||||
try:
|
||||
response = ssm_client.get_parameter(
|
||||
Name=name,
|
||||
WithDecryption=True, # Set to True if the parameter is a SecureString
|
||||
)
|
||||
json_string = response["Parameter"]["Value"]
|
||||
except ClientError as e:
|
||||
print(f"An error occurred: {e}")
|
||||
|
||||
return json_string
|
||||
|
||||
def post(self, message, dry_run=None):
|
||||
if dry_run is None:
|
||||
dry_run = self.dry_run
|
||||
print(f"Posting slack message, dry_run [{dry_run}]")
|
||||
if dry_run:
|
||||
url = self.test_channel
|
||||
else:
|
||||
url = self.dev_ci_channel
|
||||
data = {"text": message}
|
||||
try:
|
||||
requests.post(url, headers=self._HEADERS, data=json.dumps(data), timeout=10)
|
||||
except Exception as e:
|
||||
print(f"ERROR: Failed to post message, ex {e}")
|
||||
|
||||
def post_error(self, error_description, job_name="", with_instance_info=True):
|
||||
instance_id, instance_type = "unknown", "unknown"
|
||||
if with_instance_info:
|
||||
instance_id = Shell.run("ec2metadata --instance-id") or instance_id
|
||||
instance_type = Shell.run("ec2metadata --instance-type") or instance_type
|
||||
if not job_name:
|
||||
job_name = os.getenv("CHECK_NAME", "unknown")
|
||||
line_err = f":red_circle: *Error: {error_description}*\n\n"
|
||||
line_ghr = f" *Runner:* `{instance_type}`, `{instance_id}`\n"
|
||||
line_job = f" *Job:* `{job_name}`\n"
|
||||
line_pr_ = f" *PR:* <https://github.com/{self.repo}/pull/{self.pr_number}|#{self.pr_number}>\n"
|
||||
line_br_ = f" *Branch:* `{self.head_ref}`, <{self.commit_url}|commit>\n"
|
||||
message = line_err
|
||||
message += line_job
|
||||
if with_instance_info:
|
||||
message += line_ghr
|
||||
if self.pr_number > 0:
|
||||
message += line_pr_
|
||||
else:
|
||||
message += line_br_
|
||||
self.post(message)
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
# test
|
||||
buddy = CIBuddy(dry_run=True)
|
||||
buddy.post_error("Out of memory")
|
||||
@ -674,6 +674,78 @@ class CiCache:
|
||||
bucket=S3_BUILDS_BUCKET, file_path=result_json_path, s3_path=s3_path
|
||||
)
|
||||
|
||||
def filter_out_not_affected_jobs(self):
|
||||
"""
|
||||
Filter is to be applied in PRs to remove jobs that are not affected by the change
|
||||
It removes jobs from @jobs_to_do if it is a:
|
||||
1. test job and it is in @jobs_to_wait (no need to wait not affected jobs in PRs)
|
||||
2. test job and it has finished on release branch (even if failed)
|
||||
2. build job which is not required by any test job that is left in @jobs_to_do
|
||||
|
||||
:return:
|
||||
"""
|
||||
# 1.
|
||||
remove_from_await_list = []
|
||||
for job_name, job_config in self.jobs_to_wait.items():
|
||||
if CI.is_test_job(job_name) and job_name != CI.JobNames.BUILD_CHECK:
|
||||
remove_from_await_list.append(job_name)
|
||||
for job in remove_from_await_list:
|
||||
print(f"Filter job [{job}] - test job and not affected by the change")
|
||||
del self.jobs_to_wait[job]
|
||||
del self.jobs_to_do[job]
|
||||
|
||||
# 2.
|
||||
remove_from_to_do = []
|
||||
for job_name, job_config in self.jobs_to_do.items():
|
||||
if CI.is_test_job(job_name) and job_name != CI.JobNames.BUILD_CHECK:
|
||||
batches_to_remove = []
|
||||
assert job_config.batches is not None
|
||||
for batch in job_config.batches:
|
||||
if self.is_failed(
|
||||
job_name, batch, job_config.num_batches, release_branch=True
|
||||
):
|
||||
print(
|
||||
f"Filter [{job_name}/{batch}] - not affected by the change (failed on release branch)"
|
||||
)
|
||||
batches_to_remove.append(batch)
|
||||
for batch in batches_to_remove:
|
||||
job_config.batches.remove(batch)
|
||||
if not job_config.batches:
|
||||
print(
|
||||
f"Filter [{job_name}] - not affected by the change (failed on release branch)"
|
||||
)
|
||||
remove_from_to_do.append(job_name)
|
||||
for job in remove_from_to_do:
|
||||
del self.jobs_to_do[job]
