thevar1able/ClickHouse
1
0
Fork 0
mirror of https://github.com/ClickHouse/ClickHouse.git synced 2024-11-21 23:21:59 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge remote-tracking branch 'origin/master' into reject_invalid_mutation_queries

Browse Source
This commit is contained in:
Christoph Wurm 2024-11-12 13:53:01 +00:00
commit 1615cd1297
199 changed files with 4312 additions and 830 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
.github/PULL_REQUEST_TEMPLATE.md vendored
View File

@ -12,7 +12,7 @@ tests/ci/cancel_and_rerun_workflow_lambda/app.py
- Backward Incompatible Change
- Build/Testing/Packaging Improvement
- Documentation (changelog entry is not required)
- Critical Bug Fix (crash, LOGICAL_ERROR, data loss, RBAC)
- Critical Bug Fix (crash, data loss, RBAC)
- Bug Fix (user-visible misbehavior in an official stable release)
- CI Fix or Improvement (changelog entry is not required)
- Not for changelog (changelog entry is not required)

2
contrib/SimSIMD vendored

@ -1 +1 @@
Subproject commit ee3c9c9c00b51645f62a1a9e99611b78c0052a21
Subproject commit fa60f1b8e3582c50978f0ae86c2ebb6c9af957f3

30
docker/keeper/entrypoint.sh
View File

@ -1,21 +1,31 @@
#!/bin/bash
set +x
set -eo pipefail
shopt -s nullglob
DO_CHOWN=1
if [ "${CLICKHOUSE_DO_NOT_CHOWN:-0}" = "1" ]; then
if [[ "${CLICKHOUSE_RUN_AS_ROOT:=0}" = "1" || "${CLICKHOUSE_DO_NOT_CHOWN:-0}" = "1" ]]; then
DO_CHOWN=0
fi
CLICKHOUSE_UID="${CLICKHOUSE_UID:-"$(id -u clickhouse)"}"
CLICKHOUSE_GID="${CLICKHOUSE_GID:-"$(id -g clickhouse)"}"
# CLICKHOUSE_UID and CLICKHOUSE_GID are kept for backward compatibility, but deprecated
# One must use either "docker run --user" or CLICKHOUSE_RUN_AS_ROOT=1 to run the process as
# FIXME: Remove ALL CLICKHOUSE_UID CLICKHOUSE_GID before 25.3
if [[ "${CLICKHOUSE_UID:-}" || "${CLICKHOUSE_GID:-}" ]]; then
echo 'WARNING: Support for CLICKHOUSE_UID/CLICKHOUSE_GID will be removed in a couple of releases.' >&2
echo 'WARNING: Either use a proper "docker run --user=xxx:xxxx" argument instead of CLICKHOUSE_UID/CLICKHOUSE_GID' >&2
echo 'WARNING: or set "CLICKHOUSE_RUN_AS_ROOT=1" ENV to run the clickhouse-server as root:root' >&2
fi
# support --user
if [ "$(id -u)" = "0" ]; then
USER=$CLICKHOUSE_UID
GROUP=$CLICKHOUSE_GID
# support `docker run --user=xxx:xxxx`
if [[ "$(id -u)" = "0" ]]; then
if [[ "$CLICKHOUSE_RUN_AS_ROOT" = 1 ]]; then
USER=0
GROUP=0
else
USER="${CLICKHOUSE_UID:-"$(id -u clickhouse)"}"
GROUP="${CLICKHOUSE_GID:-"$(id -g clickhouse)"}"
fi
if command -v gosu &> /dev/null; then
gosu="gosu $USER:$GROUP"
elif command -v su-exec &> /dev/null; then
@ -82,11 +92,11 @@ if [[ $# -lt 1 ]] || [[ "$1" == "--"* ]]; then
# There is a config file. It is already tested with gosu (if it is readably by keeper user)
if [ -f "$KEEPER_CONFIG" ]; then
exec $gosu /usr/bin/clickhouse-keeper --config-file="$KEEPER_CONFIG" "$@"
exec $gosu clickhouse-keeper --config-file="$KEEPER_CONFIG" "$@"
fi
# There is no config file. Will use embedded one
exec $gosu /usr/bin/clickhouse-keeper --log-file="$LOG_PATH" --errorlog-file="$ERROR_LOG_PATH" "$@"
exec $gosu clickhouse-keeper --log-file="$LOG_PATH" --errorlog-file="$ERROR_LOG_PATH" "$@"
fi
# Otherwise, we assume the user want to run his own process, for example a `bash` shell to explore this image

56
docker/server/Dockerfile.ubuntu
View File

@ -88,34 +88,32 @@ RUN if [ -n "${single_binary_location_url}" ]; then \
#docker-official-library:on
# A fallback to installation from ClickHouse repository
RUN if ! clickhouse local -q "SELECT ''" > /dev/null 2>&1; then \
apt-get update \
&& apt-get install --yes --no-install-recommends \
apt-transport-https \
dirmngr \
gnupg2 \
&& mkdir -p /etc/apt/sources.list.d \
&& GNUPGHOME=$(mktemp -d) \
&& GNUPGHOME="$GNUPGHOME" gpg --batch --no-default-keyring \
--keyring /usr/share/keyrings/clickhouse-keyring.gpg \
--keyserver hkp://keyserver.ubuntu.com:80 \
--recv-keys 3a9ea1193a97b548be1457d48919f6bd2b48d754 \
&& rm -rf "$GNUPGHOME" \
&& chmod +r /usr/share/keyrings/clickhouse-keyring.gpg \
&& echo "${REPOSITORY}" > /etc/apt/sources.list.d/clickhouse.list \
&& echo "installing from repository: ${REPOSITORY}" \
&& apt-get update \
&& for package in ${PACKAGES}; do \
packages="${packages} ${package}=${VERSION}" \
; done \
&& apt-get install --allow-unauthenticated --yes --no-install-recommends ${packages} || exit 1 \
&& rm -rf \
/var/lib/apt/lists/* \
/var/cache/debconf \
/tmp/* \
&& apt-get autoremove --purge -yq libksba8 \
&& apt-get autoremove -yq \
; fi
# It works unless the clickhouse binary already exists
RUN clickhouse local -q 'SELECT 1' >/dev/null 2>&1 && exit 0 || : \
; apt-get update \
&& apt-get install --yes --no-install-recommends \
dirmngr \
gnupg2 \
&& mkdir -p /etc/apt/sources.list.d \
&& GNUPGHOME=$(mktemp -d) \
&& GNUPGHOME="$GNUPGHOME" gpg --batch --no-default-keyring \
--keyring /usr/share/keyrings/clickhouse-keyring.gpg \
--keyserver hkp://keyserver.ubuntu.com:80 \
--recv-keys 3a9ea1193a97b548be1457d48919f6bd2b48d754 \
&& rm -rf "$GNUPGHOME" \
&& chmod +r /usr/share/keyrings/clickhouse-keyring.gpg \
&& echo "${REPOSITORY}" > /etc/apt/sources.list.d/clickhouse.list \
&& echo "installing from repository: ${REPOSITORY}" \
&& apt-get update \
&& for package in ${PACKAGES}; do \
packages="${packages} ${package}=${VERSION}" \
; done \
&& apt-get install --yes --no-install-recommends ${packages} || exit 1 \
&& rm -rf \
/var/lib/apt/lists/* \
/var/cache/debconf \
/tmp/* \
&& apt-get autoremove --purge -yq dirmngr gnupg2
# post install
# we need to allow "others" access to clickhouse folder, because docker container
@ -126,8 +124,6 @@ RUN clickhouse-local -q 'SELECT * FROM system.build_options' \
RUN locale-gen en_US.UTF-8
ENV LANG en_US.UTF-8
ENV LANGUAGE en_US:en
ENV LC_ALL en_US.UTF-8
ENV TZ UTC
RUN mkdir /docker-entrypoint-initdb.d

30
docker/server/README.md
View File

@ -1,3 +1,11 @@
<!---
The README.md is generated by README.sh from the following sources:
- README.src/content.md
- README.src/license.md
If you want to change it, edit these files
-->
# ClickHouse Server Docker Image
## What is ClickHouse?
@ -8,6 +16,7 @@ ClickHouse works 100-1000x faster than traditional database management systems,
For more information and documentation see https://clickhouse.com/.
<!-- This is not related to the docker official library, remove it before commit to https://github.com/docker-library/docs -->
## Versions
- The `latest` tag points to the latest release of the latest stable branch.
@ -16,11 +25,12 @@ For more information and documentation see https://clickhouse.com/.
- The tag `head` is built from the latest commit to the default branch.
- Each tag has optional `-alpine` suffix to reflect that it's built on top of `alpine`.
<!-- REMOVE UNTIL HERE -->
### Compatibility
- The amd64 image requires support for [SSE3 instructions](https://en.wikipedia.org/wiki/SSE3). Virtually all x86 CPUs after 2005 support SSE3.
- The arm64 image requires support for the [ARMv8.2-A architecture](https://en.wikipedia.org/wiki/AArch64#ARMv8.2-A) and additionally the Load-Acquire RCpc register. The register is optional in version ARMv8.2-A and mandatory in [ARMv8.3-A](https://en.wikipedia.org/wiki/AArch64#ARMv8.3-A). Supported in Graviton >=2, Azure and GCP instances. Examples for unsupported devices are Raspberry Pi 4 (ARMv8.0-A) and Jetson AGX Xavier/Orin (ARMv8.2-A).
- Since the Clickhouse 24.11 Ubuntu images started using `ubuntu:22.04` as its base image. It requires docker version >= `20.10.10` containing [patch](https://github.com/moby/moby/commit/977283509f75303bc6612665a04abf76ff1d2468). As a workaround you could use `docker run [--privileged | --security-opt seccomp=unconfined]` instead, however that has security implications.
- Since the Clickhouse 24.11 Ubuntu images started using `ubuntu:22.04` as its base image. It requires docker version >= `20.10.10` containing [patch](https://github.com/moby/moby/commit/977283509f75303bc6612665a04abf76ff1d2468). As a workaround you could use `docker run --security-opt seccomp=unconfined` instead, however that has security implications.
## How to use this image
@ -30,7 +40,7 @@ For more information and documentation see https://clickhouse.com/.
docker run -d --name some-clickhouse-server --ulimit nofile=262144:262144 clickhouse/clickhouse-server
```
By default, ClickHouse will be accessible only via the Docker network. See the [networking section below](#networking).
By default, ClickHouse will be accessible only via the Docker network. See the **networking** section below.
By default, starting above server instance will be run as the `default` user without password.
@ -47,7 +57,7 @@ More information about the [ClickHouse client](https://clickhouse.com/docs/en/in
### connect to it using curl
```bash
echo "SELECT 'Hello, ClickHouse!'" | docker run -i --rm --link some-clickhouse-server:clickhouse-server curlimages/curl 'http://clickhouse-server:8123/?query=' -s --data-binary @-
echo "SELECT 'Hello, ClickHouse!'" | docker run -i --rm --link some-clickhouse-server:clickhouse-server buildpack-deps:curl curl 'http://clickhouse-server:8123/?query=' -s --data-binary @-
```
More information about the [ClickHouse HTTP Interface](https://clickhouse.com/docs/en/interfaces/http/).
@ -70,7 +80,7 @@ echo 'SELECT version()' | curl 'http://localhost:18123/' --data-binary @-
`22.6.3.35`
or by allowing the container to use [host ports directly](https://docs.docker.com/network/host/) using `--network=host` (also allows achieving better network performance):
Or by allowing the container to use [host ports directly](https://docs.docker.com/network/host/) using `--network=host` (also allows achieving better network performance):
```bash
docker run -d --network=host --name some-clickhouse-server --ulimit nofile=262144:262144 clickhouse/clickhouse-server
@ -88,8 +98,8 @@ Typically you may want to mount the following folders inside your container to a
```bash
docker run -d \
-v $(realpath ./ch_data):/var/lib/clickhouse/ \
-v $(realpath ./ch_logs):/var/log/clickhouse-server/ \
-v "$PWD/ch_data:/var/lib/clickhouse/" \
-v "$PWD/ch_logs:/var/log/clickhouse-server/" \
--name some-clickhouse-server --ulimit nofile=262144:262144 clickhouse/clickhouse-server
```
@ -111,6 +121,8 @@ docker run -d \
--name some-clickhouse-server --ulimit nofile=262144:262144 clickhouse/clickhouse-server
```
Read more in [knowledge base](https://clickhouse.com/docs/knowledgebase/configure_cap_ipc_lock_and_cap_sys_nice_in_docker).
## Configuration
The container exposes port 8123 for the [HTTP interface](https://clickhouse.com/docs/en/interfaces/http_interface/) and port 9000 for the [native client](https://clickhouse.com/docs/en/interfaces/tcp/).
@ -126,8 +138,8 @@ docker run -d --name some-clickhouse-server --ulimit nofile=262144:262144 -v /pa
### Start server as custom user
```bash
# $(pwd)/data/clickhouse should exist and be owned by current user
docker run --rm --user ${UID}:${GID} --name some-clickhouse-server --ulimit nofile=262144:262144 -v "$(pwd)/logs/clickhouse:/var/log/clickhouse-server" -v "$(pwd)/data/clickhouse:/var/lib/clickhouse" clickhouse/clickhouse-server
# $PWD/data/clickhouse should exist and be owned by current user
docker run --rm --user "${UID}:${GID}" --name some-clickhouse-server --ulimit nofile=262144:262144 -v "$PWD/logs/clickhouse:/var/log/clickhouse-server" -v "$PWD/data/clickhouse:/var/lib/clickhouse" clickhouse/clickhouse-server
```
When you use the image with local directories mounted, you probably want to specify the user to maintain the proper file ownership. Use the `--user` argument and mount `/var/lib/clickhouse` and `/var/log/clickhouse-server` inside the container. Otherwise, the image will complain and not start.
@ -135,7 +147,7 @@ When you use the image with local directories mounted, you probably want to spec
### Start server from root (useful in case of enabled user namespace)
```bash
docker run --rm -e CLICKHOUSE_UID=0 -e CLICKHOUSE_GID=0 --name clickhouse-server-userns -v "$(pwd)/logs/clickhouse:/var/log/clickhouse-server" -v "$(pwd)/data/clickhouse:/var/lib/clickhouse" clickhouse/clickhouse-server
docker run --rm -e CLICKHOUSE_RUN_AS_ROOT=1 --name clickhouse-server-userns -v "$PWD/logs/clickhouse:/var/log/clickhouse-server" -v "$PWD/data/clickhouse:/var/lib/clickhouse" clickhouse/clickhouse-server
```
### How to create default database and user on starting

38
docker/server/README.sh Executable file
View File

@ -0,0 +1,38 @@
#!/usr/bin/env bash
set -ueo pipefail
# A script to generate README.sh close to as it done in https://github.com/docker-library/docs
WORKDIR=$(dirname "$0")
SCRIPT_NAME=$(basename "$0")
CONTENT=README.src/content.md
LICENSE=README.src/license.md
cd "$WORKDIR"
R=README.md
cat > "$R" <<EOD
<!---
The $R is generated by $SCRIPT_NAME from the following sources:
- $CONTENT
- $LICENSE
If you want to change it, edit these files
-->
EOD
cat "$CONTENT" >> "$R"
cat >> "$R" <<EOD
## License
$(cat $LICENSE)
EOD
# Remove %%LOGO%% from the file with one line below
sed -i '/^%%LOGO%%/,+1d' "$R"
# Replace each %%IMAGE%% with our `clickhouse/clickhouse-server`
sed -i '/%%IMAGE%%/s:%%IMAGE%%:clickhouse/clickhouse-server:g' $R

1
docker/server/README.src/README-short.txt Normal file
View File

@ -0,0 +1 @@
ClickHouse is the fastest and most resource efficient OSS database for real-time apps and analytics.

170
docker/server/README.src/content.md Normal file
View File

@ -0,0 +1,170 @@
# ClickHouse Server Docker Image
## What is ClickHouse?
%%LOGO%%
ClickHouse is an open-source column-oriented DBMS (columnar database management system) for online analytical processing (OLAP) that allows users to generate analytical reports using SQL queries in real-time.
ClickHouse works 100-1000x faster than traditional database management systems, and processes hundreds of millions to over a billion rows and tens of gigabytes of data per server per second. With a widespread user base around the globe, the technology has received praise for its reliability, ease of use, and fault tolerance.
For more information and documentation see https://clickhouse.com/.
<!-- This is not related to the docker official library, remove it before commit to https://github.com/docker-library/docs -->
## Versions
- The `latest` tag points to the latest release of the latest stable branch.
- Branch tags like `22.2` point to the latest release of the corresponding branch.
- Full version tags like `22.2.3.5` point to the corresponding release.
- The tag `head` is built from the latest commit to the default branch.
- Each tag has optional `-alpine` suffix to reflect that it's built on top of `alpine`.
<!-- REMOVE UNTIL HERE -->
### Compatibility
- The amd64 image requires support for [SSE3 instructions](https://en.wikipedia.org/wiki/SSE3). Virtually all x86 CPUs after 2005 support SSE3.
- The arm64 image requires support for the [ARMv8.2-A architecture](https://en.wikipedia.org/wiki/AArch64#ARMv8.2-A) and additionally the Load-Acquire RCpc register. The register is optional in version ARMv8.2-A and mandatory in [ARMv8.3-A](https://en.wikipedia.org/wiki/AArch64#ARMv8.3-A). Supported in Graviton >=2, Azure and GCP instances. Examples for unsupported devices are Raspberry Pi 4 (ARMv8.0-A) and Jetson AGX Xavier/Orin (ARMv8.2-A).
- Since the Clickhouse 24.11 Ubuntu images started using `ubuntu:22.04` as its base image. It requires docker version >= `20.10.10` containing [patch](https://github.com/moby/moby/commit/977283509f75303bc6612665a04abf76ff1d2468). As a workaround you could use `docker run --security-opt seccomp=unconfined` instead, however that has security implications.