|
||||
|
||||
# 3.
|
||||
required_builds = [] # type: List[str]
|
||||
for job_name, job_config in self.jobs_to_do.items():
|
||||
if CI.is_test_job(job_name) and job_config.required_builds:
|
||||
required_builds += job_config.required_builds
|
||||
required_builds = list(set(required_builds))
|
||||
|
||||
remove_builds = [] # type: List[str]
|
||||
has_builds_to_do = False
|
||||
for job_name, job_config in self.jobs_to_do.items():
|
||||
if CI.is_build_job(job_name):
|
||||
if job_name not in required_builds:
|
||||
remove_builds.append(job_name)
|
||||
else:
|
||||
has_builds_to_do = True
|
||||
|
||||
for build_job in remove_builds:
|
||||
print(
|
||||
f"Filter build job [{build_job}] - not affected and not required by test jobs"
|
||||
)
|
||||
del self.jobs_to_do[build_job]
|
||||
if build_job in self.jobs_to_wait:
|
||||
del self.jobs_to_wait[build_job]
|
||||
|
||||
if not has_builds_to_do and CI.JobNames.BUILD_CHECK in self.jobs_to_do:
|
||||
print(f"Filter job [{CI.JobNames.BUILD_CHECK}] - no builds to do")
|
||||
del self.jobs_to_do[CI.JobNames.BUILD_CHECK]
|
||||
|
||||
def await_pending_jobs(self, is_release: bool, dry_run: bool = False) -> None:
|
||||
"""
|
||||
await pending jobs to be finished
|
||||
|
||||
@ -378,7 +378,7 @@ class CommonJobConfigs:
|
||||
),
|
||||
run_command='functional_test_check.py "$CHECK_NAME"',
|
||||
runner_type=Runners.FUNC_TESTER,
|
||||
timeout=10800,
|
||||
timeout=7200,
|
||||
)
|
||||
STATEFUL_TEST = JobConfig(
|
||||
job_name_keyword="stateful",
|
||||
|
||||
@ -1,4 +1,5 @@
|
||||
import os
|
||||
import subprocess
|
||||
from contextlib import contextmanager
|
||||
from pathlib import Path
|
||||
from typing import Any, Iterator, List, Union
|
||||
@ -42,3 +43,43 @@ class GHActions:
|
||||
for line in lines:
|
||||
print(line)
|
||||
print("::endgroup::")
|
||||
|
||||
|
||||
class Shell:
|
||||
@classmethod
|
||||
def run_strict(cls, command):
|
||||
subprocess.run(
|
||||
command + " 2>&1",
|
||||
shell=True,
|
||||
stdout=subprocess.PIPE,
|
||||
stderr=subprocess.PIPE,
|
||||
text=True,
|
||||
check=True,
|
||||
)
|
||||
|
||||
@classmethod
|
||||
def run(cls, command):
|
||||
res = ""
|
||||
result = subprocess.run(
|
||||
command,
|
||||
shell=True,
|
||||
stdout=subprocess.PIPE,
|
||||
stderr=subprocess.PIPE,
|
||||
text=True,
|
||||
check=False,
|
||||
)
|
||||
if result.returncode == 0:
|
||||
res = result.stdout
|
||||
return res.strip()
|
||||
|
||||
@classmethod
|
||||
def check(cls, command):
|
||||
result = subprocess.run(
|
||||
command + " 2>&1",
|
||||
shell=True,
|
||||
stdout=subprocess.PIPE,
|
||||
stderr=subprocess.PIPE,
|
||||
text=True,
|
||||
check=False,
|
||||
)
|
||||
return result.returncode == 0
|
||||
|
||||
@ -108,6 +108,7 @@ def get_run_command(
|
||||
"--privileged "
|
||||
f"{ci_logs_args}"
|
||||
f"--volume={repo_path}/tests:/usr/share/clickhouse-test "
|
||||
f"--volume={repo_path}/utils/grpc-client:/usr/share/clickhouse-utils/grpc-client "
|
||||
f"{volume_with_broken_test}"
|
||||
f"--volume={result_path}:/test_output "
|
||||
f"--volume={server_log_path}:/var/log/clickhouse-server "
|
||||
|
||||
@ -1,4 +1,5 @@
|
||||
"""Module to get the token for GitHub"""
|
||||
|
||||
from dataclasses import dataclass
|
||||
import json
|
||||
import time
|
||||
|
||||
@ -417,7 +417,7 @@ class TestCIConfig(unittest.TestCase):
|
||||
assert not ci_cache.jobs_to_skip
|
||||
assert not ci_cache.jobs_to_wait
|
||||
|
||||
# pretend there are pending jobs that we neet to wait
|
||||
# pretend there are pending jobs that we need to wait
|
||||
ci_cache.jobs_to_wait = dict(ci_cache.jobs_to_do)
|
||||
for job, config in ci_cache.jobs_to_wait.items():
|
||||
assert not config.pending_batches
|
||||
@ -489,3 +489,87 @@ class TestCIConfig(unittest.TestCase):
|
||||
self.assertCountEqual(
|
||||
list(ci_cache.jobs_to_do) + ci_cache.jobs_to_skip, all_jobs_in_wf
|
||||
)
|
||||
|
||||