## How to use this image
### start server instance
```bash
docker run -d --name some-clickhouse-server --ulimit nofile=262144:262144 %%IMAGE%%
```
By default, ClickHouse will be accessible only via the Docker network. See the **networking** section below.
By default, starting above server instance will be run as the `default` user without password.
### connect to it from a native client
```bash
docker run -it --rm --link some-clickhouse-server:clickhouse-server --entrypoint clickhouse-client %%IMAGE%% --host clickhouse-server
# OR
docker exec -it some-clickhouse-server clickhouse-client
```
More information about the [ClickHouse client](https://clickhouse.com/docs/en/interfaces/cli/).
### connect to it using curl
```bash
echo "SELECT 'Hello, ClickHouse!'" | docker run -i --rm --link some-clickhouse-server:clickhouse-server buildpack-deps:curl curl 'http://clickhouse-server:8123/?query=' -s --data-binary @-
```
More information about the [ClickHouse HTTP Interface](https://clickhouse.com/docs/en/interfaces/http/).
### stopping / removing the container
```bash
docker stop some-clickhouse-server
docker rm some-clickhouse-server
```
### networking
You can expose your ClickHouse running in docker by [mapping a particular port](https://docs.docker.com/config/containers/container-networking/) from inside the container using host ports:
```bash
docker run -d -p 18123:8123 -p19000:9000 --name some-clickhouse-server --ulimit nofile=262144:262144 %%IMAGE%%
echo 'SELECT version()' | curl 'http://localhost:18123/' --data-binary @-
```
`22.6.3.35`
Or by allowing the container to use [host ports directly](https://docs.docker.com/network/host/) using `--network=host` (also allows achieving better network performance):
```bash
docker run -d --network=host --name some-clickhouse-server --ulimit nofile=262144:262144 %%IMAGE%%
echo 'SELECT version()' | curl 'http://localhost:8123/' --data-binary @-
```
`22.6.3.35`
### Volumes
Typically you may want to mount the following folders inside your container to achieve persistency:
- `/var/lib/clickhouse/` - main folder where ClickHouse stores the data
- `/var/log/clickhouse-server/` - logs
```bash
docker run -d \
-v "$PWD/ch_data:/var/lib/clickhouse/" \
-v "$PWD/ch_logs:/var/log/clickhouse-server/" \
--name some-clickhouse-server --ulimit nofile=262144:262144 %%IMAGE%%
```
You may also want to mount:
- `/etc/clickhouse-server/config.d/*.xml` - files with server configuration adjustments
- `/etc/clickhouse-server/users.d/*.xml` - files with user settings adjustments
- `/docker-entrypoint-initdb.d/` - folder with database initialization scripts (see below).
### Linux capabilities
ClickHouse has some advanced functionality, which requires enabling several [Linux capabilities](https://man7.org/linux/man-pages/man7/capabilities.7.html).
They are optional and can be enabled using the following [docker command-line arguments](https://docs.docker.com/engine/reference/run/#runtime-privilege-and-linux-capabilities):
```bash
docker run -d \
--cap-add=SYS_NICE --cap-add=NET_ADMIN --cap-add=IPC_LOCK \
--name some-clickhouse-server --ulimit nofile=262144:262144 %%IMAGE%%
```
Read more in [knowledge base](https://clickhouse.com/docs/knowledgebase/configure_cap_ipc_lock_and_cap_sys_nice_in_docker).
## Configuration
The container exposes port 8123 for the [HTTP interface](https://clickhouse.com/docs/en/interfaces/http_interface/) and port 9000 for the [native client](https://clickhouse.com/docs/en/interfaces/tcp/).
ClickHouse configuration is represented with a file "config.xml" ([documentation](https://clickhouse.com/docs/en/operations/configuration_files/))
### Start server instance with custom configuration
```bash
docker run -d --name some-clickhouse-server --ulimit nofile=262144:262144 -v /path/to/your/config.xml:/etc/clickhouse-server/config.xml %%IMAGE%%
```
### Start server as custom user
```bash
# $PWD/data/clickhouse should exist and be owned by current user
docker run --rm --user "${UID}:${GID}" --name some-clickhouse-server --ulimit nofile=262144:262144 -v "$PWD/logs/clickhouse:/var/log/clickhouse-server" -v "$PWD/data/clickhouse:/var/lib/clickhouse" %%IMAGE%%
```
When you use the image with local directories mounted, you probably want to specify the user to maintain the proper file ownership. Use the `--user` argument and mount `/var/lib/clickhouse` and `/var/log/clickhouse-server` inside the container. Otherwise, the image will complain and not start.
### Start server from root (useful in case of enabled user namespace)
```bash
docker run --rm -e CLICKHOUSE_RUN_AS_ROOT=1 --name clickhouse-server-userns -v "$PWD/logs/clickhouse:/var/log/clickhouse-server" -v "$PWD/data/clickhouse:/var/lib/clickhouse" %%IMAGE%%
```
### How to create default database and user on starting
Sometimes you may want to create a user (user named `default` is used by default) and database on a container start. You can do it using environment variables `CLICKHOUSE_DB`, `CLICKHOUSE_USER`, `CLICKHOUSE_DEFAULT_ACCESS_MANAGEMENT` and `CLICKHOUSE_PASSWORD`:
```bash
docker run --rm -e CLICKHOUSE_DB=my_database -e CLICKHOUSE_USER=username -e CLICKHOUSE_DEFAULT_ACCESS_MANAGEMENT=1 -e CLICKHOUSE_PASSWORD=password -p 9000:9000/tcp %%IMAGE%%
```
## How to extend this image
To perform additional initialization in an image derived from this one, add one or more `*.sql`, `*.sql.gz`, or `*.sh` scripts under `/docker-entrypoint-initdb.d`. After the entrypoint calls `initdb`, it will run any `*.sql` files, run any executable `*.sh` scripts, and source any non-executable `*.sh` scripts found in that directory to do further initialization before starting the service.
Also, you can provide environment variables `CLICKHOUSE_USER` & `CLICKHOUSE_PASSWORD` that will be used for clickhouse-client during initialization.
For example, to add an additional user and database, add the following to `/docker-entrypoint-initdb.d/init-db.sh`:
```bash
#!/bin/bash
set -e
clickhouse client -n <<-EOSQL
CREATE DATABASE docker;
CREATE TABLE docker.docker (x Int32) ENGINE = Log;
EOSQL
```

1
docker/server/README.src/github-repo Normal file
View File

@ -0,0 +1 @@
https://github.com/ClickHouse/ClickHouse

1
docker/server/README.src/license.md Normal file
View File

@ -0,0 +1 @@
View [license information](https://github.com/ClickHouse/ClickHouse/blob/master/LICENSE) for the software contained in this image.

43
docker/server/README.src/logo.svg Normal file
View File

@ -0,0 +1,43 @@
<?xml version="1.0" encoding="UTF-8"?>
<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" viewBox="0 0 616 616">
<defs>
<style>
.cls-1 {
clip-path: url(#clippath);
}
.cls-2 {
fill: none;
}
.cls-2, .cls-3, .cls-4 {
stroke-width: 0px;
}
.cls-3 {
fill: #1e1e1e;
}
.cls-4 {
fill: #faff69;
}
</style>
<clipPath id="clippath">
<rect class="cls-2" x="83.23" y="71.73" width="472.55" height="472.55"/>
</clipPath>
</defs>
<g id="Layer_2" data-name="Layer 2">
<rect class="cls-4" width="616" height="616"/>
</g>
<g id="Layer_1" data-name="Layer 1">
<g class="cls-1">
<g>
<path class="cls-3" d="m120.14,113.3c0-2.57,2.09-4.66,4.66-4.66h34.98c2.57,0,4.66,2.09,4.66,4.66v389.38c0,2.57-2.09,4.66-4.66,4.66h-34.98c-2.57,0-4.66-2.09-4.66-4.66V113.3Z"/>
<path class="cls-3" d="m208.75,113.3c0-2.57,2.09-4.66,4.66-4.66h34.98c2.57,0,4.66,2.09,4.66,4.66v389.38c0,2.57-2.09,4.66-4.66,4.66h-34.98c-2.57,0-4.66-2.09-4.66-4.66V113.3Z"/>
<path class="cls-3" d="m297.35,113.3c0-2.57,2.09-4.66,4.66-4.66h34.98c2.57,0,4.66,2.09,4.66,4.66v389.38c0,2.57-2.09,4.66-4.66,4.66h-34.98c-2.57,0-4.66-2.09-4.66-4.66V113.3Z"/>
<path class="cls-3" d="m385.94,113.3c0-2.57,2.09-4.66,4.66-4.66h34.98c2.57,0,4.66,2.09,4.66,4.66v389.38c0,2.57-2.09,4.66-4.66,4.66h-34.98c-2.57,0-4.66-2.09-4.66-4.66V113.3Z"/>
<path class="cls-3" d="m474.56,268.36c0-2.57,2.09-4.66,4.66-4.66h34.98c2.57,0,4.65,2.09,4.65,4.66v79.28c0,2.57-2.09,4.66-4.65,4.66h-34.98c-2.57,0-4.66-2.09-4.66-4.66v-79.28Z"/>
</g>
</g>
</g>
</svg>
Side by Side

After

Width:  |  Height:  |  Size: 1.7 KiB

1
docker/server/README.src/maintainer.md Normal file
View File

@ -0,0 +1 @@
[ClickHouse Inc.](%%GITHUB-REPO%%)

7
docker/server/README.src/metadata.json Normal file
View File

@ -0,0 +1,7 @@
{
"hub": {
"categories": [
"databases-and-storage"
]
}
}

49
docker/server/entrypoint.sh
View File

@ -4,17 +4,28 @@ set -eo pipefail
shopt -s nullglob
DO_CHOWN=1
if [ "${CLICKHOUSE_DO_NOT_CHOWN:-0}" = "1" ]; then
if [[ "${CLICKHOUSE_RUN_AS_ROOT:=0}" = "1" || "${CLICKHOUSE_DO_NOT_CHOWN:-0}" = "1" ]]; then
DO_CHOWN=0
fi
CLICKHOUSE_UID="${CLICKHOUSE_UID:-"$(id -u clickhouse)"}"
CLICKHOUSE_GID="${CLICKHOUSE_GID:-"$(id -g clickhouse)"}"
# CLICKHOUSE_UID and CLICKHOUSE_GID are kept for backward compatibility, but deprecated
# One must use either "docker run --user" or CLICKHOUSE_RUN_AS_ROOT=1 to run the process as
# FIXME: Remove ALL CLICKHOUSE_UID CLICKHOUSE_GID before 25.3
if [[ "${CLICKHOUSE_UID:-}" || "${CLICKHOUSE_GID:-}" ]]; then
echo 'WARNING: Support for CLICKHOUSE_UID/CLICKHOUSE_GID will be removed in a couple of releases.' >&2
echo 'WARNING: Either use a proper "docker run --user=xxx:xxxx" argument instead of CLICKHOUSE_UID/CLICKHOUSE_GID' >&2
echo 'WARNING: or set "CLICKHOUSE_RUN_AS_ROOT=1" ENV to run the clickhouse-server as root:root' >&2
fi
# support --user
if [ "$(id -u)" = "0" ]; then
USER=$CLICKHOUSE_UID
GROUP=$CLICKHOUSE_GID
# support `docker run --user=xxx:xxxx`
if [[ "$(id -u)" = "0" ]]; then
if [[ "$CLICKHOUSE_RUN_AS_ROOT" = 1 ]]; then
USER=0
GROUP=0
else
USER="${CLICKHOUSE_UID:-"$(id -u clickhouse)"}"
GROUP="${CLICKHOUSE_GID:-"$(id -g clickhouse)"}"
fi
else
USER="$(id -u)"
GROUP="$(id -g)"
@ -55,14 +66,14 @@ function create_directory_and_do_chown() {
[ -z "$dir" ] && return
# ensure directories exist
if [ "$DO_CHOWN" = "1" ]; then
mkdir="mkdir"
mkdir=( mkdir )
else
# if DO_CHOWN=0 it means that the system does not map root user to "admin" permissions
# it mainly happens on NFS mounts where root==nobody for security reasons
# thus mkdir MUST run with user id/gid and not from nobody that has zero permissions
mkdir="/usr/bin/clickhouse su "${USER}:${GROUP}" mkdir"
mkdir=( clickhouse su "${USER}:${GROUP}" mkdir )
fi
if ! $mkdir -p "$dir"; then
if ! "${mkdir[@]}" -p "$dir"; then
echo "Couldn't create necessary directory: $dir"
exit 1
fi
@ -143,7 +154,7 @@ if [ -n "${RUN_INITDB_SCRIPTS}" ]; then
fi
# Listen only on localhost until the initialization is done
/usr/bin/clickhouse su "${USER}:${GROUP}" /usr/bin/clickhouse-server --config-file="$CLICKHOUSE_CONFIG" -- --listen_host=127.0.0.1 &
clickhouse su "${USER}:${GROUP}" clickhouse-server --config-file="$CLICKHOUSE_CONFIG" -- --listen_host=127.0.0.1 &
pid="$!"
# check if clickhouse is ready to accept connections
@ -151,7 +162,7 @@ if [ -n "${RUN_INITDB_SCRIPTS}" ]; then
tries=${CLICKHOUSE_INIT_TIMEOUT:-1000}
while ! wget --spider --no-check-certificate -T 1 -q "$URL" 2>/dev/null; do
if [ "$tries" -le "0" ]; then
echo >&2 'ClickHouse init process failed.'
echo >&2 'ClickHouse init process timeout.'
exit 1
fi
tries=$(( tries-1 ))
@ -203,18 +214,8 @@ if [[ $# -lt 1 ]] || [[ "$1" == "--"* ]]; then
CLICKHOUSE_WATCHDOG_ENABLE=${CLICKHOUSE_WATCHDOG_ENABLE:-0}
export CLICKHOUSE_WATCHDOG_ENABLE
# An option for easy restarting and replacing clickhouse-server in a container, especially in Kubernetes.
# For example, you can replace the clickhouse-server binary to another and restart it while keeping the container running.
if [[ "${CLICKHOUSE_DOCKER_RESTART_ON_EXIT:-0}" -eq "1" ]]; then
while true; do
# This runs the server as a child process of the shell script:
/usr/bin/clickhouse su "${USER}:${GROUP}" /usr/bin/clickhouse-server --config-file="$CLICKHOUSE_CONFIG" "$@" ||:
echo >&2 'ClickHouse Server exited, and the environment variable CLICKHOUSE_DOCKER_RESTART_ON_EXIT is set to 1. Restarting the server.'
done
else
# This replaces the shell script with the server:
exec /usr/bin/clickhouse su "${USER}:${GROUP}" /usr/bin/clickhouse-server --config-file="$CLICKHOUSE_CONFIG" "$@"
fi
# This replaces the shell script with the server:
exec clickhouse su "${USER}:${GROUP}" clickhouse-server --config-file="$CLICKHOUSE_CONFIG" "$@"
fi
# Otherwise, we assume the user want to run his own process, for example a `bash` shell to explore this image

44
docs/en/getting-started/example-datasets/tpch.md
View File

@ -33,6 +33,21 @@ Then, generate the data. Parameter `-s` specifies the scale factor. For example,
./dbgen -s 100
```
Detailed table sizes with scale factor 100:
| Table | size (in rows) | size (compressed in ClickHouse) |
|----------|----------------|---------------------------------|
| nation | 25 | 2 kB |
| region | 5 | 1 kB |
| part | 20.000.000 | 895 MB |
| supplier | 1.000.000 | 75 MB |
| partsupp | 80.000.000 | 4.37 GB |
| customer | 15.000.000 | 1.19 GB |
| orders | 150.000.000 | 6.15 GB |
| lineitem | 600.00.00 | 26.69 GB |
(Compressed sizes in ClickHouse are taken from `system.tables.total_bytes` and based on below table definitions.)
Now create tables in ClickHouse.
We stick as closely as possible to the rules of the TPC-H specification:
@ -151,10 +166,37 @@ clickhouse-client --format_csv_delimiter '|' --query "INSERT INTO orders FORMAT
clickhouse-client --format_csv_delimiter '|' --query "INSERT INTO lineitem FORMAT CSV" < lineitem.tbl
```
The queries are generated by `./qgen -s <scaling_factor>`. Example queries for `s = 100`:
:::note
Instead of using tpch-kit and generating the tables by yourself, you can alternatively import the data from a public S3 bucket. Make sure
to create empty tables first using above `CREATE` statements.
```sql
-- Scaling factor 1
INSERT INTO nation SELECT * FROM s3('https://clickhouse-datasets.s3.amazonaws.com/h/1/nation.tbl', NOSIGN, CSV) SETTINGS format_csv_delimiter = '|', input_format_defaults_for_omitted_fields = 1, input_format_csv_empty_as_default = 1;
INSERT INTO region SELECT * FROM s3('https://clickhouse-datasets.s3.amazonaws.com/h/1/region.tbl', NOSIGN, CSV) SETTINGS format_csv_delimiter = '|', input_format_defaults_for_omitted_fields = 1, input_format_csv_empty_as_default = 1;
INSERT INTO part SELECT * FROM s3('https://clickhouse-datasets.s3.amazonaws.com/h/1/part.tbl', NOSIGN, CSV) SETTINGS format_csv_delimiter = '|', input_format_defaults_for_omitted_fields = 1, input_format_csv_empty_as_default = 1;
INSERT INTO supplier SELECT * FROM s3('https://clickhouse-datasets.s3.amazonaws.com/h/1/supplier.tbl', NOSIGN, CSV) SETTINGS format_csv_delimiter = '|', input_format_defaults_for_omitted_fields = 1, input_format_csv_empty_as_default = 1;
INSERT INTO partsupp SELECT * FROM s3('https://clickhouse-datasets.s3.amazonaws.com/h/1/partsupp.tbl', NOSIGN, CSV) SETTINGS format_csv_delimiter = '|', input_format_defaults_for_omitted_fields = 1, input_format_csv_empty_as_default = 1;
INSERT INTO customer SELECT * FROM s3('https://clickhouse-datasets.s3.amazonaws.com/h/1/customer.tbl', NOSIGN, CSV) SETTINGS format_csv_delimiter = '|', input_format_defaults_for_omitted_fields = 1, input_format_csv_empty_as_default = 1;
INSERT INTO orders SELECT * FROM s3('https://clickhouse-datasets.s3.amazonaws.com/h/1/orders.tbl', NOSIGN, CSV) SETTINGS format_csv_delimiter = '|', input_format_defaults_for_omitted_fields = 1, input_format_csv_empty_as_default = 1;
INSERT INTO lineitem SELECT * FROM s3('https://clickhouse-datasets.s3.amazonaws.com/h/1/lineitem.tbl', NOSIGN, CSV) SETTINGS format_csv_delimiter = '|', input_format_defaults_for_omitted_fields = 1, input_format_csv_empty_as_default = 1;
-- Scaling factor 100
INSERT INTO nation SELECT * FROM s3('https://clickhouse-datasets.s3.amazonaws.com/h/100/nation.tbl.gz', NOSIGN, CSV) SETTINGS format_csv_delimiter = '|', input_format_defaults_for_omitted_fields = 1, input_format_csv_empty_as_default = 1;
INSERT INTO region SELECT * FROM s3('https://clickhouse-datasets.s3.amazonaws.com/h/100/region.tbl.gz', NOSIGN, CSV) SETTINGS format_csv_delimiter = '|', input_format_defaults_for_omitted_fields = 1, input_format_csv_empty_as_default = 1;
INSERT INTO part SELECT * FROM s3('https://clickhouse-datasets.s3.amazonaws.com/h/100/part.tbl.gz', NOSIGN, CSV) SETTINGS format_csv_delimiter = '|', input_format_defaults_for_omitted_fields = 1, input_format_csv_empty_as_default = 1;
INSERT INTO supplier SELECT * FROM s3('https://clickhouse-datasets.s3.amazonaws.com/h/100/supplier.tbl.gz', NOSIGN, CSV) SETTINGS format_csv_delimiter = '|', input_format_defaults_for_omitted_fields = 1, input_format_csv_empty_as_default = 1;
INSERT INTO partsupp SELECT * FROM s3('https://clickhouse-datasets.s3.amazonaws.com/h/100/partsupp.tbl.gz', NOSIGN, CSV) SETTINGS format_csv_delimiter = '|', input_format_defaults_for_omitted_fields = 1, input_format_csv_empty_as_default = 1;
INSERT INTO customer SELECT * FROM s3('https://clickhouse-datasets.s3.amazonaws.com/h/100/customer.tbl.gz', NOSIGN, CSV) SETTINGS format_csv_delimiter = '|', input_format_defaults_for_omitted_fields = 1, input_format_csv_empty_as_default = 1;
INSERT INTO orders SELECT * FROM s3('https://clickhouse-datasets.s3.amazonaws.com/h/100/orders.tbl.gz', NOSIGN, CSV) SETTINGS format_csv_delimiter = '|', input_format_defaults_for_omitted_fields = 1, input_format_csv_empty_as_default = 1;
INSERT INTO lineitem SELECT * FROM s3('https://clickhouse-datasets.s3.amazonaws.com/h/100/lineitem.tbl.gz', NOSIGN, CSV) SETTINGS format_csv_delimiter = '|', input_format_defaults_for_omitted_fields = 1, input_format_csv_empty_as_default = 1;
````
:::
## Queries
The queries are generated by `./qgen -s <scaling_factor>`. Example queries for `s = 100`:
**Correctness**
The result of the queries agrees with the official results unless mentioned otherwise. To verify, generate a TPC-H database with scale

117
docs/en/sql-reference/aggregate-functions/reference/index.md
View File

@ -7,119 +7,4 @@ toc_hidden: true
# List of Aggregate Functions
Standard aggregate functions:
- [count](../reference/count.md)
- [min](../reference/min.md)
- [max](../reference/max.md)
- [sum](../reference/sum.md)
- [avg](../reference/avg.md)
- [any](../reference/any.md)
- [stddevPop](../reference/stddevpop.md)
- [stddevPopStable](../reference/stddevpopstable.md)
- [stddevSamp](../reference/stddevsamp.md)
- [stddevSampStable](../reference/stddevsampstable.md)
- [varPop](../reference/varpop.md)
- [varSamp](../reference/varsamp.md)
- [corr](../reference/corr.md)
- [corr](../reference/corrstable.md)
- [corrMatrix](../reference/corrmatrix.md)
- [covarPop](../reference/covarpop.md)
- [covarStable](../reference/covarpopstable.md)
- [covarPopMatrix](../reference/covarpopmatrix.md)
- [covarSamp](../reference/covarsamp.md)
- [covarSampStable](../reference/covarsampstable.md)
- [covarSampMatrix](../reference/covarsampmatrix.md)
- [entropy](../reference/entropy.md)
- [exponentialMovingAverage](../reference/exponentialmovingaverage.md)
- [intervalLengthSum](../reference/intervalLengthSum.md)
- [kolmogorovSmirnovTest](../reference/kolmogorovsmirnovtest.md)
- [mannwhitneyutest](../reference/mannwhitneyutest.md)
- [median](../reference/median.md)
- [rankCorr](../reference/rankCorr.md)
- [sumKahan](../reference/sumkahan.md)
- [studentTTest](../reference/studentttest.md)
- [welchTTest](../reference/welchttest.md)
ClickHouse-specific aggregate functions:
- [aggThrow](../reference/aggthrow.md)
- [analysisOfVariance](../reference/analysis_of_variance.md)
- [any](../reference/any.md)
- [anyHeavy](../reference/anyheavy.md)
- [anyLast](../reference/anylast.md)
- [boundingRatio](../reference/boundrat.md)
- [first_value](../reference/first_value.md)
- [last_value](../reference/last_value.md)
- [argMin](../reference/argmin.md)
- [argMax](../reference/argmax.md)
- [avgWeighted](../reference/avgweighted.md)
- [topK](../reference/topk.md)
- [topKWeighted](../reference/topkweighted.md)
- [deltaSum](../reference/deltasum.md)
- [deltaSumTimestamp](../reference/deltasumtimestamp.md)
- [flameGraph](../reference/flame_graph.md)
- [groupArray](../reference/grouparray.md)
- [groupArrayLast](../reference/grouparraylast.md)
- [groupUniqArray](../reference/groupuniqarray.md)
- [groupArrayInsertAt](../reference/grouparrayinsertat.md)
- [groupArrayMovingAvg](../reference/grouparraymovingavg.md)
- [groupArrayMovingSum](../reference/grouparraymovingsum.md)
- [groupArraySample](../reference/grouparraysample.md)
- [groupArraySorted](../reference/grouparraysorted.md)
- [groupArrayIntersect](../reference/grouparrayintersect.md)
- [groupBitAnd](../reference/groupbitand.md)
- [groupBitOr](../reference/groupbitor.md)
- [groupBitXor](../reference/groupbitxor.md)
- [groupBitmap](../reference/groupbitmap.md)
- [groupBitmapAnd](../reference/groupbitmapand.md)
- [groupBitmapOr](../reference/groupbitmapor.md)
- [groupBitmapXor](../reference/groupbitmapxor.md)
- [sumWithOverflow](../reference/sumwithoverflow.md)
- [sumMap](../reference/summap.md)
- [sumMapWithOverflow](../reference/summapwithoverflow.md)
- [sumMapFiltered](../parametric-functions.md/#summapfiltered)
- [sumMapFilteredWithOverflow](../parametric-functions.md/#summapfilteredwithoverflow)
- [minMap](../reference/minmap.md)
- [maxMap](../reference/maxmap.md)
- [skewSamp](../reference/skewsamp.md)
- [skewPop](../reference/skewpop.md)
- [kurtSamp](../reference/kurtsamp.md)
- [kurtPop](../reference/kurtpop.md)
- [uniq](../reference/uniq.md)
- [uniqExact](../reference/uniqexact.md)
- [uniqCombined](../reference/uniqcombined.md)
- [uniqCombined64](../reference/uniqcombined64.md)
- [uniqHLL12](../reference/uniqhll12.md)
- [uniqTheta](../reference/uniqthetasketch.md)
- [quantile](../reference/quantile.md)
- [quantiles](../reference/quantiles.md)
- [quantileExact](../reference/quantileexact.md)
- [quantileExactLow](../reference/quantileexact.md#quantileexactlow)
- [quantileExactHigh](../reference/quantileexact.md#quantileexacthigh)
- [quantileExactWeighted](../reference/quantileexactweighted.md)
- [quantileTiming](../reference/quantiletiming.md)
- [quantileTimingWeighted](../reference/quantiletimingweighted.md)
- [quantileDeterministic](../reference/quantiledeterministic.md)
- [quantileTDigest](../reference/quantiletdigest.md)
- [quantileTDigestWeighted](../reference/quantiletdigestweighted.md)
- [quantileBFloat16](../reference/quantilebfloat16.md#quantilebfloat16)
- [quantileBFloat16Weighted](../reference/quantilebfloat16.md#quantilebfloat16weighted)
- [quantileDD](../reference/quantileddsketch.md#quantileddsketch)
- [simpleLinearRegression](../reference/simplelinearregression.md)
- [singleValueOrNull](../reference/singlevalueornull.md)
- [stochasticLinearRegression](../reference/stochasticlinearregression.md)
- [stochasticLogisticRegression](../reference/stochasticlogisticregression.md)
- [categoricalInformationValue](../reference/categoricalinformationvalue.md)
- [contingency](../reference/contingency.md)
- [cramersV](../reference/cramersv.md)
- [cramersVBiasCorrected](../reference/cramersvbiascorrected.md)
- [theilsU](../reference/theilsu.md)
- [maxIntersections](../reference/maxintersections.md)
- [maxIntersectionsPosition](../reference/maxintersectionsposition.md)
- [meanZTest](../reference/meanztest.md)
- [quantileGK](../reference/quantileGK.md)
- [quantileInterpolatedWeighted](../reference/quantileinterpolatedweighted.md)
- [sparkBar](../reference/sparkbar.md)
- [sumCount](../reference/sumcount.md)
- [largestTriangleThreeBuckets](../reference/largestTriangleThreeBuckets.md)
ClickHouse supports all standard SQL aggregate functions ([sum](../reference/sum.md), [avg](../reference/avg.md), [min](../reference/min.md), [max](../reference/max.md), [count](../reference/count.md)), as well as a wide range of other aggregate functions.

4
docs/en/sql-reference/data-types/aggregatefunction.md
View File

@ -6,7 +6,9 @@ sidebar_label: AggregateFunction
# AggregateFunction
Aggregate functions can have an implementation-defined intermediate state that can be serialized to an `AggregateFunction(...)` data type and stored in a table, usually, by means of [a materialized view](../../sql-reference/statements/create/view.md). The common way to produce an aggregate function state is by calling the aggregate function with the `-State` suffix. To get the final result of aggregation in the future, you must use the same aggregate function with the `-Merge`suffix.
Aggregate functions have an implementation-defined intermediate state that can be serialized to an `AggregateFunction(...)` data type and stored in a table, usually, by means of [a materialized view](../../sql-reference/statements/create/view.md).
The common way to produce an aggregate function state is by calling the aggregate function with the `-State` suffix.
To get the final result of aggregation in the future, you must use the same aggregate function with the `-Merge`suffix.
`AggregateFunction(name, types_of_arguments...)` — parametric data type.

29
docs/en/sql-reference/data-types/index.md
View File

@ -6,29 +6,8 @@ sidebar_position: 1
# Data Types in ClickHouse
ClickHouse can store various kinds of data in table cells. This section describes the supported data types and special considerations for using and/or implementing them if any.
This section describes the data types supported by ClickHouse, for example [integers](int-uint.md), [floats](float.md) and [strings](string.md).
:::note
You can check whether a data type name is case-sensitive in the [system.data_type_families](../../operations/system-tables/data_type_families.md#system_tables-data_type_families) table.
:::
ClickHouse data types include:
- **Integer types**: [signed and unsigned integers](./int-uint.md) (`UInt8`, `UInt16`, `UInt32`, `UInt64`, `UInt128`, `UInt256`, `Int8`, `Int16`, `Int32`, `Int64`, `Int128`, `Int256`)
- **Floating-point numbers**: [floats](./float.md)(`Float32` and `Float64`) and [`Decimal` values](./decimal.md)
- **Boolean**: ClickHouse has a [`Boolean` type](./boolean.md)
- **Strings**: [`String`](./string.md) and [`FixedString`](./fixedstring.md)
- **Dates**: use [`Date`](./date.md) and [`Date32`](./date32.md) for days, and [`DateTime`](./datetime.md) and [`DateTime64`](./datetime64.md) for instances in time
- **Object**: the [`Object`](./json.md) stores a JSON document in a single column (deprecated)
- **JSON**: the [`JSON` object](./newjson.md) stores a JSON document in a single column
- **UUID**: a performant option for storing [`UUID` values](./uuid.md)
- **Low cardinality types**: use an [`Enum`](./enum.md) when you have a handful of unique values, or use [`LowCardinality`](./lowcardinality.md) when you have up to 10,000 unique values of a column
- **Arrays**: any column can be defined as an [`Array` of values](./array.md)
- **Maps**: use [`Map`](./map.md) for storing key/value pairs
- **Aggregation function types**: use [`SimpleAggregateFunction`](./simpleaggregatefunction.md) and [`AggregateFunction`](./aggregatefunction.md) for storing the intermediate status of aggregate function results
- **Nested data structures**: A [`Nested` data structure](./nested-data-structures/index.md) is like a table inside a cell
- **Tuples**: A [`Tuple` of elements](./tuple.md), each having an individual type.
- **Nullable**: [`Nullable`](./nullable.md) allows you to store a value as `NULL` when a value is "missing" (instead of the column settings its default value for the data type)
- **IP addresses**: use [`IPv4`](./ipv4.md) and [`IPv6`](./ipv6.md) to efficiently store IP addresses
- **Geo types**: for [geographical data](./geo.md), including `Point`, `Ring`, `Polygon` and `MultiPolygon`
- **Special data types**: including [`Expression`](./special-data-types/expression.md), [`Set`](./special-data-types/set.md), [`Nothing`](./special-data-types/nothing.md) and [`Interval`](./special-data-types/interval.md)
System table [system.data_type_families](../../operations/system-tables/data_type_families.md#system_tables-data_type_families) provides an
overview of all available data types.
It also shows whether a data type is an alias to another data type and its name is case-sensitive (e.g. `bool` vs. `BOOL`).

2
docs/en/sql-reference/data-types/json.md
View File

@ -7,7 +7,7 @@ keywords: [object, data type]
# Object Data Type (deprecated)
**This feature is not production-ready and is now deprecated.** If you need to work with JSON documents, consider using [this guide](/docs/en/integrations/data-formats/json/overview) instead. A new implementation to support JSON object is in progress and can be tracked [here](https://github.com/ClickHouse/ClickHouse/issues/54864).
**This feature is not production-ready and deprecated.** If you need to work with JSON documents, consider using [this guide](/docs/en/integrations/data-formats/json/overview) instead. A new implementation to support JSON object is in progress and can be tracked [here](https://github.com/ClickHouse/ClickHouse/issues/54864).
<hr />

4
docs/en/sql-reference/data-types/simpleaggregatefunction.md
View File

@ -5,7 +5,9 @@ sidebar_label: SimpleAggregateFunction
---
# SimpleAggregateFunction
`SimpleAggregateFunction(name, types_of_arguments...)` data type stores current value of the aggregate function, and does not store its full state as [`AggregateFunction`](../../sql-reference/data-types/aggregatefunction.md) does. This optimization can be applied to functions for which the following property holds: the result of applying a function `f` to a row set `S1 UNION ALL S2` can be obtained by applying `f` to parts of the row set separately, and then again applying `f` to the results: `f(S1 UNION ALL S2) = f(f(S1) UNION ALL f(S2))`. This property guarantees that partial aggregation results are enough to compute the combined one, so we do not have to store and process any extra data.
`SimpleAggregateFunction(name, types_of_arguments...)` data type stores current value (intermediate state) of the aggregate function, but not its full state as [`AggregateFunction`](../../sql-reference/data-types/aggregatefunction.md) does.
This optimization can be applied to functions for which the following property holds: the result of applying a function `f` to a row set `S1 UNION ALL S2` can be obtained by applying `f` to parts of the row set separately, and then again applying `f` to the results: `f(S1 UNION ALL S2) = f(f(S1) UNION ALL f(S2))`.
This property guarantees that partial aggregation results are enough to compute the combined one, so we do not have to store and process any extra data.
The common way to produce an aggregate function value is by calling the aggregate function with the [-SimpleState](../../sql-reference/aggregate-functions/combinators.md#agg-functions-combinator-simplestate) suffix.