def test_ci_py_filters_not_affected_jobs_in_prs(self):
|
||||
"""
|
||||
checks ci.py filters not affected jobs in PRs
|
||||
"""
|
||||
settings = CiSettings()
|
||||
settings.no_ci_cache = True
|
||||
pr_info = PRInfo(github_event=_TEST_EVENT_JSON)
|
||||
pr_info.event_type = EventType.PUSH
|
||||
pr_info.number = 0
|
||||
assert pr_info.is_release and not pr_info.is_merge_queue
|
||||
ci_cache = CIPY._configure_jobs(
|
||||
S3Helper(), pr_info, settings, skip_jobs=False, dry_run=True
|
||||
)
|
||||
self.assertTrue(not ci_cache.jobs_to_skip, "Must be no jobs in skip list")
|
||||
all_jobs_in_wf = list(ci_cache.jobs_to_do)
|
||||
assert not ci_cache.jobs_to_wait
|
||||
assert not ci_cache.jobs_to_skip
|
||||
|
||||
# pretend there are pending jobs that we need to wait
|
||||
for job, job_config in ci_cache.jobs_to_do.items():
|
||||
ci_cache.jobs_to_wait[job] = job_config
|
||||
|
||||
# remove couple tests from to_wait and
|
||||
# expect they are preserved in @jobs_to_to along with required package_asan
|
||||
del ci_cache.jobs_to_wait[CI.JobNames.STATELESS_TEST_ASAN]
|
||||
del ci_cache.jobs_to_wait[CI.JobNames.INTEGRATION_TEST_TSAN]
|
||||
del ci_cache.jobs_to_wait[CI.JobNames.STATELESS_TEST_MSAN]
|
||||
|
||||
# pretend we have some batches failed for one of the job from the to_do list
|
||||
failed_job = CI.JobNames.INTEGRATION_TEST_TSAN
|
||||
failed_job_config = ci_cache.jobs_to_do[failed_job]
|
||||
FAILED_BATCHES = [0, 3]
|
||||
for batch in FAILED_BATCHES:
|
||||
assert batch < failed_job_config.num_batches
|
||||
record = CiCache.Record(
|
||||
record_type=CiCache.RecordType.FAILED,
|
||||
job_name=failed_job,
|
||||
job_digest=ci_cache.job_digests[failed_job],
|
||||
batch=batch,
|
||||
num_batches=failed_job_config.num_batches,
|
||||
release_branch=True,
|
||||
)
|
||||
for record_t_, records_ in ci_cache.records.items():
|
||||
if record_t_.value == CiCache.RecordType.FAILED.value:
|
||||
records_[record.to_str_key()] = record
|
||||
|
||||
# pretend we have all batches failed for one of the job from the to_do list
|
||||
failed_job = CI.JobNames.STATELESS_TEST_MSAN
|
||||
failed_job_config = ci_cache.jobs_to_do[failed_job]
|
||||
assert failed_job_config.num_batches > 1
|
||||
for batch in range(failed_job_config.num_batches):
|
||||
record = CiCache.Record(
|
||||
record_type=CiCache.RecordType.FAILED,
|
||||
job_name=failed_job,
|
||||
job_digest=ci_cache.job_digests[failed_job],
|
||||
batch=batch,
|
||||
num_batches=failed_job_config.num_batches,
|
||||
release_branch=True,
|
||||
)
|
||||
for record_t_, records_ in ci_cache.records.items():
|
||||
if record_t_.value == CiCache.RecordType.FAILED.value:
|
||||
records_[record.to_str_key()] = record
|
||||
|
||||
ci_cache.filter_out_not_affected_jobs()
|
||||
expected_to_do = [
|
||||
CI.JobNames.STATELESS_TEST_ASAN,
|
||||
CI.BuildNames.PACKAGE_ASAN,
|
||||
CI.JobNames.INTEGRATION_TEST_TSAN,
|
||||
CI.BuildNames.PACKAGE_TSAN,
|
||||
CI.JobNames.BUILD_CHECK,
|
||||
]
|
||||
self.assertCountEqual(
|
||||
list(ci_cache.jobs_to_wait),
|
||||
[
|
||||
CI.BuildNames.PACKAGE_ASAN,
|
||||
CI.BuildNames.PACKAGE_TSAN,
|
||||
CI.JobNames.BUILD_CHECK,
|
||||
],
|
||||
)
|
||||
self.assertCountEqual(list(ci_cache.jobs_to_do), expected_to_do)
|
||||
self.assertTrue(ci_cache.jobs_to_do[CI.JobNames.INTEGRATION_TEST_TSAN].batches)
|
||||
for batch in ci_cache.jobs_to_do[CI.JobNames.INTEGRATION_TEST_TSAN].batches:
|
||||
self.assertTrue(batch not in FAILED_BATCHES)
|
||||
|
||||
@ -1750,7 +1750,7 @@ class TestCase:
|
||||
return TestResult(
|
||||
self.name,
|
||||
TestStatus.FAIL,
|
||||
FailureReason.INTERNAL_QUERY_FAIL,
|
||||
FailureReason.TIMEOUT,
|
||||
total_time,
|
||||
self.add_info_about_settings(
|
||||
self.get_description_from_exception_info(sys.exc_info())
|
||||
@ -2189,11 +2189,26 @@ def run_tests_array(all_tests_with_params: Tuple[List[str], int, TestSuite, bool
|
||||
sys.stdout.flush()
|
||||
|
||||
while True:
|
||||