68
docs/en/sql-reference/functions/geo/index.md
View File

@ -5,70 +5,4 @@ sidebar_position: 62
title: "Geo Functions"
---
## Geographical Coordinates Functions
- [greatCircleDistance](./coordinates.md#greatcircledistance)
- [geoDistance](./coordinates.md#geodistance)
- [greatCircleAngle](./coordinates.md#greatcircleangle)
- [pointInEllipses](./coordinates.md#pointinellipses)
- [pointInPolygon](./coordinates.md#pointinpolygon)
## Geohash Functions
- [geohashEncode](./geohash.md#geohashencode)
- [geohashDecode](./geohash.md#geohashdecode)
- [geohashesInBox](./geohash.md#geohashesinbox)
## H3 Indexes Functions
- [h3IsValid](./h3.md#h3isvalid)
- [h3GetResolution](./h3.md#h3getresolution)
- [h3EdgeAngle](./h3.md#h3edgeangle)
- [h3EdgeLengthM](./h3.md#h3edgelengthm)
- [h3EdgeLengthKm](./h3.md#h3edgelengthkm)
- [geoToH3](./h3.md#geotoh3)
- [h3ToGeo](./h3.md#h3togeo)
- [h3ToGeoBoundary](./h3.md#h3togeoboundary)
- [h3kRing](./h3.md#h3kring)
- [h3GetBaseCell](./h3.md#h3getbasecell)
- [h3HexAreaM2](./h3.md#h3hexaream2)
- [h3HexAreaKm2](./h3.md#h3hexareakm2)
- [h3IndexesAreNeighbors](./h3.md#h3indexesareneighbors)
- [h3ToChildren](./h3.md#h3tochildren)
- [h3ToParent](./h3.md#h3toparent)
- [h3ToString](./h3.md#h3tostring)
- [stringToH3](./h3.md#stringtoh3)
- [h3GetResolution](./h3.md#h3getresolution)
- [h3IsResClassIII](./h3.md#h3isresclassiii)
- [h3IsPentagon](./h3.md#h3ispentagon)
- [h3GetFaces](./h3.md#h3getfaces)
- [h3CellAreaM2](./h3.md#h3cellaream2)
- [h3CellAreaRads2](./h3.md#h3cellarearads2)
- [h3ToCenterChild](./h3.md#h3tocenterchild)
- [h3ExactEdgeLengthM](./h3.md#h3exactedgelengthm)
- [h3ExactEdgeLengthKm](./h3.md#h3exactedgelengthkm)
- [h3ExactEdgeLengthRads](./h3.md#h3exactedgelengthrads)
- [h3NumHexagons](./h3.md#h3numhexagons)
- [h3Line](./h3.md#h3line)
- [h3Distance](./h3.md#h3distance)
- [h3HexRing](./h3.md#h3hexring)
- [h3GetUnidirectionalEdge](./h3.md#h3getunidirectionaledge)
- [h3UnidirectionalEdgeIsValid](./h3.md#h3unidirectionaledgeisvalid)
- [h3GetOriginIndexFromUnidirectionalEdge](./h3.md#h3getoriginindexfromunidirectionaledge)
- [h3GetDestinationIndexFromUnidirectionalEdge](./h3.md#h3getdestinationindexfromunidirectionaledge)
- [h3GetIndexesFromUnidirectionalEdge](./h3.md#h3getindexesfromunidirectionaledge)
- [h3GetUnidirectionalEdgesFromHexagon](./h3.md#h3getunidirectionaledgesfromhexagon)
- [h3GetUnidirectionalEdgeBoundary](./h3.md#h3getunidirectionaledgeboundary)
## S2 Index Functions
- [geoToS2](./s2.md#geotos2)
- [s2ToGeo](./s2.md#s2togeo)
- [s2GetNeighbors](./s2.md#s2getneighbors)
- [s2CellsIntersect](./s2.md#s2cellsintersect)
- [s2CapContains](./s2.md#s2capcontains)
- [s2CapUnion](./s2.md#s2capunion)
- [s2RectAdd](./s2.md#s2rectadd)
- [s2RectContains](./s2.md#s2rectcontains)
- [s2RectUnion](./s2.md#s2rectunion)
- [s2RectIntersection](./s2.md#s2rectintersection)
Functions for working with geometric objects, for example [to calculate distances between points on a sphere](./coordinates.md), [compute geohashes](./geohash.md), and work with [h3 indexes](./h3.md).

2
docs/en/sql-reference/functions/index.md
View File

@ -24,7 +24,7 @@ All expressions in a query that have the same AST (the same record or same resul
## Types of Results
All functions return a single return as the result (not several values, and not zero values). The type of result is usually defined only by the types of arguments, not by the values. Exceptions are the tupleElement function (the a.N operator), and the toFixedString function.
All functions return a single value as the result (not several values, and not zero values). The type of result is usually defined only by the types of arguments, not by the values. Exceptions are the tupleElement function (the a.N operator), and the toFixedString function.
## Constants

2
docs/en/sql-reference/statements/alter/column.md
View File

@ -279,7 +279,7 @@ For columns with a new or updated `MATERIALIZED` value expression, all existing
For columns with a new or updated `DEFAULT` value expression, the behavior depends on the ClickHouse version:
- In ClickHouse < v24.2, all existing rows are rewritten.
- ClickHouse >= v24.2 distinguishes if a row value in a column with `DEFAULT` value expression was explicitly specified when it was inserted, or not, i.e. calculated from the `DEFAULT` value expression. If the value was explicitly specified, ClickHouse keeps it as is. If the value was was calculated, ClickHouse changes it to the new or updated `MATERIALIZED` value expression.
- ClickHouse >= v24.2 distinguishes if a row value in a column with `DEFAULT` value expression was explicitly specified when it was inserted, or not, i.e. calculated from the `DEFAULT` value expression. If the value was explicitly specified, ClickHouse keeps it as is. If the value was calculated, ClickHouse changes it to the new or updated `MATERIALIZED` value expression.
Syntax:

14
docs/en/sql-reference/statements/create/index.md
View File

@ -6,16 +6,4 @@ sidebar_label: CREATE
# CREATE Queries
Create queries make a new entity of one of the following kinds:
- [DATABASE](/docs/en/sql-reference/statements/create/database.md)
- [TABLE](/docs/en/sql-reference/statements/create/table.md)
- [VIEW](/docs/en/sql-reference/statements/create/view.md)
- [DICTIONARY](/docs/en/sql-reference/statements/create/dictionary.md)
- [FUNCTION](/docs/en/sql-reference/statements/create/function.md)
- [USER](/docs/en/sql-reference/statements/create/user.md)
- [ROLE](/docs/en/sql-reference/statements/create/role.md)
- [ROW POLICY](/docs/en/sql-reference/statements/create/row-policy.md)
- [QUOTA](/docs/en/sql-reference/statements/create/quota.md)
- [SETTINGS PROFILE](/docs/en/sql-reference/statements/create/settings-profile.md)
- [NAMED COLLECTION](/docs/en/sql-reference/statements/create/named-collection.md)
CREATE queries create (for example) new [databases](/docs/en/sql-reference/statements/create/database.md), [tables](/docs/en/sql-reference/statements/create/table.md) and [views](/docs/en/sql-reference/statements/create/view.md).

25
docs/en/sql-reference/statements/index.md
View File

@ -6,27 +6,4 @@ sidebar_label: List of statements
# ClickHouse SQL Statements
Statements represent various kinds of action you can perform using SQL queries. Each kind of statement has it’s own syntax and usage details that are described separately:
- [SELECT](/docs/en/sql-reference/statements/select/index.md)
- [INSERT INTO](/docs/en/sql-reference/statements/insert-into.md)
- [CREATE](/docs/en/sql-reference/statements/create/index.md)
- [ALTER](/docs/en/sql-reference/statements/alter/index.md)
- [SYSTEM](/docs/en/sql-reference/statements/system.md)
- [SHOW](/docs/en/sql-reference/statements/show.md)
- [GRANT](/docs/en/sql-reference/statements/grant.md)
- [REVOKE](/docs/en/sql-reference/statements/revoke.md)
- [ATTACH](/docs/en/sql-reference/statements/attach.md)
- [CHECK TABLE](/docs/en/sql-reference/statements/check-table.md)
- [DESCRIBE TABLE](/docs/en/sql-reference/statements/describe-table.md)
- [DETACH](/docs/en/sql-reference/statements/detach.md)
- [DROP](/docs/en/sql-reference/statements/drop.md)
- [EXISTS](/docs/en/sql-reference/statements/exists.md)
- [KILL](/docs/en/sql-reference/statements/kill.md)
- [OPTIMIZE](/docs/en/sql-reference/statements/optimize.md)
- [RENAME](/docs/en/sql-reference/statements/rename.md)
- [SET](/docs/en/sql-reference/statements/set.md)
- [SET ROLE](/docs/en/sql-reference/statements/set-role.md)
- [TRUNCATE](/docs/en/sql-reference/statements/truncate.md)
- [USE](/docs/en/sql-reference/statements/use.md)
- [EXPLAIN](/docs/en/sql-reference/statements/explain.md)
Users interact with ClickHouse using SQL statements. ClickHouse supports common SQL statements like [SELECT](select/index.md) and [CREATE](create/index.md), but it also provides specialized statements like [KILL](kill.md) and [OPTIMIZE](optimize.md).

10
programs/client/Client.cpp
View File

@ -431,7 +431,7 @@ catch (const Exception & e)
bool need_print_stack_trace = config().getBool("stacktrace", false) && e.code() != ErrorCodes::NETWORK_ERROR;
std::cerr << getExceptionMessage(e, need_print_stack_trace, true) << std::endl << std::endl;
/// If exception code isn't zero, we should return non-zero return code anyway.
return e.code() ? e.code() : -1;
return static_cast<UInt8>(e.code()) ? e.code() : -1;
}
catch (...)
{
@ -1390,7 +1390,8 @@ int mainEntryClickHouseClient(int argc, char ** argv)
catch (const DB::Exception & e)
{
std::cerr << DB::getExceptionMessage(e, false) << std::endl;
return 1;
auto code = DB::getCurrentExceptionCode();
return static_cast<UInt8>(code) ? code : 1;
}
catch (const boost::program_options::error & e)
{
@ -1399,7 +1400,8 @@ int mainEntryClickHouseClient(int argc, char ** argv)
}
catch (...)
{
std::cerr << DB::getCurrentExceptionMessage(true) << std::endl;
return 1;
std::cerr << DB::getCurrentExceptionMessage(true) << '\n';
auto code = DB::getCurrentExceptionCode();
return static_cast<UInt8>(code) ? code : 1;
}
}

40
programs/compressor/Compressor.cpp
View File

@ -9,6 +9,7 @@
#include <IO/WriteBufferFromFile.h>
#include <IO/ReadBufferFromFile.h>
#include <Compression/CompressedWriteBuffer.h>
#include <Compression/ParallelCompressedWriteBuffer.h>
#include <Compression/CompressedReadBuffer.h>
#include <Compression/CompressedReadBufferFromFile.h>
#include <IO/WriteHelpers.h>
@ -17,6 +18,8 @@
#include <Parsers/ExpressionElementParsers.h>
#include <Compression/CompressionFactory.h>
#include <Common/TerminalSize.h>
#include <Common/ThreadPool.h>
#include <Common/CurrentMetrics.h>
#include <Core/Defines.h>
@ -29,6 +32,13 @@ namespace DB
}
}
namespace CurrentMetrics
{
extern const Metric LocalThread;
extern const Metric LocalThreadActive;
extern const Metric LocalThreadScheduled;
}
namespace
{
@ -77,11 +87,12 @@ int mainEntryClickHouseCompressor(int argc, char ** argv)
("decompress,d", "decompress")
("offset-in-compressed-file", po::value<size_t>()->default_value(0ULL), "offset to the compressed block (i.e. physical file offset)")
("offset-in-decompressed-block", po::value<size_t>()->default_value(0ULL), "offset to the decompressed block (i.e. virtual offset)")
("block-size,b", po::value<unsigned>()->default_value(DBMS_DEFAULT_BUFFER_SIZE), "compress in blocks of specified size")
("block-size,b", po::value<size_t>()->default_value(DBMS_DEFAULT_BUFFER_SIZE), "compress in blocks of specified size")
("hc", "use LZ4HC instead of LZ4")
("zstd", "use ZSTD instead of LZ4")
("codec", po::value<std::vector<std::string>>()->multitoken(), "use codecs combination instead of LZ4")
("level", po::value<int>(), "compression level for codecs specified via flags")
("threads", po::value<size_t>()->default_value(1), "number of threads for parallel compression")
("none", "use no compression instead of LZ4")
("stat", "print block statistics of compressed data")
("stacktrace", "print stacktrace of exception")
@ -109,7 +120,8 @@ int mainEntryClickHouseCompressor(int argc, char ** argv)
bool stat_mode = options.count("stat");
bool use_none = options.count("none");
print_stacktrace = options.count("stacktrace");
unsigned block_size = options["block-size"].as<unsigned>();
size_t block_size = options["block-size"].as<size_t>();
size_t num_threads = options["threads"].as<size_t>();
std::vector<std::string> codecs;
if (options.count("codec"))
codecs = options["codec"].as<std::vector<std::string>>();
@ -117,6 +129,12 @@ int mainEntryClickHouseCompressor(int argc, char ** argv)
if ((use_lz4hc || use_zstd || use_none) && !codecs.empty())
throw Exception(ErrorCodes::BAD_ARGUMENTS, "Wrong options, codec flags like --zstd and --codec options are mutually exclusive");
if (num_threads < 1)
throw Exception(ErrorCodes::BAD_ARGUMENTS, "Invalid value of `threads` parameter");
if (num_threads > 1 && decompress)
throw Exception(ErrorCodes::BAD_ARGUMENTS, "Parallel mode is only implemented for compression (not for decompression)");
if (!codecs.empty() && options.count("level"))
throw Exception(ErrorCodes::BAD_ARGUMENTS, "Wrong options, --level is not compatible with --codec list");
@ -145,7 +163,6 @@ int mainEntryClickHouseCompressor(int argc, char ** argv)
else
codec = CompressionCodecFactory::instance().get(method_family, level);
std::unique_ptr<ReadBufferFromFileBase> rb;
std::unique_ptr<WriteBufferFromFileBase> wb;
@ -186,9 +203,20 @@ int mainEntryClickHouseCompressor(int argc, char ** argv)
else
{
/// Compression
CompressedWriteBuffer to(*wb, codec, block_size);
copyData(*rb, to);
to.finalize();
if (num_threads == 1)
{
CompressedWriteBuffer to(*wb, codec, block_size);
copyData(*rb, to);
to.finalize();
}
else
{
ThreadPool pool(CurrentMetrics::LocalThread, CurrentMetrics::LocalThreadActive, CurrentMetrics::LocalThreadScheduled, num_threads);
ParallelCompressedWriteBuffer to(*wb, codec, block_size, num_threads, pool);
copyData(*rb, to);
to.finalize();
}
}
}
catch (...)

8
programs/disks/DisksApp.cpp
View File

@ -546,16 +546,18 @@ int mainEntryClickHouseDisks(int argc, char ** argv)
catch (const DB::Exception & e)
{
std::cerr << DB::getExceptionMessage(e, false) << std::endl;
return 0;
auto code = DB::getCurrentExceptionCode();
return static_cast<UInt8>(code) ? code : 1;
}
catch (const boost::program_options::error & e)
{
std::cerr << "Bad arguments: " << e.what() << std::endl;
return 0;
return DB::ErrorCodes::BAD_ARGUMENTS;
}
catch (...)
{
std::cerr << DB::getCurrentExceptionMessage(true) << std::endl;
return 0;
auto code = DB::getCurrentExceptionCode();
return static_cast<UInt8>(code) ? code : 1;
}
}

6
programs/keeper-client/KeeperClient.cpp
View File

@ -448,7 +448,8 @@ int mainEntryClickHouseKeeperClient(int argc, char ** argv)
catch (const DB::Exception & e)
{
std::cerr << DB::getExceptionMessage(e, false) << std::endl;
return 1;
auto code = DB::getCurrentExceptionCode();
return static_cast<UInt8>(code) ? code : 1;
}
catch (const boost::program_options::error & e)
{
@ -458,6 +459,7 @@ int mainEntryClickHouseKeeperClient(int argc, char ** argv)
catch (...)
{
std::cerr << DB::getCurrentExceptionMessage(true) << std::endl;
return 1;
auto code = DB::getCurrentExceptionCode();
return static_cast<UInt8>(code) ? code : 1;
}
}

4
programs/keeper/Keeper.cpp
View File

@ -81,7 +81,7 @@ int mainEntryClickHouseKeeper(int argc, char ** argv)
{
std::cerr << DB::getCurrentExceptionMessage(true) << "\n";
auto code = DB::getCurrentExceptionCode();
return code ? code : 1;
return static_cast<UInt8>(code) ? code : 1;
}
}
@ -672,7 +672,7 @@ catch (...)
/// Poco does not provide stacktrace.
tryLogCurrentException("Application");
auto code = getCurrentExceptionCode();
return code ? code : -1;
return static_cast<UInt8>(code) ? code : -1;
}

2
programs/library-bridge/LibraryBridge.cpp
View File

@ -13,7 +13,7 @@ int mainEntryClickHouseLibraryBridge(int argc, char ** argv)
{
std::cerr << DB::getCurrentExceptionMessage(true) << "\n";
auto code = DB::getCurrentExceptionCode();
return code ? code : 1;
return static_cast<UInt8>(code) ? code : 1;
}
}

36
programs/local/LocalServer.cpp
View File

@ -14,6 +14,7 @@
#include <Databases/registerDatabases.h>
#include <Databases/DatabaseFilesystem.h>
#include <Databases/DatabaseMemory.h>
#include <Databases/DatabaseAtomic.h>
#include <Databases/DatabasesOverlay.h>
#include <Storages/System/attachSystemTables.h>
#include <Storages/System/attachInformationSchemaTables.h>
@ -22,7 +23,6 @@
#include <Interpreters/ProcessList.h>
#include <Interpreters/loadMetadata.h>
#include <Interpreters/registerInterpreters.h>
#include <base/getFQDNOrHostName.h>
#include <Access/AccessControl.h>
#include <Common/PoolId.h>
#include <Common/Exception.h>
@ -31,7 +31,6 @@
#include <Common/ThreadStatus.h>
#include <Common/TLDListsHolder.h>
#include <Common/quoteString.h>
#include <Common/randomSeed.h>
#include <Common/ThreadPool.h>
#include <Common/CurrentMetrics.h>
#include <Loggers/OwnFormattingChannel.h>
@ -50,7 +49,6 @@
#include <Dictionaries/registerDictionaries.h>
#include <Disks/registerDisks.h>
#include <Formats/registerFormats.h>
#include <boost/algorithm/string/replace.hpp>
#include <boost/program_options/options_description.hpp>
#include <base/argsToConfig.h>
#include <filesystem>
@ -71,9 +69,11 @@ namespace CurrentMetrics
namespace DB
{
namespace Setting
{
extern const SettingsBool allow_introspection_functions;
extern const SettingsBool implicit_select;
extern const SettingsLocalFSReadMethod storage_file_read_method;
}
@ -126,6 +126,7 @@ void applySettingsOverridesForLocal(ContextMutablePtr context)
settings[Setting::allow_introspection_functions] = true;
settings[Setting::storage_file_read_method] = LocalFSReadMethod::mmap;
settings[Setting::implicit_select] = true;
context->setSettings(settings);
}
@ -257,12 +258,12 @@ static DatabasePtr createMemoryDatabaseIfNotExists(ContextPtr context, const Str
return system_database;
}
static DatabasePtr createClickHouseLocalDatabaseOverlay(const String & name_, ContextPtr context_)
static DatabasePtr createClickHouseLocalDatabaseOverlay(const String & name_, ContextPtr context)
{
auto databaseCombiner = std::make_shared<DatabasesOverlay>(name_, context_);
databaseCombiner->registerNextDatabase(std::make_shared<DatabaseFilesystem>(name_, "", context_));
databaseCombiner->registerNextDatabase(std::make_shared<DatabaseMemory>(name_, context_));
return databaseCombiner;
auto overlay = std::make_shared<DatabasesOverlay>(name_, context);
overlay->registerNextDatabase(std::make_shared<DatabaseAtomic>(name_, fs::weakly_canonical(context->getPath()), UUIDHelpers::generateV4(), context));
overlay->registerNextDatabase(std::make_shared<DatabaseFilesystem>(name_, "", context));
return overlay;
}
/// If path is specified and not empty, will try to setup server environment and load existing metadata
@ -615,12 +616,14 @@ catch (const DB::Exception & e)
{
bool need_print_stack_trace = getClientConfiguration().getBool("stacktrace", false);
std::cerr << getExceptionMessage(e, need_print_stack_trace, true) << std::endl;
return e.code() ? e.code() : -1;
auto code = DB::getCurrentExceptionCode();
return static_cast<UInt8>(code) ? code : 1;
}
catch (...)
{
std::cerr << getCurrentExceptionMessage(false) << std::endl;
return getCurrentExceptionCode();
std::cerr << DB::getCurrentExceptionMessage(true) << '\n';
auto code = DB::getCurrentExceptionCode();
return static_cast<UInt8>(code) ? code : 1;
}
void LocalServer::updateLoggerLevel(const String & logs_level)
@ -809,7 +812,12 @@ void LocalServer::processConfig()
DatabaseCatalog::instance().initializeAndLoadTemporaryDatabase();
std::string default_database = server_settings[ServerSetting::default_database];
DatabaseCatalog::instance().attachDatabase(default_database, createClickHouseLocalDatabaseOverlay(default_database, global_context));
{
DatabasePtr database = createClickHouseLocalDatabaseOverlay(default_database, global_context);
if (UUID uuid = database->getUUID(); uuid != UUIDHelpers::Nil)
DatabaseCatalog::instance().addUUIDMapping(uuid);
DatabaseCatalog::instance().attachDatabase(default_database, database);
}
global_context->setCurrentDatabase(default_database);
if (getClientConfiguration().has("path"))
@ -1029,7 +1037,7 @@ int mainEntryClickHouseLocal(int argc, char ** argv)
{
std::cerr << DB::getExceptionMessage(e, false) << std::endl;
auto code = DB::getCurrentExceptionCode();
return code ? code : 1;
return static_cast<UInt8>(code) ? code : 1;
}
catch (const boost::program_options::error & e)
{
@ -1040,6 +1048,6 @@ int mainEntryClickHouseLocal(int argc, char ** argv)
{
std::cerr << DB::getCurrentExceptionMessage(true) << '\n';
auto code = DB::getCurrentExceptionCode();
return code ? code : 1;
return static_cast<UInt8>(code) ? code : 1;
}
}

40
programs/main.cpp
View File

@ -1,27 +1,22 @@
#include <unistd.h>
#include <fcntl.h>
#include <base/phdr_cache.h>
#include <Common/EnvironmentChecks.h>
#include <Common/StringUtils.h>
#include <Common/getHashOfLoadedBinary.h>
#include <new>
#include <iostream>
#include <vector>
#include <string>
#include <string_view>
#include <utility> /// pair
#include <fmt/format.h>
#if defined(SANITIZE_COVERAGE)
# include <Common/Coverage.h>
#endif
#include "config.h"
#include "config_tools.h"
#include <Common/EnvironmentChecks.h>
#include <Common/Coverage.h>
#include <Common/StringUtils.h>
#include <Common/getHashOfLoadedBinary.h>
#include <Common/IO.h>
#include <base/phdr_cache.h>
#include <base/coverage.h>
#include <filesystem>
#include <iostream>
#include <new>
#include <string>
#include <string_view>
#include <utility> /// pair
#include <vector>
/// Universal executable for various clickhouse applications
int mainEntryClickHouseServer(int argc, char ** argv);
@ -238,9 +233,12 @@ int main(int argc_, char ** argv_)
/// clickhouse # spawn local
/// clickhouse local # spawn local
/// clickhouse "select ..." # spawn local
/// clickhouse /tmp/repro --enable-analyzer
///
if (main_func == printHelp && !argv.empty() && (argv.size() == 1 || argv[1][0] == '-'
|| std::string_view(argv[1]).contains(' ')))
std::error_code ec;
if (main_func == printHelp && !argv.empty()
&& (argv.size() == 1 || argv[1][0] == '-' || std::string_view(argv[1]).contains(' ')
|| std::filesystem::is_regular_file(std::filesystem::path{argv[1]}, ec)))
{
main_func = mainEntryClickHouseLocal;
}

2
programs/obfuscator/Obfuscator.cpp
View File

@ -1480,5 +1480,5 @@ catch (...)
{
std::cerr << DB::getCurrentExceptionMessage(true) << "\n";
auto code = DB::getCurrentExceptionCode();
return code ? code : 1;
return static_cast<UInt8>(code) ? code : 1;
}

2
programs/odbc-bridge/ODBCBridge.cpp
View File

@ -13,7 +13,7 @@ int mainEntryClickHouseODBCBridge(int argc, char ** argv)
{
std::cerr << DB::getCurrentExceptionMessage(true) << "\n";
auto code = DB::getCurrentExceptionCode();
return code ? code : 1;
return static_cast<UInt8>(code) ? code : 1;
}
}

4
programs/server/Server.cpp
View File

@ -343,7 +343,7 @@ int mainEntryClickHouseServer(int argc, char ** argv)
{
std::cerr << DB::getCurrentExceptionMessage(true) << "\n";
auto code = DB::getCurrentExceptionCode();
return code ? code : 1;
return static_cast<UInt8>(code) ? code : 1;
}
}
@ -2537,7 +2537,7 @@ catch (...)
/// Poco does not provide stacktrace.
tryLogCurrentException("Application");
auto code = getCurrentExceptionCode();
return code ? code : -1;
return static_cast<UInt8>(code) ? code : -1;
}
std::unique_ptr<TCPProtocolStackFactory> Server::buildProtocolStackFromConfig(

10
src/Analyzer/JoinNode.cpp
View File

@ -48,9 +48,15 @@ ASTPtr JoinNode::toASTTableJoin() const
auto join_expression_ast = children[join_expression_child_index]->toAST();
if (is_using_join_expression)
join_ast->using_expression_list = std::move(join_expression_ast);
{
join_ast->using_expression_list = join_expression_ast;
join_ast->children.push_back(join_ast->using_expression_list);
}
else
join_ast->on_expression = std::move(join_expression_ast);
{
join_ast->on_expression = join_expression_ast;
join_ast->children.push_back(join_ast->on_expression);
}
}
return join_ast;

59
src/Analyzer/QueryTreePassManager.cpp
View File

@ -3,7 +3,6 @@
#include <memory>
#include <Common/Exception.h>
#include "Analyzer/Passes/OptimizeGroupByInjectiveFunctionsPass.h"
#include <IO/WriteHelpers.h>
#include <IO/Operators.h>
@ -16,39 +15,39 @@
#include <Analyzer/ColumnNode.h>
#include <Analyzer/FunctionNode.h>
#include <Analyzer/InDepthQueryTreeVisitor.h>
#include <Analyzer/Utils.h>
#include <Analyzer/Passes/AggregateFunctionOfGroupByKeysPass.h>
#include <Analyzer/Passes/AggregateFunctionsArithmericOperationsPass.h>
#include <Analyzer/Passes/ArrayExistsToHasPass.h>
#include <Analyzer/Passes/AutoFinalOnQueryPass.h>
#include <Analyzer/Passes/ComparisonTupleEliminationPass.h>
#include <Analyzer/Passes/ConvertOrLikeChainPass.h>
#include <Analyzer/Passes/ConvertQueryToCNFPass.h>
#include <Analyzer/Passes/CountDistinctPass.h>
#include <Analyzer/Passes/CrossToInnerJoinPass.h>
#include <Analyzer/Passes/FunctionToSubcolumnsPass.h>
#include <Analyzer/Passes/FuseFunctionsPass.h>
#include <Analyzer/Passes/GroupingFunctionsResolvePass.h>
#include <Analyzer/Passes/IfChainToMultiIfPass.h>
#include <Analyzer/Passes/IfConstantConditionPass.h>
#include <Analyzer/Passes/IfTransformStringsToEnumPass.h>
#include <Analyzer/Passes/LogicalExpressionOptimizerPass.h>
#include <Analyzer/Passes/MultiIfToIfPass.h>
#include <Analyzer/Passes/NormalizeCountVariantsPass.h>
#include <Analyzer/Passes/OptimizeDateOrDateTimeConverterWithPreimagePass.h>
#include <Analyzer/Passes/OptimizeGroupByFunctionKeysPass.h>
#include <Analyzer/Passes/OptimizeGroupByInjectiveFunctionsPass.h>
#include <Analyzer/Passes/OptimizeRedundantFunctionsInOrderByPass.h>
#include <Analyzer/Passes/OrderByLimitByDuplicateEliminationPass.h>
#include <Analyzer/Passes/OrderByTupleEliminationPass.h>
#include <Analyzer/Passes/QueryAnalysisPass.h>
#include <Analyzer/Passes/RemoveUnusedProjectionColumnsPass.h>
#include <Analyzer/Passes/RewriteSumFunctionWithSumAndCountPass.h>
#include <Analyzer/Passes/CountDistinctPass.h>
#include <Analyzer/Passes/UniqToCountPass.h>
#include <Analyzer/Passes/FunctionToSubcolumnsPass.h>
#include <Analyzer/Passes/RewriteAggregateFunctionWithIfPass.h>
#include <Analyzer/Passes/SumIfToCountIfPass.h>
#include <Analyzer/Passes/MultiIfToIfPass.h>
#include <Analyzer/Passes/IfConstantConditionPass.h>
#include <Analyzer/Passes/IfChainToMultiIfPass.h>
#include <Analyzer/Passes/OrderByTupleEliminationPass.h>
#include <Analyzer/Passes/NormalizeCountVariantsPass.h>
#include <Analyzer/Passes/AggregateFunctionsArithmericOperationsPass.h>
#include <Analyzer/Passes/UniqInjectiveFunctionsEliminationPass.h>
#include <Analyzer/Passes/OrderByLimitByDuplicateEliminationPass.h>
#include <Analyzer/Passes/FuseFunctionsPass.h>
#include <Analyzer/Passes/OptimizeGroupByFunctionKeysPass.h>
#include <Analyzer/Passes/IfTransformStringsToEnumPass.h>
#include <Analyzer/Passes/ConvertOrLikeChainPass.h>
#include <Analyzer/Passes/OptimizeRedundantFunctionsInOrderByPass.h>
#include <Analyzer/Passes/GroupingFunctionsResolvePass.h>
#include <Analyzer/Passes/AutoFinalOnQueryPass.h>
#include <Analyzer/Passes/ArrayExistsToHasPass.h>
#include <Analyzer/Passes/ComparisonTupleEliminationPass.h>
#include <Analyzer/Passes/LogicalExpressionOptimizerPass.h>
#include <Analyzer/Passes/CrossToInnerJoinPass.h>
#include <Analyzer/Passes/RewriteSumFunctionWithSumAndCountPass.h>
#include <Analyzer/Passes/ShardNumColumnToFunctionPass.h>
#include <Analyzer/Passes/ConvertQueryToCNFPass.h>
#include <Analyzer/Passes/AggregateFunctionOfGroupByKeysPass.h>
#include <Analyzer/Passes/OptimizeDateOrDateTimeConverterWithPreimagePass.h>
#include <Analyzer/Passes/SumIfToCountIfPass.h>
#include <Analyzer/Passes/UniqInjectiveFunctionsEliminationPass.h>
#include <Analyzer/Passes/UniqToCountPass.h>
#include <Analyzer/Utils.h>
namespace DB
{

19
src/Analyzer/Resolve/QueryAnalyzer.cpp
View File

@ -1,3 +1,4 @@
#include <Interpreters/ProcessorsProfileLog.h>
#include <Common/FieldVisitorToString.h>
#include <DataTypes/DataTypesNumber.h>
@ -51,7 +52,6 @@
#include <Analyzer/ArrayJoinNode.h>
#include <Analyzer/JoinNode.h>
#include <Analyzer/UnionNode.h>
#include <Analyzer/InDepthQueryTreeVisitor.h>
#include <Analyzer/QueryTreeBuilder.h>
#include <Analyzer/IQueryTreeNode.h>
#include <Analyzer/Identifier.h>
@ -677,6 +677,8 @@ void QueryAnalyzer::evaluateScalarSubqueryIfNeeded(QueryTreeNodePtr & node, Iden
"tuple"});
}
}
logProcessorProfile(context, io.pipeline.getProcessors());
}
scalars_cache.emplace(node_with_hash, scalar_block);
@ -3023,9 +3025,10 @@ ProjectionNames QueryAnalyzer::resolveFunction(QueryTreeNodePtr & node, Identifi
argument_column.name = arguments_projection_names[function_argument_index];
/** If function argument is lambda, save lambda argument index and initialize argument type as DataTypeFunction
* where function argument types are initialized with empty array of lambda arguments size.
* where function argument types are initialized with empty arrays of lambda arguments size.
*/
if (const auto * lambda_node = function_argument->as<const LambdaNode>())
const auto * lambda_node = function_argument->as<const LambdaNode>();
if (lambda_node)
{
size_t lambda_arguments_size = lambda_node->getArguments().getNodes().size();
argument_column.type = std::make_shared<DataTypeFunction>(DataTypes(lambda_arguments_size, nullptr), nullptr);
@ -3497,15 +3500,11 @@ ProjectionNames QueryAnalyzer::resolveFunction(QueryTreeNodePtr & node, Identifi
else
function_base = function->build(argument_columns);
/// Do not constant fold get scalar functions
// bool disable_constant_folding = function_name == "__getScalar" || function_name == "shardNum" ||
// function_name == "shardCount" || function_name == "hostName" || function_name == "tcpPort";
/** If function is suitable for constant folding try to convert it to constant.
* Example: SELECT plus(1, 1);
* Result: SELECT 2;
*/
if (function_base->isSuitableForConstantFolding()) // && !disable_constant_folding)
if (function_base->isSuitableForConstantFolding())
{
auto result_type = function_base->getResultType();
auto executable_function = function_base->prepare(argument_columns);
@ -3514,7 +3513,9 @@ ProjectionNames QueryAnalyzer::resolveFunction(QueryTreeNodePtr & node, Identifi
if (all_arguments_constants)
{
size_t num_rows = function_arguments.empty() ? 0 : argument_columns.front().column->size();
size_t num_rows = 0;
if (!argument_columns.empty())
num_rows = argument_columns.front().column->size();
column = executable_function->execute(argument_columns, result_type, num_rows, true);
}
else

10
src/Client/ClientBase.cpp
View File

@ -1650,6 +1650,11 @@ void ClientBase::sendData(Block & sample, const ColumnsDescription & columns_des
if (!parsed_insert_query)
return;
/// If it's clickhouse-local, and the input data reading is already baked into the query pipeline,
/// don't read the data again here. This happens in some cases (e.g. input() table function) but not others (e.g. INFILE).
if (!connection->isSendDataNeeded())
return;
bool have_data_in_stdin = !is_interactive && !stdin_is_a_tty && isStdinNotEmptyAndValid(std_in);
if (need_render_progress)
@ -2674,7 +2679,10 @@ void ClientBase::runInteractive()
#if USE_REPLXX
replxx::Replxx::highlighter_callback_t highlight_callback{};
if (getClientConfiguration().getBool("highlight", true))
highlight_callback = highlight;
highlight_callback = [this](const String & query, std::vector<replxx::Replxx::Color> & colors)
{
highlight(query, colors, *client_context);
};
ReplxxLineReader lr(
*suggest,

29
src/Client/ClientBaseHelpers.cpp
View File

@ -5,6 +5,8 @@
#include <Parsers/ParserQuery.h>
#include <Parsers/parseQuery.h>
#include <Common/UTF8Helpers.h>
#include <Core/Settings.h>
#include <Interpreters/Context.h>
#include <iostream>
@ -12,6 +14,11 @@
namespace DB
{
namespace Setting
{
extern const SettingsBool implicit_select;
}
/// Should we celebrate a bit?
bool isNewYearMode()
{
@ -95,7 +102,7 @@ bool isChineseNewYearMode(const String & local_tz)
}
#if USE_REPLXX
void highlight(const String & query, std::vector<replxx::Replxx::Color> & colors)
void highlight(const String & query, std::vector<replxx::Replxx::Color> & colors, const Context & context)
{
using namespace replxx;
@ -135,13 +142,27 @@ void highlight(const String & query, std::vector<replxx::Replxx::Color> & colors
/// Currently we highlight only the first query in the multi-query mode.