test_result = test_case.run(
|
||||
args, test_suite, client_options, server_logs_level
|
||||
)
|
||||
test_result = test_case.process_result(test_result, MESSAGES)
|
||||
if not test_result.need_retry:
|
||||
# This is the upper level timeout
|
||||
# It helps with completely frozen processes, like in case of gdb errors
|
||||
def timeout_handler(signum, frame):
|
||||
raise TimeoutError("Test execution timed out")
|
||||
|
||||
signal.signal(signal.SIGALRM, timeout_handler)
|
||||
signal.alarm(int(args.timeout * 1.1))
|
||||
test_result = None
|
||||
try:
|
||||
test_result = test_case.run(
|
||||
args, test_suite, client_options, server_logs_level
|
||||
)
|
||||
test_result = test_case.process_result(test_result, MESSAGES)
|
||||
break
|
||||
except TimeoutError:
|
||||
break
|
||||
finally:
|
||||
signal.alarm(0)
|
||||
|
||||
if not test_result or not test_result.need_retry:
|
||||
break
|
||||
restarted_tests.append(test_result)
|
||||
|
||||
@ -2452,6 +2467,10 @@ def override_envs(*args_, **kwargs):
|
||||
run_tests_array(*args_, **kwargs)
|
||||
|
||||
|
||||
def run_tests_process(*args, **kwargs):
|
||||
return run_tests_array(*args, **kwargs)
|
||||
|
||||
|
||||
def do_run_tests(jobs, test_suite: TestSuite):
|
||||
if jobs > 1 and len(test_suite.parallel_tests) > 0:
|
||||
print(
|
||||
@ -2475,39 +2494,70 @@ def do_run_tests(jobs, test_suite: TestSuite):
|
||||
# of failures will be nearly the same for all tests from the group.
|
||||
random.shuffle(test_suite.parallel_tests)
|
||||
|
||||
batch_size = max(1, len(test_suite.parallel_tests) // jobs)
|
||||
batch_size = max(1, (len(test_suite.parallel_tests) // jobs) + 1)
|
||||
parallel_tests_array = []
|
||||
for job in range(jobs):
|
||||
range_ = job * batch_size, job * batch_size + batch_size
|
||||
batch = test_suite.parallel_tests[range_[0] : range_[1]]
|
||||
parallel_tests_array.append((batch, batch_size, test_suite, True))
|
||||
|
||||
try:
|
||||
with multiprocessing.Pool(processes=jobs + 1) as pool:
|
||||
future = pool.map_async(run_tests_array, parallel_tests_array)
|
||||
processes = []
|
||||
|
||||
if args.run_sequential_tests_in_parallel:
|
||||
# Run parallel tests and sequential tests at the same time
|
||||
# Sequential tests will use different ClickHouse instance
|
||||
# In this process we can safely override values in `args` and `os.environ`
|
||||
future_seq = pool.map_async(
|
||||
override_envs,
|
||||
[
|
||||
(
|
||||
test_suite.sequential_tests,
|
||||
len(test_suite.sequential_tests),
|
||||
test_suite,
|
||||
False,
|
||||
)
|
||||
],
|
||||
)
|
||||
future_seq.wait()
|
||||
for test_batch in parallel_tests_array:
|
||||
process = multiprocessing.Process(
|
||||
target=run_tests_process, args=(test_batch,)
|
||||
)
|
||||
processes.append(process)
|
||||
process.start()
|
||||
|
||||
future.wait()
|
||||
finally:
|
||||
pool.terminate()
|
||||
pool.close()
|
||||
pool.join()
|
||||
if args.run_sequential_tests_in_parallel:
|
||||
# Run parallel tests and sequential tests at the same time
|
||||
# Sequential tests will use different ClickHouse instance
|
||||
# In this process we can safely override values in `args` and `os.environ`
|
||||
process = multiprocessing.Process(
|
||||
target=override_envs,
|
||||
args=(
|
||||
(
|
||||
test_suite.sequential_tests,
|
||||
len(test_suite.sequential_tests),
|
||||
test_suite,
|
||||
False,
|
||||
),
|
||||
),
|
||||
)
|
||||
processes.append(process)
|
||||
process.start()
|
||||
|
||||
while processes:
|
||||
sys.stdout.flush()
|
||||
# Periodically check the server for hangs
|
||||
# and stop all processes in this case
|
||||
try:
|
||||
clickhouse_execute(
|
||||
args,
|
||||
query="SELECT 1 /*hang up check*/",
|
||||
max_http_retries=5,
|
||||
timeout=20,
|
||||
)
|
||||
except Exception:
|
||||
print("Hang up check failed")
|
||||
server_died.set()
|
||||
|
||||
if server_died.is_set():
|
||||
print("Server died, terminating all processes...")
|
||||
kill_gdb_if_any()
|
||||
# Wait for test results
|
||||
sleep(args.timeout)
|
||||
for p in processes:
|
||||
if p.is_alive():
|
||||
p.terminate()
|
||||
break
|
||||
|
||||
for p in processes[:]:
|
||||
if not p.is_alive():
|
||||
processes.remove(p)
|
||||
|
||||
sleep(5)
|
||||
|
||||
if not args.run_sequential_tests_in_parallel:
|
||||
run_tests_array(
|
||||
@ -3358,6 +3408,14 @@ def parse_args():
|
||||
return parser.parse_args()
|
||||
|
||||
|
||||
class Terminated(KeyboardInterrupt):
|
||||
pass
|
||||
|
||||
|
||||
def signal_handler(sig, frame):
|
||||
raise Terminated(f"Terminated with {sig} signal")
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
stop_time = None
|
||||
exit_code = multiprocessing.Value("i", 0)
|
||||
@ -3369,6 +3427,9 @@ if __name__ == "__main__":
|
||||
# infinite tests processes left
|
||||
# (new process group is required to avoid killing some parent processes)
|
||||
os.setpgid(0, 0)
|
||||
signal.signal(signal.SIGTERM, signal_handler)
|
||||
signal.signal(signal.SIGINT, signal_handler)
|
||||
signal.signal(signal.SIGHUP, signal_handler)
|
||||
|
||||
try:
|
||||
args = parse_args()
|
||||
|
||||
3
tests/config/config.d/grpc_protocol.xml
Normal file
3
tests/config/config.d/grpc_protocol.xml
Normal file
@ -0,0 +1,3 @@
|
||||
<clickhouse>
|
||||
<grpc_port>9100</grpc_port>
|
||||
</clickhouse>
|
||||
@ -27,6 +27,7 @@ ln -sf $SRC_PATH/config.d/secure_ports.xml $DEST_SERVER_PATH/config.d/
|
||||
ln -sf $SRC_PATH/config.d/clusters.xml $DEST_SERVER_PATH/config.d/
|
||||
ln -sf $SRC_PATH/config.d/graphite.xml $DEST_SERVER_PATH/config.d/
|
||||
ln -sf $SRC_PATH/config.d/graphite_alternative.xml $DEST_SERVER_PATH/config.d/
|
||||
ln -sf $SRC_PATH/config.d/grpc_protocol.xml $DEST_SERVER_PATH/config.d/
|
||||
ln -sf $SRC_PATH/config.d/database_atomic.xml $DEST_SERVER_PATH/config.d/
|
||||
ln -sf $SRC_PATH/config.d/max_concurrent_queries.xml $DEST_SERVER_PATH/config.d/
|
||||
ln -sf $SRC_PATH/config.d/merge_tree_settings.xml $DEST_SERVER_PATH/config.d/
|
||||
|
||||
@ -78,13 +78,13 @@ def wait_rabbitmq_to_start(rabbitmq_docker_id, cookie, timeout=180):
|
||||
|
||||
def kill_rabbitmq(rabbitmq_id):
|
||||
p = subprocess.Popen(("docker", "stop", rabbitmq_id), stdout=subprocess.PIPE)
|
||||
p.communicate()
|
||||
p.wait(timeout=60)
|
||||
return p.returncode == 0
|
||||
|
||||
|
||||
def revive_rabbitmq(rabbitmq_id, cookie):
|
||||
p = subprocess.Popen(("docker", "start", rabbitmq_id), stdout=subprocess.PIPE)
|
||||
p.communicate()
|
||||
p.wait(timeout=60)
|
||||
wait_rabbitmq_to_start(rabbitmq_id, cookie)
|
||||
|
||||
|
||||
|
||||
@ -1 +1,2 @@
|
||||
42
|
||||
42
|
||||
|
||||
@ -17,3 +17,27 @@ ENGINE = MergeTree ORDER BY conversation;
|
||||
INSERT INTO t2(conversation) VALUES (42);
|
||||
|
||||
select * from t2;
|
||||
|
||||
drop table t1;
|
||||
|
||||
INSERT INTO t2(conversation) VALUES (42); -- { serverError UNKNOWN_TABLE }
|
||||
|
||||
drop table t2;
|
||||
|
||||
CREATE TABLE t2 (
|
||||
`conversation` UInt64,
|
||||
CONSTRAINT constraint_conversation CHECK conversation IN (SELECT id FROM t1)
|
||||
)
|
||||
ENGINE = MergeTree ORDER BY conversation;
|
||||
|
||||
INSERT INTO t2(conversation) VALUES (42); -- { serverError UNKNOWN_TABLE }
|
||||
|
||||
CREATE TABLE t1 (
|
||||
`id` UInt64
|
||||
)
|
||||
ENGINE = MergeTree ORDER BY id;
|
||||
|
||||
INSERT INTO t1(id) VALUES (42);
|
||||
|
||||
INSERT INTO t2(conversation) VALUES (42);
|
||||
select * from t2;
|
||||
|
||||
@ -0,0 +1,21 @@
|
||||
2020-01-01
|
||||
2020-01-01
|
||||
2020-01-01 00:00:00
|
||||
2020-01-01 00:00:00.000000
|
||||
127.0.0.1
|
||||