ParserQuery parser(end);
ParserQuery parser(end, false, context.getSettingsRef()[Setting::implicit_select]);
ASTPtr ast;
bool parse_res = false;
try
{
parse_res = parser.parse(token_iterator, ast, expected);
while (!token_iterator->isEnd())
{
parse_res = parser.parse(token_iterator, ast, expected);
if (!parse_res)
break;
if (!token_iterator->isEnd() && token_iterator->type != TokenType::Semicolon)
{
parse_res = false;
break;
}
while (token_iterator->type == TokenType::Semicolon)
++token_iterator;
}
}
catch (...)
{
@ -175,7 +196,7 @@ void highlight(const String & query, std::vector<replxx::Replxx::Color> & colors
/// Highlight the last error in red. If the parser failed or the lexer found an invalid token,
/// or if it didn't parse all the data (except, the data for INSERT query, which is legitimately unparsed)
if ((!parse_res || last_token.isError() || (!token_iterator->isEnd() && token_iterator->type != TokenType::Semicolon))
if ((!parse_res || last_token.isError())
&& !(insert_data && expected.max_parsed_pos >= insert_data)
&& expected.max_parsed_pos >= prev)
{

4
src/Client/ClientBaseHelpers.h
View File

@ -11,13 +11,15 @@
namespace DB
{
class Context;
/// Should we celebrate a bit?
bool isNewYearMode();
bool isChineseNewYearMode(const String & local_tz);
#if USE_REPLXX
void highlight(const String & query, std::vector<replxx::Replxx::Color> & colors);
void highlight(const String & query, std::vector<replxx::Replxx::Color> & colors, const Context & context);
#endif
}

6
src/Client/ClientBaseOptimizedParts.cpp
View File

@ -1,5 +1,7 @@
#include <Client/ClientApplicationBase.h>
#include <filesystem>
namespace DB
{
@ -108,6 +110,7 @@ void ClientApplicationBase::parseAndCheckOptions(OptionsDescription & options_de
{
/// Two special cases for better usability:
/// - if the option contains a whitespace, it might be a query: clickhouse "SELECT 1"
/// - if the option is a filesystem file, then it's likely a queries file (clickhouse repro.sql)
/// These are relevant for interactive usage - user-friendly, but questionable in general.
/// In case of ambiguity or for scripts, prefer using proper options.
@ -115,8 +118,11 @@ void ClientApplicationBase::parseAndCheckOptions(OptionsDescription & options_de
po::variable_value value(boost::any(op.value), false);
const char * option;
std::error_code ec;
if (token.contains(' '))
option = "query";
else if (std::filesystem::is_regular_file(std::filesystem::path{token}, ec))
option = "queries-file";
else
throw Exception(ErrorCodes::BAD_ARGUMENTS, "Positional option `{}` is not supported.", token);

4
src/Client/IServerConnection.h
View File

@ -109,6 +109,10 @@ public:
/// Send block of data; if name is specified, server will write it to external (temporary) table of that name.
virtual void sendData(const Block & block, const String & name, bool scalar) = 0;
/// Whether the client needs to read and send the data for the INSERT.
/// False if the server will read the data through other means (in particular if clickhouse-local added input reading step directly into the query pipeline).
virtual bool isSendDataNeeded() const { return true; }
/// Send all contents of external (temporary) tables.
virtual void sendExternalTablesData(ExternalTablesData & data) = 0;

5
src/Client/LocalConnection.cpp
View File

@ -328,6 +328,11 @@ void LocalConnection::sendData(const Block & block, const String &, bool)
sendProfileEvents();
}
bool LocalConnection::isSendDataNeeded() const
{
return !state || state->input_pipeline == nullptr;
}
void LocalConnection::sendCancel()
{
state->is_cancelled = true;

2
src/Client/LocalConnection.h
View File

@ -120,6 +120,8 @@ public:
void sendData(const Block & block, const String & name/* = "" */, bool scalar/* = false */) override;
bool isSendDataNeeded() const override;
void sendExternalTablesData(ExternalTablesData &) override;
void sendMergeTreeReadTaskResponse(const ParallelReadResponse & response) override;

20
src/Columns/ColumnFunction.cpp
View File

@ -72,6 +72,26 @@ ColumnPtr ColumnFunction::cut(size_t start, size_t length) const
return ColumnFunction::create(length, function, capture, is_short_circuit_argument, is_function_compiled);
}
Field ColumnFunction::operator[](size_t n) const
{
Field res;
get(n, res);
return res;
}
void ColumnFunction::get(size_t n, Field & res) const
{
const size_t tuple_size = captured_columns.size();
res = Tuple();
Tuple & res_tuple = res.safeGet<Tuple &>();
res_tuple.reserve(tuple_size);
for (size_t i = 0; i < tuple_size; ++i)
res_tuple.push_back((*captured_columns[i].column)[n]);
}
#if !defined(DEBUG_OR_SANITIZER_BUILD)
void ColumnFunction::insertFrom(const IColumn & src, size_t n)
#else

10
src/Columns/ColumnFunction.h
View File

@ -60,15 +60,9 @@ public:
void appendArguments(const ColumnsWithTypeAndName & columns);
ColumnWithTypeAndName reduce() const;
Field operator[](size_t) const override
{
throw Exception(ErrorCodes::NOT_IMPLEMENTED, "Cannot get value from {}", getName());
}
Field operator[](size_t n) const override;
void get(size_t, Field &) const override
{
throw Exception(ErrorCodes::NOT_IMPLEMENTED, "Cannot get value from {}", getName());
}
void get(size_t n, Field & res) const override;
StringRef getDataAt(size_t) const override
{

7
src/Common/CurrentMetrics.cpp
View File

@ -41,6 +41,10 @@
M(PostgreSQLConnection, "Number of client connections using PostgreSQL protocol") \
M(OpenFileForRead, "Number of files open for reading") \
M(OpenFileForWrite, "Number of files open for writing") \
M(Compressing, "Number of compress operations using internal compression codecs") \
M(Decompressing, "Number of decompress operations using internal compression codecs") \
M(ParallelCompressedWriteBufferThreads, "Number of threads in all instances of ParallelCompressedWriteBuffer - these threads are doing parallel compression and writing") \
M(ParallelCompressedWriteBufferWait, "Number of threads in all instances of ParallelCompressedWriteBuffer that are currently waiting for buffer to become available for writing") \
M(TotalTemporaryFiles, "Number of temporary files created") \
M(TemporaryFilesForSort, "Number of temporary files created for external sorting") \
M(TemporaryFilesForAggregation, "Number of temporary files created for external aggregation") \
@ -99,6 +103,9 @@
M(IOThreads, "Number of threads in the IO thread pool.") \
M(IOThreadsActive, "Number of threads in the IO thread pool running a task.") \
M(IOThreadsScheduled, "Number of queued or active jobs in the IO thread pool.") \
M(CompressionThread, "Number of threads in compression thread pools.") \
M(CompressionThreadActive, "Number of threads in compression thread pools running a task.") \
M(CompressionThreadScheduled, "Number of queued or active jobs in compression thread pools.") \
M(ThreadPoolRemoteFSReaderThreads, "Number of threads in the thread pool for remote_filesystem_read_method=threadpool.") \
M(ThreadPoolRemoteFSReaderThreadsActive, "Number of threads in the thread pool for remote_filesystem_read_method=threadpool running a task.") \
M(ThreadPoolRemoteFSReaderThreadsScheduled, "Number of queued or active jobs in the thread pool for remote_filesystem_read_method=threadpool.") \

4
src/Common/ThreadPool.h
View File

@ -122,7 +122,7 @@ public:
void scheduleOrThrowOnError(Job job, Priority priority = {});
/// Similar to scheduleOrThrowOnError(...). Wait for specified amount of time and schedule a job or return false.
bool trySchedule(Job job, Priority priority = {}, uint64_t wait_microseconds = 0) noexcept;
[[nodiscard]] bool trySchedule(Job job, Priority priority = {}, uint64_t wait_microseconds = 0) noexcept;
/// Similar to scheduleOrThrowOnError(...). Wait for specified amount of time and schedule a job or throw an exception.
void scheduleOrThrow(Job job, Priority priority = {}, uint64_t wait_microseconds = 0, bool propagate_opentelemetry_tracing_context = true);
@ -142,7 +142,7 @@ public:
/// Returns true if the pool already terminated
/// (and any further scheduling will produce CANNOT_SCHEDULE_TASK exception)
bool finished() const;
[[nodiscard]] bool finished() const;
void setMaxThreads(size_t value);
void setMaxFreeThreads(size_t value);

1
src/Compression/CompressedWriteBuffer.cpp
View File

@ -2,7 +2,6 @@
#include <cstring>
#include <base/types.h>
#include <base/unaligned.h>
#include <base/defines.h>
#include <IO/WriteHelpers.h>

11
src/Compression/ICompressionCodec.cpp
View File

@ -5,11 +5,18 @@
#include <Parsers/ASTFunction.h>
#include <base/unaligned.h>
#include <Common/Exception.h>
#include <Common/CurrentMetrics.h>
#include <Parsers/queryToString.h>
#include <Parsers/ASTIdentifier.h>
#include <Compression/CompressionCodecMultiple.h>
namespace CurrentMetrics
{
extern const Metric Compressing;
extern const Metric Decompressing;
}
namespace DB
{
@ -80,6 +87,8 @@ UInt32 ICompressionCodec::compress(const char * source, UInt32 source_size, char
{
assert(source != nullptr && dest != nullptr);
CurrentMetrics::Increment metric_increment(CurrentMetrics::Compressing);
dest[0] = getMethodByte();
UInt8 header_size = getHeaderSize();
/// Write data from header_size
@ -93,6 +102,8 @@ UInt32 ICompressionCodec::decompress(const char * source, UInt32 source_size, ch
{
assert(source != nullptr && dest != nullptr);
CurrentMetrics::Increment metric_increment(CurrentMetrics::Decompressing);
UInt8 header_size = getHeaderSize();
if (source_size < header_size)
throw Exception(decompression_error_code,

166
src/Compression/ParallelCompressedWriteBuffer.cpp Normal file
View File

@ -0,0 +1,166 @@
#include <city.h>
#include <base/types.h>
#include <base/defines.h>
#include <IO/WriteHelpers.h>
#include <Common/setThreadName.h>
#include <Common/scope_guard_safe.h>
#include <Common/CurrentThread.h>
#include <Common/CurrentMetrics.h>
#include <Compression/ParallelCompressedWriteBuffer.h>
namespace CurrentMetrics
{
extern const Metric ParallelCompressedWriteBufferThreads;
extern const Metric ParallelCompressedWriteBufferWait;
}
namespace DB
{
ParallelCompressedWriteBuffer::ParallelCompressedWriteBuffer(
WriteBuffer & out_,
CompressionCodecPtr codec_,
size_t buf_size_,
size_t num_threads_,
ThreadPool & pool_)
: WriteBuffer(nullptr, 0), out(out_), codec(codec_), buf_size(buf_size_), num_threads(num_threads_), pool(pool_)
{
buffers.emplace_back(buf_size);
current_buffer = buffers.begin();
BufferBase::set(current_buffer->uncompressed.data(), buf_size, 0);
}
void ParallelCompressedWriteBuffer::nextImpl()
{
if (!offset())
return;
std::unique_lock lock(mutex);
/// The buffer will be compressed and processed in the thread.
current_buffer->busy = true;
current_buffer->sequence_num = current_sequence_num;
++current_sequence_num;
current_buffer->uncompressed_size = offset();
pool.scheduleOrThrowOnError([this, my_current_buffer = current_buffer, thread_group = CurrentThread::getGroup()]
{
SCOPE_EXIT_SAFE(
if (thread_group)
CurrentThread::detachFromGroupIfNotDetached();
);
if (thread_group)
CurrentThread::attachToGroupIfDetached(thread_group);
setThreadName("ParallelCompres");
compress(my_current_buffer);
});
BufferPair * previous_buffer = &*current_buffer;
++current_buffer;
if (current_buffer == buffers.end())
{
if (buffers.size() < num_threads)
{
/// If we didn't use all num_threads buffers yet, create a new one.
current_buffer = buffers.emplace(current_buffer, buf_size);
}
else
{
/// Otherwise, wrap around to the first buffer in the list.
current_buffer = buffers.begin();
}
}
/// Wait while the buffer becomes not busy
if (current_buffer->busy)
{
CurrentMetrics::Increment metric_increment(CurrentMetrics::ParallelCompressedWriteBufferWait);
cond.wait(lock, [&]{ return !current_buffer->busy; });
}
/// Now this buffer can be used.
current_buffer->previous = previous_buffer;
BufferBase::set(current_buffer->uncompressed.data(), buf_size, 0);
}
void ParallelCompressedWriteBuffer::finalizeImpl()
{
next();
pool.wait();
}
void ParallelCompressedWriteBuffer::compress(Iterator buffer)
{
CurrentMetrics::Increment metric_increment(CurrentMetrics::ParallelCompressedWriteBufferThreads);
chassert(buffer->uncompressed_size <= INT_MAX);
UInt32 uncompressed_size = static_cast<UInt32>(buffer->uncompressed_size);
UInt32 compressed_reserve_size = codec->getCompressedReserveSize(uncompressed_size);
/// If all previous buffers have been written,
/// and if the output buffer has the required capacity,
/// we can compress data directly into the output buffer.
size_t required_out_capacity = compressed_reserve_size + sizeof(CityHash_v1_0_2::uint128);
bool can_write_directly = false;
if (!buffer->previous)
{
can_write_directly = out.available() >= required_out_capacity;
}
else
{
std::unique_lock lock(mutex);
can_write_directly = (!buffer->previous->busy || buffer->previous->sequence_num > buffer->sequence_num)
&& out.available() >= required_out_capacity;
}
if (can_write_directly)
{
char * out_compressed_ptr = out.position() + sizeof(CityHash_v1_0_2::uint128);
UInt32 compressed_size = codec->compress(buffer->uncompressed.data(), uncompressed_size, out_compressed_ptr);
CityHash_v1_0_2::uint128 checksum = CityHash_v1_0_2::CityHash128(out_compressed_ptr, compressed_size);
writeBinaryLittleEndian(checksum.low64, out);
writeBinaryLittleEndian(checksum.high64, out);
out.position() += compressed_size;
}
else
{
buffer->compressed.resize(compressed_reserve_size);
UInt32 compressed_size = codec->compress(buffer->uncompressed.data(), uncompressed_size, buffer->compressed.data());
CityHash_v1_0_2::uint128 checksum = CityHash_v1_0_2::CityHash128(buffer->compressed.data(), compressed_size);
/// Wait while all previous buffers have been written.
if (buffer->previous)
{
CurrentMetrics::Increment metric_wait_increment(CurrentMetrics::ParallelCompressedWriteBufferWait);
std::unique_lock lock(mutex);
cond.wait(lock, [&]{ return !buffer->previous->busy || buffer->previous->sequence_num > buffer->sequence_num; });
}
writeBinaryLittleEndian(checksum.low64, out);
writeBinaryLittleEndian(checksum.high64, out);
out.write(buffer->compressed.data(), compressed_size);
}
std::unique_lock lock(mutex);
buffer->busy = false;
cond.notify_all();
}
ParallelCompressedWriteBuffer::~ParallelCompressedWriteBuffer()
{
if (!canceled)
finalize();
}
}

70
src/Compression/ParallelCompressedWriteBuffer.h Normal file
View File

@ -0,0 +1,70 @@
#pragma once
#include <list>
#include <memory>
#include <Common/PODArray.h>
#include <IO/WriteBuffer.h>
#include <IO/BufferWithOwnMemory.h>
#include <Compression/ICompressionCodec.h>
#include <Compression/CompressionFactory.h>
#include <Common/ThreadPool.h>
namespace DB
{
/** Uses multi-buffering for parallel compression.
* When the buffer is filled, it will be compressed in the background,
* and a new buffer is created for the next input data.
*/
class ParallelCompressedWriteBuffer final : public WriteBuffer
{
public:
explicit ParallelCompressedWriteBuffer(
WriteBuffer & out_,
CompressionCodecPtr codec_,
size_t buf_size_,
size_t num_threads_,
ThreadPool & pool_);
~ParallelCompressedWriteBuffer() override;
private:
void nextImpl() override;
void finalizeImpl() override;
WriteBuffer & out;
CompressionCodecPtr codec;
size_t buf_size;
size_t num_threads;
ThreadPool & pool;
struct BufferPair
{
explicit BufferPair(size_t input_size)
: uncompressed(input_size)
{
}
Memory<> uncompressed;
size_t uncompressed_size = 0;
PODArray<char> compressed;
BufferPair * previous = nullptr;
size_t sequence_num = 0;
bool busy = false;
};
std::mutex mutex;
std::condition_variable cond;
std::list<BufferPair> buffers;
using Iterator = std::list<BufferPair>::iterator;
Iterator current_buffer;
size_t current_sequence_num = 0;
void compress(Iterator buffer);
};
}

8
src/Core/Settings.cpp
View File

@ -2892,6 +2892,9 @@ Possible values:
**See Also**
- [ORDER BY Clause](../../sql-reference/statements/select/order-by.md/#optimize_read_in_order)
)", 0) \
DECLARE(Bool, read_in_order_use_virtual_row, false, R"(
Use virtual row while reading in order of primary key or its monotonic function fashion. It is useful when searching over multiple parts as only relevant ones are touched.