2001:db8:85a3::8a2e:370:7334
|
||||
42
|
||||
42
|
||||
42
|
||||
42
|
||||
42
|
||||
42
|
||||
42
|
||||
42
|
||||
42
|
||||
42
|
||||
Hello
|
||||
Hello
|
||||
\0\0\0
|
||||
Hello\0\0\0\0\0
|
||||
5801c962-1182-458a-89f8-d077da5074f9
|
||||
29
tests/queries/0_stateless/03198_json_extract_more_types.sql
Normal file
29
tests/queries/0_stateless/03198_json_extract_more_types.sql
Normal file
@ -0,0 +1,29 @@
|
||||
set allow_suspicious_low_cardinality_types=1;
|
||||
|
||||
select JSONExtract('{"a" : "2020-01-01"}', 'a', 'Date');
|
||||
select JSONExtract('{"a" : "2020-01-01"}', 'a', 'Date32');
|
||||
select JSONExtract('{"a" : "2020-01-01 00:00:00"}', 'a', 'DateTime');
|
||||
select JSONExtract('{"a" : "2020-01-01 00:00:00.000000"}', 'a', 'DateTime64(6)');
|
||||
select JSONExtract('{"a" : "127.0.0.1"}', 'a', 'IPv4');
|
||||
select JSONExtract('{"a" : "2001:0db8:85a3:0000:0000:8a2e:0370:7334"}', 'a', 'IPv6');
|
||||
|
||||
|
||||
select JSONExtract('{"a" : 42}', 'a', 'LowCardinality(UInt8)');
|
||||
select JSONExtract('{"a" : 42}', 'a', 'LowCardinality(Int8)');
|
||||
select JSONExtract('{"a" : 42}', 'a', 'LowCardinality(UInt16)');
|
||||
select JSONExtract('{"a" : 42}', 'a', 'LowCardinality(Int16)');
|
||||
select JSONExtract('{"a" : 42}', 'a', 'LowCardinality(UInt32)');
|
||||
select JSONExtract('{"a" : 42}', 'a', 'LowCardinality(Int32)');
|
||||
select JSONExtract('{"a" : 42}', 'a', 'LowCardinality(UInt64)');
|
||||
select JSONExtract('{"a" : 42}', 'a', 'LowCardinality(Int64)');
|
||||
|
||||
select JSONExtract('{"a" : 42}', 'a', 'LowCardinality(Float32)');
|
||||
select JSONExtract('{"a" : 42}', 'a', 'LowCardinality(Float32)');
|
||||
|
||||
select JSONExtract('{"a" : "Hello"}', 'a', 'LowCardinality(String)');
|
||||
select JSONExtract('{"a" : "Hello"}', 'a', 'LowCardinality(FixedString(5))');
|
||||
select JSONExtract('{"a" : "Hello"}', 'a', 'LowCardinality(FixedString(3))');
|
||||
select JSONExtract('{"a" : "Hello"}', 'a', 'LowCardinality(FixedString(10))');
|
||||
|
||||
select JSONExtract('{"a" : "5801c962-1182-458a-89f8-d077da5074f9"}', 'a', 'LowCardinality(UUID)');
|
||||
|
||||
@ -0,0 +1,30 @@
|
||||
true Bool
|
||||
42 Int64
|
||||
-42 Int64
|
||||
18446744073709551615 UInt64
|
||||
42.42 Float64
|
||||
42 Int64
|
||||
-42 Int64
|
||||
18446744073709551615 UInt64
|
||||
Hello String
|
||||
2020-01-01 Date
|
||||
2020-01-01 00:00:00.000000000 DateTime64(9)
|
||||
[1,2,3] Array(Nullable(Int64))
|
||||
['str1','str2','str3'] Array(Nullable(String))
|
||||
[[[1],[2,3,4]],[[5,6],[7]]] Array(Array(Array(Nullable(Int64))))
|
||||
['2020-01-01 00:00:00.000000000','2020-01-01 00:00:00.000000000'] Array(Nullable(DateTime64(9)))
|
||||
['2020-01-01','2020-01-01 date'] Array(Nullable(String))
|
||||
['2020-01-01','2020-01-01 00:00:00','str'] Array(Nullable(String))
|
||||
['2020-01-01','2020-01-01 00:00:00','42'] Array(Nullable(String))
|
||||
['str','42'] Array(Nullable(String))
|
||||
[42,42.42] Array(Nullable(Float64))
|
||||
[42,18446744073709552000,42.42] Array(Nullable(Float64))
|
||||
[42,42.42] Array(Nullable(Float64))
|
||||
[NULL,NULL] Array(Nullable(String))
|
||||
[NULL,42] Array(Nullable(Int64))
|
||||
[[NULL],[],[42]] Array(Array(Nullable(Int64)))
|
||||
[[],[NULL,NULL],[1,NULL,3],[NULL,2,NULL]] Array(Array(Nullable(Int64)))
|
||||
[[],[NULL,NULL],['1',NULL,'3'],[NULL,'2',NULL],['2020-01-01']] Array(Array(Nullable(String)))
|
||||
('str',42,[42]) Tuple(Nullable(String), Nullable(Int64), Array(Nullable(Int64)))
|
||||
[42,18446744073709551615] Array(Nullable(UInt64))
|
||||
(-42,18446744073709551615) Tuple(Nullable(Int64), Nullable(UInt64))
|
||||
37
tests/queries/0_stateless/03199_json_extract_dynamic.sql
Normal file
37
tests/queries/0_stateless/03199_json_extract_dynamic.sql
Normal file
@ -0,0 +1,37 @@
|
||||
set input_format_json_try_infer_numbers_from_strings=1;
|
||||
|
||||
select JSONExtract(materialize('{"d" : true}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : 42}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : -42}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : 18446744073709551615}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : 42.42}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : "42"}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : "-42"}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : "18446744073709551615"}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
|
||||
select JSONExtract(materialize('{"d" : "Hello"}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : "2020-01-01"}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : "2020-01-01 00:00:00.000"}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
|
||||
select JSONExtract(materialize('{"d" : [1, 2, 3]}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : ["str1", "str2", "str3"]}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : [[[1], [2, 3, 4]], [[5, 6], [7]]]}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
|
||||
select JSONExtract(materialize('{"d" : ["2020-01-01", "2020-01-01 00:00:00"]}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : ["2020-01-01", "2020-01-01 date"]}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : ["2020-01-01", "2020-01-01 00:00:00", "str"]}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : ["2020-01-01", "2020-01-01 00:00:00", "42"]}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : ["str", "42"]}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : [42, 42.42]}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : [42, 18446744073709551615, 42.42]}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : [42, 42.42]}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
|
||||
select JSONExtract(materialize('{"d" : [null, null]}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : [null, 42]}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : [[null], [], [42]]}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"a" : [[], [null, null], ["1", null, "3"], [null, "2", null]]}'), 'a', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"a" : [[], [null, null], ["1", null, "3"], [null, "2", null], ["2020-01-01"]]}'), 'a', 'Dynamic') as d, dynamicType(d);
|
||||
|
||||
select JSONExtract(materialize('{"d" : ["str", 42, [42]]}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : [42, 18446744073709551615]}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
select JSONExtract(materialize('{"d" : [-42, 18446744073709551615]}'), 'd', 'Dynamic') as d, dynamicType(d);
|
||||
1
tests/queries/0_stateless/03203_grpc_protocol.reference
Normal file
1
tests/queries/0_stateless/03203_grpc_protocol.reference
Normal file
@ -0,0 +1 @@
|
||||
ok
|
||||
14
tests/queries/0_stateless/03203_grpc_protocol.sh
Executable file
14
tests/queries/0_stateless/03203_grpc_protocol.sh
Executable file
@ -0,0 +1,14 @@
|
||||
#!/usr/bin/env bash
|
||||
# Tags: no-fasttest
|
||||
# Tag no-fasttest: In fasttest, ENABLE_LIBRARIES=0, so the grpc library is not built
|
||||
|
||||
CURDIR=$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)
|
||||
# shellcheck source=../shell_config.sh
|
||||
. "$CURDIR"/../shell_config.sh
|
||||
|
||||
if [[ -z "$CLICKHOUSE_GRPC_CLIENT" ]]; then
|
||||
CLICKHOUSE_GRPC_CLIENT="$CURDIR/../../../utils/grpc-client/clickhouse-grpc-client.py"