)", 0) \
DECLARE(Bool, optimize_read_in_window_order, true, R"(
Enable ORDER BY optimization in window clause for reading data in corresponding order in MergeTree tables.
@ -4877,6 +4880,9 @@ Limit on size of a single batch of file segments that a read buffer can request
)", 0) \
DECLARE(UInt64, filesystem_cache_reserve_space_wait_lock_timeout_milliseconds, 1000, R"(
Wait time to lock cache for space reservation in filesystem cache
)", 0) \
DECLARE(Bool, filesystem_cache_prefer_bigger_buffer_size, true, R"(
Prefer bigger buffer size if filesystem cache is enabled to avoid writing small file segments which deteriorate cache performance. On the other hand, enabling this setting might increase memory usage.
)", 0) \
DECLARE(UInt64, temporary_data_in_cache_reserve_space_wait_lock_timeout_milliseconds, (10 * 60 * 1000), R"(
Wait time to lock cache for space reservation for temporary data in filesystem cache
@ -5713,6 +5719,8 @@ If enabled, MongoDB tables will return an error when a MongoDB query cannot be b
)", 0) \
DECLARE(Bool, implicit_select, false, R"(
Allow writing simple SELECT queries without the leading SELECT keyword, which makes it simple for calculator-style usage, e.g. `1 + 2` becomes a valid query.
In `clickhouse-local` it is enabled by default and can be explicitly disabled.
)", 0) \
\
\

2
src/Core/SettingsChangesHistory.cpp
View File

@ -78,6 +78,8 @@ static std::initializer_list<std::pair<ClickHouseVersion, SettingsChangesHistory
{"backup_restore_keeper_max_retries_while_handling_error", 0, 20, "New setting."},
{"backup_restore_finish_timeout_after_error_sec", 0, 180, "New setting."},
{"parallel_replicas_local_plan", false, true, "Use local plan for local replica in a query with parallel replicas"},
{"filesystem_cache_prefer_bigger_buffer_size", true, true, "New setting"},
{"read_in_order_use_virtual_row", false, false, "Use virtual row while reading in order of primary key or its monotonic function fashion. It is useful when searching over multiple parts as only relevant ones are touched."},
}
},
{"24.10",

30
src/Databases/DatabaseAtomic.cpp
View File

@ -6,7 +6,6 @@
#include <Databases/DatabaseReplicated.h>
#include <IO/ReadBufferFromFile.h>
#include <IO/ReadHelpers.h>
#include <IO/WriteHelpers.h>
#include <Interpreters/Context.h>
#include <Interpreters/DDLTask.h>
#include <Interpreters/DatabaseCatalog.h>
@ -19,6 +18,7 @@
#include <Common/filesystemHelpers.h>
#include <Core/Settings.h>
namespace fs = std::filesystem;
namespace DB
@ -60,9 +60,6 @@ DatabaseAtomic::DatabaseAtomic(String name_, String metadata_path_, UUID uuid, c
, db_uuid(uuid)
{
assert(db_uuid != UUIDHelpers::Nil);
fs::create_directories(fs::path(getContext()->getPath()) / "metadata");
fs::create_directories(path_to_table_symlinks);
tryCreateMetadataSymlink();
}
DatabaseAtomic::DatabaseAtomic(String name_, String metadata_path_, UUID uuid, ContextPtr context_)
@ -70,6 +67,20 @@ DatabaseAtomic::DatabaseAtomic(String name_, String metadata_path_, UUID uuid, C
{
}
void DatabaseAtomic::createDirectories()
{
std::lock_guard lock(mutex);
createDirectoriesUnlocked();
}
void DatabaseAtomic::createDirectoriesUnlocked()
{
DatabaseOnDisk::createDirectoriesUnlocked();
fs::create_directories(fs::path(getContext()->getPath()) / "metadata");
fs::create_directories(path_to_table_symlinks);
tryCreateMetadataSymlink();
}
String DatabaseAtomic::getTableDataPath(const String & table_name) const
{
std::lock_guard lock(mutex);
@ -108,6 +119,7 @@ void DatabaseAtomic::attachTable(ContextPtr /* context_ */, const String & name,
assert(relative_table_path != data_path && !relative_table_path.empty());
DetachedTables not_in_use;
std::lock_guard lock(mutex);
createDirectoriesUnlocked();
not_in_use = cleanupDetachedTables();
auto table_id = table->getStorageID();
assertDetachedTableNotInUse(table_id.uuid);
@ -208,11 +220,15 @@ void DatabaseAtomic::renameTable(ContextPtr local_context, const String & table_
if (exchange && !supportsAtomicRename(&message))
throw Exception(ErrorCodes::NOT_IMPLEMENTED, "RENAME EXCHANGE is not supported because exchanging files is not supported by the OS ({})", message);
createDirectories();
waitDatabaseStarted();
auto & other_db = dynamic_cast<DatabaseAtomic &>(to_database);
bool inside_database = this == &other_db;
if (!inside_database)
other_db.createDirectories();
String old_metadata_path = getObjectMetadataPath(table_name);
String new_metadata_path = to_database.getObjectMetadataPath(to_table_name);
@ -333,6 +349,7 @@ void DatabaseAtomic::commitCreateTable(const ASTCreateQuery & query, const Stora
const String & table_metadata_tmp_path, const String & table_metadata_path,
ContextPtr query_context)
{
createDirectories();
DetachedTables not_in_use;
auto table_data_path = getTableDataPath(query);
try
@ -469,6 +486,9 @@ void DatabaseAtomic::beforeLoadingMetadata(ContextMutablePtr /*context*/, Loadin
if (mode < LoadingStrictnessLevel::FORCE_RESTORE)
return;
if (!fs::exists(path_to_table_symlinks))
return;
/// Recreate symlinks to table data dirs in case of force restore, because some of them may be broken
for (const auto & table_path : fs::directory_iterator(path_to_table_symlinks))
{
@ -611,6 +631,7 @@ void DatabaseAtomic::renameDatabase(ContextPtr query_context, const String & new
{
/// CREATE, ATTACH, DROP, DETACH and RENAME DATABASE must hold DDLGuard
createDirectories();
waitDatabaseStarted();
bool check_ref_deps = query_context->getSettingsRef()[Setting::check_referential_table_dependencies];
@ -702,4 +723,5 @@ void registerDatabaseAtomic(DatabaseFactory & factory)
};
factory.registerDatabase("Atomic", create_fn);
}
}

3
src/Databases/DatabaseAtomic.h
View File

@ -76,6 +76,9 @@ protected:
using DetachedTables = std::unordered_map<UUID, StoragePtr>;
[[nodiscard]] DetachedTables cleanupDetachedTables() TSA_REQUIRES(mutex);
void createDirectories();
void createDirectoriesUnlocked() TSA_REQUIRES(mutex);
void tryCreateMetadataSymlink();
virtual bool allowMoveTableToOtherDatabaseEngine(IDatabase & /*to_database*/) const { return false; }

1
src/Databases/DatabaseLazy.cpp
View File

@ -47,6 +47,7 @@ DatabaseLazy::DatabaseLazy(const String & name_, const String & metadata_path_,
: DatabaseOnDisk(name_, metadata_path_, std::filesystem::path("data") / escapeForFileName(name_) / "", "DatabaseLazy (" + name_ + ")", context_)
, expiration_time(expiration_time_)
{
createDirectories();
}

22
src/Databases/DatabaseOnDisk.cpp
View File

@ -180,7 +180,18 @@ DatabaseOnDisk::DatabaseOnDisk(
, metadata_path(metadata_path_)
, data_path(data_path_)
{
fs::create_directories(local_context->getPath() + data_path);
}
void DatabaseOnDisk::createDirectories()
{
std::lock_guard lock(mutex);
createDirectoriesUnlocked();
}
void DatabaseOnDisk::createDirectoriesUnlocked()
{
fs::create_directories(std::filesystem::path(getContext()->getPath()) / data_path);
fs::create_directories(metadata_path);
}
@ -198,6 +209,8 @@ void DatabaseOnDisk::createTable(
const StoragePtr & table,
const ASTPtr & query)
{
createDirectories();
const auto & settings = local_context->getSettingsRef();
const auto & create = query->as<ASTCreateQuery &>();
assert(table_name == create.getTable());
@ -265,7 +278,6 @@ void DatabaseOnDisk::createTable(
}
commitCreateTable(create, table, table_metadata_tmp_path, table_metadata_path, local_context);
removeDetachedPermanentlyFlag(local_context, table_name, table_metadata_path, false);
}
@ -293,6 +305,8 @@ void DatabaseOnDisk::commitCreateTable(const ASTCreateQuery & query, const Stora
{
try
{
createDirectories();
/// Add a table to the map of known tables.
attachTable(query_context, query.getTable(), table, getTableDataPath(query));
@ -426,6 +440,7 @@ void DatabaseOnDisk::renameTable(
throw Exception(ErrorCodes::NOT_IMPLEMENTED, "Moving tables between databases of different engines is not supported");
}
createDirectories();
waitDatabaseStarted();
auto table_data_relative_path = getTableDataPath(table_name);
@ -621,6 +636,9 @@ time_t DatabaseOnDisk::getObjectMetadataModificationTime(const String & object_n
void DatabaseOnDisk::iterateMetadataFiles(const IteratingFunction & process_metadata_file) const
{
if (!fs::exists(metadata_path))
return;
auto process_tmp_drop_metadata_file = [&](const String & file_name)
{
assert(getUUID() == UUIDHelpers::Nil);

3
src/Databases/DatabaseOnDisk.h
View File

@ -99,6 +99,9 @@ protected:
virtual void removeDetachedPermanentlyFlag(ContextPtr context, const String & table_name, const String & table_metadata_path, bool attach);
virtual void setDetachedTableNotInUseForce(const UUID & /*uuid*/) {}
void createDirectories();
void createDirectoriesUnlocked() TSA_REQUIRES(mutex);
const String metadata_path;
const String data_path;
};

1
src/Databases/IDatabase.h
View File

@ -416,6 +416,7 @@ public:
std::lock_guard lock{mutex};
return database_name;
}
/// Get UUID of database.
virtual UUID getUUID() const { return UUIDHelpers::Nil; }

1
src/Databases/MySQL/DatabaseMaterializedMySQL.cpp
View File

@ -62,6 +62,7 @@ DatabaseMaterializedMySQL::DatabaseMaterializedMySQL(
, settings(std::move(settings_))
, materialize_thread(context_, database_name_, mysql_database_name_, std::move(pool_), std::move(client_), binlog_client_, settings.get())
{
createDirectories();
}
DatabaseMaterializedMySQL::~DatabaseMaterializedMySQL() = default;

28
src/Dictionaries/HashedDictionary.h
View File

@ -334,22 +334,26 @@ HashedDictionary<dictionary_key_type, sparse, sharded>::~HashedDictionary()
if (container.empty())
return;
pool.trySchedule([&container, thread_group = CurrentThread::getGroup()]
{
SCOPE_EXIT_SAFE(
if (!pool.trySchedule([&container, thread_group = CurrentThread::getGroup()]
{
SCOPE_EXIT_SAFE(
if (thread_group)
CurrentThread::detachFromGroupIfNotDetached();
);
/// Do not account memory that was occupied by the dictionaries for the query/user context.
MemoryTrackerBlockerInThread memory_blocker;
if (thread_group)
CurrentThread::detachFromGroupIfNotDetached();
);
CurrentThread::attachToGroupIfDetached(thread_group);
setThreadName("HashedDictDtor");
/// Do not account memory that was occupied by the dictionaries for the query/user context.
clearContainer(container);
}))
{
MemoryTrackerBlockerInThread memory_blocker;
if (thread_group)
CurrentThread::attachToGroupIfDetached(thread_group);
setThreadName("HashedDictDtor");
clearContainer(container);
});
}
++hash_tables_count;
};

5
src/Disks/ObjectStorages/DiskObjectStorage.cpp
View File

@ -642,7 +642,10 @@ std::unique_ptr<ReadBufferFromFileBase> DiskObjectStorage::readFile(
};
/// Avoid cache fragmentation by choosing bigger buffer size.
bool prefer_bigger_buffer_size = object_storage->supportsCache() && read_settings.enable_filesystem_cache;
bool prefer_bigger_buffer_size = read_settings.filesystem_cache_prefer_bigger_buffer_size
&& object_storage->supportsCache()
&& read_settings.enable_filesystem_cache;
size_t buffer_size = prefer_bigger_buffer_size
? std::max<size_t>(settings.remote_fs_buffer_size, DBMS_DEFAULT_BUFFER_SIZE)
: settings.remote_fs_buffer_size;

107
src/Formats/PrettyFormatHelpers.cpp Normal file
View File

@ -0,0 +1,107 @@
#include <Formats/PrettyFormatHelpers.h>
#include <IO/WriteBuffer.h>
#include <IO/WriteHelpers.h>
#include <Processors/Chunk.h>
#include <Common/formatReadable.h>
static constexpr const char * GRAY_COLOR = "\033[90m";
static constexpr const char * UNDERSCORE = "\033[4m";
static constexpr const char * RESET_COLOR = "\033[0m";
namespace DB
{
void writeReadableNumberTipIfSingleValue(WriteBuffer & out, const Chunk & chunk, const FormatSettings & settings, bool color)
{
if (chunk.getNumRows() == 1 && chunk.getNumColumns() == 1)
writeReadableNumberTip(out, *chunk.getColumns()[0], 0, settings, color);
}
void writeReadableNumberTip(WriteBuffer & out, const IColumn & column, size_t row, const FormatSettings & settings, bool color)
{
if (column.isNullAt(row))
return;
auto value = column.getFloat64(row);
auto threshold = settings.pretty.output_format_pretty_single_large_number_tip_threshold;
if (threshold && isFinite(value) && abs(value) > threshold)
{
if (color)
writeCString(GRAY_COLOR, out);
writeCString(" -- ", out);
formatReadableQuantity(value, out, 2);
if (color)
writeCString(RESET_COLOR, out);
}
}
String highlightDigitGroups(String source)
{
if (source.size() <= 4)
return source;
bool is_regular_number = true;
size_t num_digits_before_decimal = 0;
for (auto c : source)
{
if (c == '-' || c == ' ')
continue;
if (c == '.')
break;
if (c >= '0' && c <= '9')
{
++num_digits_before_decimal;
}
else
{
is_regular_number = false;
break;
}
}
if (!is_regular_number || num_digits_before_decimal <= 4)
return source;
String result;
size_t size = source.size();
result.reserve(2 * size);
bool before_decimal = true;
size_t digit_num = 0;
for (size_t i = 0; i < size; ++i)
{
auto c = source[i];
if (before_decimal && c >= '0' && c <= '9')
{
++digit_num;
size_t offset = num_digits_before_decimal - digit_num;
if (offset && offset % 3 == 0)
{
result += UNDERSCORE;
result += c;
result += RESET_COLOR;
}
else
{
result += c;
}
}
else if (c == '.')
{
before_decimal = false;
result += c;
}
else
{
result += c;
}
}
return result;
}
}

21
src/Formats/PrettyFormatHelpers.h Normal file
View File

@ -0,0 +1,21 @@
#pragma once
#include <base/types.h>
namespace DB
{
class Chunk;
class IColumn;
class WriteBuffer;
struct FormatSettings;
/// Prints text describing the number in the form of: -- 12.34 million
void writeReadableNumberTip(WriteBuffer & out, const IColumn & column, size_t row, const FormatSettings & settings, bool color);
void writeReadableNumberTipIfSingleValue(WriteBuffer & out, const Chunk & chunk, const FormatSettings & settings, bool color);
/// Underscores digit groups related to thousands using terminal ANSI escape sequences.
String highlightDigitGroups(String source);
}

2
src/Functions/FunctionsConversion.cpp
View File

@ -4410,7 +4410,7 @@ private:
variant_column = IColumn::mutate(column);
/// Otherwise we should filter column.
else
variant_column = column->filter(filter, variant_size_hint)->assumeMutable();
variant_column = IColumn::mutate(column->filter(filter, variant_size_hint));
assert_cast<ColumnLowCardinality &>(*variant_column).nestedRemoveNullable();
return createVariantFromDescriptorsAndOneNonEmptyVariant(variant_types, std::move(discriminators), std::move(variant_column), variant_discr);

28
src/Functions/FunctionsMiscellaneous.h
View File

@ -6,6 +6,7 @@
#include <IO/WriteBufferFromString.h>
#include <IO/Operators.h>
#include <Columns/ColumnFunction.h>
#include <Columns/ColumnConst.h>
#include <DataTypes/DataTypesNumber.h>
@ -112,6 +113,7 @@ public:
NamesAndTypesList lambda_arguments;
String return_name;
DataTypePtr return_type;
bool allow_constant_folding;
};
using CapturePtr = std::shared_ptr<Capture>;
@ -122,6 +124,7 @@ public:
String getName() const override { return "FunctionCapture"; }
bool useDefaultImplementationForNulls() const override { return false; }
/// It's possible if expression_actions contains function that don't use
/// default implementation for Nothing and one of captured columns can be Nothing
/// Example: SELECT arrayMap(x -> [x, arrayElement(y, 0)], []), [] as y
@ -148,7 +151,26 @@ public:
auto function = std::make_unique<FunctionExpression>(expression_actions, types, names,
capture->return_type, capture->return_name);
return ColumnFunction::create(input_rows_count, std::move(function), arguments);
/// If all the captured arguments are constant, let's also return ColumnConst (with ColumnFunction inside it).
/// Consequently, it allows to treat higher order functions with constant arrays and constant captured columns
/// as constant expressions.