|
||||
fi
|
||||
|
||||
# Simple test.
|
||||
$CLICKHOUSE_GRPC_CLIENT --query "SELECT 'ok'"
|
||||
@ -1900,11 +1900,13 @@ kurtosis
|
||||
kurtpop
|
||||
kurtsamp
|
||||
laion
|
||||
lagInFrame
|
||||
lang
|
||||
laravel
|
||||
largestTriangleThreeBuckets
|
||||
latencies
|
||||
ldap
|
||||
leadInFrame
|
||||
leftPad
|
||||
leftPadUTF
|
||||
leftUTF
|
||||
|
||||
52
utils/grpc-client/generate_pb2.py
Executable file
52
utils/grpc-client/generate_pb2.py
Executable file
@ -0,0 +1,52 @@
|
||||
#!/usr/bin/env python3
|
||||
|
||||
# This is a helper utility.
|
||||
# It generates files in the "pb2" folder using the protocol buffer compiler.
|
||||
# This script must be called manually after any change pf "clickhouse_grpc.proto"
|
||||
|
||||
import grpc_tools # pip3 install grpcio-tools
|
||||
|
||||
import os, shutil, subprocess
|
||||
|
||||
|
||||
# Settings.
|
||||
script_path = os.path.realpath(__file__)
|
||||
script_name = os.path.basename(script_path)
|
||||
script_dir = os.path.dirname(script_path)
|
||||
root_dir = os.path.abspath(os.path.join(script_dir, "../.."))
|
||||
|
||||
grpc_proto_dir = os.path.abspath(os.path.join(root_dir, "src/Server/grpc_protos"))
|
||||
grpc_proto_filename = "clickhouse_grpc.proto"
|
||||
|
||||
# Files in the "pb2" folder which will be generated by this script.
|
||||
pb2_filenames = ["clickhouse_grpc_pb2.py", "clickhouse_grpc_pb2_grpc.py"]
|
||||
pb2_dir = os.path.join(script_dir, "pb2")
|
||||
|
||||
|
||||
# Processes the protobuf schema with the protocol buffer compiler and generates the "pb2" folder.
|
||||
def generate_pb2():
|
||||
print(f"Generating files:")
|
||||
for pb2_filename in pb2_filenames:
|
||||
print(os.path.join(pb2_dir, pb2_filename))
|
||||
|
||||
os.makedirs(pb2_dir, exist_ok=True)
|
||||
|
||||
cmd = [
|
||||
"python3",
|
||||
"-m",
|
||||
"grpc_tools.protoc",
|
||||
"-I" + grpc_proto_dir,
|
||||
"--python_out=" + pb2_dir,
|
||||
"--grpc_python_out=" + pb2_dir,
|
||||
os.path.join(grpc_proto_dir, grpc_proto_filename),
|
||||
]
|
||||
subprocess.run(cmd)
|
||||
|
||||
for pb2_filename in pb2_filenames:
|
||||
assert os.path.exists(os.path.join(pb2_dir, pb2_filename))
|
||||
print("Done! (generate_pb2)")
|
||||
|
||||
|
||||
# MAIN
|
||||
if __name__ == "__main__":
|
||||
generate_pb2()
|
||||
@ -1,29 +0,0 @@
|
||||
#!/usr/bin/env python3
|
||||
|
||||
import grpc_tools # pip3 install grpcio-tools
|
||||
|
||||
import os
|
||||
import subprocess
|
||||
|
||||
|
||||
script_dir = os.path.dirname(os.path.realpath(__file__))
|
||||
dest_dir = script_dir
|
||||
src_dir = os.path.abspath(os.path.join(script_dir, "../../../src/Server/grpc_protos"))
|
||||
src_filename = "clickhouse_grpc.proto"
|
||||
|
||||
|
||||
def generate():
|
||||
cmd = [
|
||||
"python3",
|
||||
"-m",
|
||||
"grpc_tools.protoc",
|
||||
"-I" + src_dir,
|
||||
"--python_out=" + dest_dir,
|
||||
"--grpc_python_out=" + dest_dir,
|
||||
os.path.join(src_dir, src_filename),
|
||||
]
|
||||
subprocess.run(cmd)
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
generate()
|
||||
Loading…
Reference in New Issue
Block a user