/// Consequently, it allows its usage in contexts requiring constants, such as the right hand side of IN.
bool constant_folding = capture->allow_constant_folding
&& std::all_of(arguments.begin(), arguments.end(),
[](const auto & arg) { return arg.column->isConst(); });
if (constant_folding)
{
ColumnsWithTypeAndName arguments_resized = arguments;
for (auto & elem : arguments_resized)
elem.column = elem.column->cloneResized(1);
return ColumnConst::create(ColumnFunction::create(1, std::move(function), arguments_resized), input_rows_count);
}
else
{
return ColumnFunction::create(input_rows_count, std::move(function), arguments);
}
}
private:
@ -203,7 +225,8 @@ public:
const Names & captured_names,
const NamesAndTypesList & lambda_arguments,
const DataTypePtr & function_return_type,
const String & expression_return_name)
const String & expression_return_name,
bool allow_constant_folding)
: expression_actions(std::move(expression_actions_))
{
/// Check that expression does not contain unusual actions that will break columns structure.
@ -246,6 +269,7 @@ public:
.lambda_arguments = lambda_arguments,
.return_name = expression_return_name,
.return_type = function_return_type,
.allow_constant_folding = allow_constant_folding,
});
}

2
src/Functions/IFunction.h
View File

@ -184,7 +184,7 @@ public:
/** If function isSuitableForConstantFolding then, this method will be called during query analysis
* if some arguments are constants. For example logical functions (AndFunction, OrFunction) can
* return they result based on some constant arguments.
* return the result based on some constant arguments.
* Arguments are passed without modifications, useDefaultImplementationForNulls, useDefaultImplementationForNothing,
* useDefaultImplementationForConstants, useDefaultImplementationForLowCardinality are not applied.
*/

4
src/Functions/array/FunctionArrayMapped.h
View File

@ -282,7 +282,9 @@ public:
if (!column_with_type_and_name.column)
throw Exception(ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT, "First argument for function {} must be a function.", getName());
const auto * column_function = typeid_cast<const ColumnFunction *>(column_with_type_and_name.column.get());
auto column_function_materialized = column_with_type_and_name.column->convertToFullColumnIfConst();
const auto * column_function = typeid_cast<const ColumnFunction *>(column_function_materialized.get());
if (!column_function)
throw Exception(ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT, "First argument for function {} must be a function.", getName());

4
src/Functions/array/arrayFold.cpp
View File

@ -87,7 +87,9 @@ public:
if (!lambda_function_with_type_and_name.column)
throw Exception(ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT, "First argument for function {} must be a function", getName());
const auto * lambda_function = typeid_cast<const ColumnFunction *>(lambda_function_with_type_and_name.column.get());
auto lambda_function_materialized = lambda_function_with_type_and_name.column->convertToFullColumnIfConst();
const auto * lambda_function = typeid_cast<const ColumnFunction *>(lambda_function_materialized.get());
if (!lambda_function)
throw Exception(ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT, "First argument for function {} must be a function", getName());

19
src/Functions/nested.cpp
View File

@ -108,27 +108,29 @@ public:
{
size_t arguments_size = arguments.size();
const auto * lhs_array = assert_cast<const ColumnArray *>(arguments.at(1).column.get());
ColumnPtr first_array_materialized = arguments[1].column->convertToFullColumnIfConst();
const ColumnArray & first_array = assert_cast<const ColumnArray &>(*first_array_materialized);
Columns data_columns;
data_columns.reserve(arguments_size);
data_columns.push_back(lhs_array->getDataPtr());
data_columns.push_back(first_array.getDataPtr());
for (size_t i = 2; i < arguments_size; ++i)
{
const auto * rhs_array = assert_cast<const ColumnArray *>(arguments[i].column.get());
ColumnPtr other_array_materialized = arguments[i].column->convertToFullColumnIfConst();
const ColumnArray & other_array = assert_cast<const ColumnArray &>(*other_array_materialized);
if (!lhs_array->hasEqualOffsets(*rhs_array))
if (!first_array.hasEqualOffsets(other_array))
throw Exception(ErrorCodes::SIZES_OF_ARRAYS_DONT_MATCH,
"The argument 2 and argument {} of function {} have different array offsets",
i + 1,
getName());
data_columns.push_back(rhs_array->getDataPtr());
data_columns.push_back(other_array.getDataPtr());
}
auto tuple_column = ColumnTuple::create(std::move(data_columns));
auto array_column = ColumnArray::create(std::move(tuple_column), lhs_array->getOffsetsPtr());
auto array_column = ColumnArray::create(std::move(tuple_column), first_array.getOffsetsPtr());
return array_column;
}
@ -168,7 +170,12 @@ REGISTER_FUNCTION(Nested)
{
factory.registerFunction<FunctionNested>(FunctionDocumentation{
.description=R"(
This is a function used internally by the ClickHouse engine and not meant to be used directly.
Returns the array of tuples from multiple arrays.
The first argument must be a constant array of Strings determining the names of the resulting Tuple.
The other arguments must be arrays of the same size.
)",
.examples{{"nested", "SELECT nested(['keys', 'values'], ['key_1', 'key_2'], ['value_1','value_2'])", ""}},
.categories{"OtherFunctions"}

1
src/IO/ReadSettings.h
View File

@ -61,6 +61,7 @@ struct ReadSettings
bool filesystem_cache_allow_background_download = true;
bool filesystem_cache_allow_background_download_for_metadata_files_in_packed_storage = true;
bool filesystem_cache_allow_background_download_during_fetch = true;
bool filesystem_cache_prefer_bigger_buffer_size = true;
bool use_page_cache_for_disks_without_file_cache = false;
bool read_from_page_cache_if_exists_otherwise_bypass_cache = false;

5
src/IO/S3/Client.cpp
View File

@ -645,7 +645,7 @@ Client::doRequestWithRetryNetworkErrors(RequestType & request, RequestFn request
try
{
/// S3 does retries network errors actually.
/// But it is matter when errors occur.
/// But it does matter when errors occur.
/// This code retries a specific case when
/// network error happens when XML document is being read from the response body.
/// Hence, the response body is a stream, network errors are possible at reading.
@ -656,8 +656,9 @@ Client::doRequestWithRetryNetworkErrors(RequestType & request, RequestFn request
/// Requests that expose the response stream as an answer are not retried with that code. E.g. GetObject.
return request_fn_(request_);
}
catch (Poco::Net::ConnectionResetException &)
catch (Poco::Net::NetException &)
{
/// This includes "connection reset", "malformed message", and possibly other exceptions.
if constexpr (IsReadMethod)
{

2
src/Interpreters/ActionsVisitor.cpp
View File

@ -1308,7 +1308,7 @@ void ActionsMatcher::visit(const ASTFunction & node, const ASTPtr & ast, Data &
String lambda_name = data.getUniqueName("__lambda");
auto function_capture = std::make_shared<FunctionCaptureOverloadResolver>(
lambda_actions, captured, lambda_arguments, result_type, result_name);
lambda_actions, captured, lambda_arguments, result_type, result_name, false);
data.addFunction(function_capture, captured, lambda_name);
argument_types[i] = std::make_shared<DataTypeFunction>(lambda_type->getArgumentTypes(), result_type);

2
src/Interpreters/Context.cpp
View File

@ -196,6 +196,7 @@ namespace Setting
extern const SettingsUInt64 filesystem_cache_segments_batch_size;
extern const SettingsBool filesystem_cache_enable_background_download_for_metadata_files_in_packed_storage;
extern const SettingsBool filesystem_cache_enable_background_download_during_fetch;
extern const SettingsBool filesystem_cache_prefer_bigger_buffer_size;
extern const SettingsBool http_make_head_request;
extern const SettingsUInt64 http_max_fields;
extern const SettingsUInt64 http_max_field_name_size;
@ -5751,6 +5752,7 @@ ReadSettings Context::getReadSettings() const
res.filesystem_cache_allow_background_download_for_metadata_files_in_packed_storage
= settings_ref[Setting::filesystem_cache_enable_background_download_for_metadata_files_in_packed_storage];
res.filesystem_cache_allow_background_download_during_fetch = settings_ref[Setting::filesystem_cache_enable_background_download_during_fetch];
res.filesystem_cache_prefer_bigger_buffer_size = settings_ref[Setting::filesystem_cache_prefer_bigger_buffer_size];
res.filesystem_cache_max_download_size = settings_ref[Setting::filesystem_cache_max_download_size];
res.skip_download_if_exceeds_query_cache = settings_ref[Setting::skip_download_if_exceeds_query_cache];

19
src/Interpreters/ExecuteScalarSubqueriesVisitor.cpp
View File

@ -5,9 +5,11 @@
#include <Core/Settings.h>
#include <DataTypes/DataTypeNullable.h>
#include <DataTypes/DataTypeTuple.h>
#include <IO/WriteBufferFromString.h>
#include <IO/WriteHelpers.h>
#include <Interpreters/Context.h>
#include <Interpreters/InterpreterSelectWithUnionQuery.h>
#include <Interpreters/ProcessorsProfileLog.h>
#include <Interpreters/addTypeConversionToAST.h>
#include <Interpreters/misc.h>
#include <Parsers/ASTExpressionList.h>
@ -19,9 +21,8 @@
#include <Parsers/ASTWithElement.h>
#include <Parsers/queryToString.h>
#include <Processors/Executors/PullingAsyncPipelineExecutor.h>
#include <Common/ProfileEvents.h>
#include <Common/FieldVisitorToString.h>
#include <IO/WriteBufferFromString.h>
#include <Common/ProfileEvents.h>
namespace ProfileEvents
{
@ -67,6 +68,18 @@ bool ExecuteScalarSubqueriesMatcher::needChildVisit(ASTPtr & node, const ASTPtr
return false;
}
if (auto * tables = node->as<ASTTablesInSelectQueryElement>())
{
/// Contrary to what's said in the code block above, ARRAY JOIN needs to resolve the subquery if possible
/// and assign an alias for 02367_optimize_trivial_count_with_array_join to pass. Otherwise it will fail in
/// ArrayJoinedColumnsVisitor (`No alias for non-trivial value in ARRAY JOIN: _a`)
/// This looks 100% as a incomplete code working on top of a bug, but this code has already been made obsolete
/// by the new analyzer, so it's an inconvenience we can live with until we deprecate it.
if (child == tables->array_join)
return true;
return false;
}
return true;
}
@ -246,6 +259,8 @@ void ExecuteScalarSubqueriesMatcher::visit(const ASTSubquery & subquery, ASTPtr
if (tmp_block.rows() != 0)
throw Exception(ErrorCodes::INCORRECT_RESULT_OF_SCALAR_SUBQUERY, "Scalar subquery returned more than one row");
logProcessorProfile(data.getContext(), io.pipeline.getProcessors());
}
block = materializeBlock(block);

2
src/Interpreters/InterpreterCreateQuery.cpp
View File

@ -1467,7 +1467,7 @@ BlockIO InterpreterCreateQuery::createTable(ASTCreateQuery & create)
bool is_secondary_query = getContext()->getZooKeeperMetadataTransaction() && !getContext()->getZooKeeperMetadataTransaction()->isInitialQuery();
auto mode = getLoadingStrictnessLevel(create.attach, /*force_attach*/ false, /*has_force_restore_data_flag*/ false, is_secondary_query || is_restore_from_backup);
if (!create.sql_security && create.supportSQLSecurity() && !getContext()->getServerSettings()[ServerSetting::ignore_empty_sql_security_in_create_view_query])
if (!create.sql_security && create.supportSQLSecurity() && (create.refresh_strategy || !getContext()->getServerSettings()[ServerSetting::ignore_empty_sql_security_in_create_view_query]))
create.sql_security = std::make_shared<ASTSQLSecurity>();
if (create.sql_security)

3
src/Interpreters/InterpreterInsertQuery.cpp
View File

@ -121,8 +121,7 @@ StoragePtr InterpreterInsertQuery::getTable(ASTInsertQuery & query)
if (current_context->getSettingsRef()[Setting::allow_experimental_analyzer])
{
InterpreterSelectQueryAnalyzer interpreter_select(query.select, current_context, select_query_options);
header_block = interpreter_select.getSampleBlock();
header_block = InterpreterSelectQueryAnalyzer::getSampleBlock(query.select, current_context, select_query_options);
}
else
{

4
src/Interpreters/InterpreterSystemQuery.cpp
View File

@ -795,9 +795,9 @@ BlockIO InterpreterSystemQuery::execute()
case Type::WAIT_FAILPOINT:
{
getContext()->checkAccess(AccessType::SYSTEM_FAILPOINT);
LOG_TRACE(log, "waiting for failpoint {}", query.fail_point_name);
LOG_TRACE(log, "Waiting for failpoint {}", query.fail_point_name);
FailPointInjection::pauseFailPoint(query.fail_point_name);
LOG_TRACE(log, "finished failpoint {}", query.fail_point_name);
LOG_TRACE(log, "Finished waiting for failpoint {}", query.fail_point_name);
break;
}
case Type::RESET_COVERAGE:

19
src/Interpreters/PreparedSets.cpp
View File

@ -1,21 +1,22 @@
#include <chrono>
#include <variant>
#include <Interpreters/PreparedSets.h>
#include <Interpreters/Set.h>
#include <Interpreters/InterpreterSelectWithUnionQuery.h>
#include <IO/Operators.h>
#include <Common/logger_useful.h>
#include <Core/Block.h>
#include <Core/Settings.h>
#include <Processors/QueryPlan/CreatingSetsStep.h>
#include <IO/Operators.h>
#include <Interpreters/InterpreterSelectWithUnionQuery.h>
#include <Interpreters/PreparedSets.h>
#include <Interpreters/ProcessorsProfileLog.h>
#include <Interpreters/Set.h>
#include <Processors/Executors/CompletedPipelineExecutor.h>
#include <Processors/QueryPlan/BuildQueryPipelineSettings.h>
#include <Processors/QueryPlan/CreatingSetsStep.h>
#include <Processors/QueryPlan/Optimizations/QueryPlanOptimizationSettings.h>
#include <Processors/QueryPlan/QueryPlan.h>
#include <Processors/Sinks/EmptySink.h>
#include <Processors/Sinks/NullSink.h>
#include <Processors/QueryPlan/QueryPlan.h>
#include <Core/Block.h>
#include <QueryPipeline/QueryPipelineBuilder.h>
#include <QueryPipeline/SizeLimits.h>
#include <Common/logger_useful.h>
namespace DB
{
@ -239,6 +240,8 @@ SetPtr FutureSetFromSubquery::buildOrderedSetInplace(const ContextPtr & context)
if (!set_and_key->set->isCreated())
return nullptr;
logProcessorProfile(context, pipeline.getProcessors());
return set_and_key->set;
}

62
src/Interpreters/ProcessorsProfileLog.cpp
View File

@ -1,5 +1,6 @@
#include <Interpreters/ProcessorsProfileLog.h>
#include <Core/Settings.h>
#include <DataTypes/DataTypeArray.h>
#include <DataTypes/DataTypeDate.h>
#include <DataTypes/DataTypeDateTime.h>
@ -8,16 +9,19 @@
#include <DataTypes/DataTypeNullable.h>
#include <DataTypes/DataTypeString.h>
#include <DataTypes/DataTypesNumber.h>
#include <Interpreters/Context.h>
#include <base/getFQDNOrHostName.h>
#include <Common/ClickHouseRevision.h>
#include <Common/DateLUTImpl.h>
#include <Common/logger_useful.h>
#include <array>
namespace DB
{
namespace Setting
{
extern const SettingsBool log_processors_profiles;
}
ColumnsDescription ProcessorProfileLogElement::getColumnsDescription()
{
return ColumnsDescription
@ -81,5 +85,57 @@ void ProcessorProfileLogElement::appendToBlock(MutableColumns & columns) const
columns[i++]->insert(output_bytes);
}
void logProcessorProfile(ContextPtr context, const Processors & processors)
{
const Settings & settings = context->getSettingsRef();
if (settings[Setting::log_processors_profiles])
{
if (auto processors_profile_log = context->getProcessorsProfileLog())
{
ProcessorProfileLogElement processor_elem;
const auto time_now = std::chrono::system_clock::now();
processor_elem.event_time = timeInSeconds(time_now);
processor_elem.event_time_microseconds = timeInMicroseconds(time_now);
processor_elem.initial_query_id = context->getInitialQueryId();
processor_elem.query_id = context->getCurrentQueryId();
auto get_proc_id = [](const IProcessor & proc) -> UInt64 { return reinterpret_cast<std::uintptr_t>(&proc); };
for (const auto & processor : processors)
{
std::vector<UInt64> parents;
for (const auto & port : processor->getOutputs())
{
if (!port.isConnected())
continue;
const IProcessor & next = port.getInputPort().getProcessor();
parents.push_back(get_proc_id(next));
}
processor_elem.id = get_proc_id(*processor);
processor_elem.parent_ids = std::move(parents);
processor_elem.plan_step = reinterpret_cast<std::uintptr_t>(processor->getQueryPlanStep());
processor_elem.plan_step_name = processor->getPlanStepName();
processor_elem.plan_step_description = processor->getPlanStepDescription();
processor_elem.plan_group = processor->getQueryPlanStepGroup();
processor_elem.processor_name = processor->getName();
processor_elem.elapsed_us = static_cast<UInt64>(processor->getElapsedNs() / 1000U);
processor_elem.input_wait_elapsed_us = static_cast<UInt64>(processor->getInputWaitElapsedNs() / 1000U);
processor_elem.output_wait_elapsed_us = static_cast<UInt64>(processor->getOutputWaitElapsedNs() / 1000U);
auto stats = processor->getProcessorDataStats();
processor_elem.input_rows = stats.input_rows;
processor_elem.input_bytes = stats.input_bytes;
processor_elem.output_rows = stats.output_rows;
processor_elem.output_bytes = stats.output_bytes;
processors_profile_log->add(processor_elem);
}
}
}
}
}

1
src/Interpreters/ProcessorsProfileLog.h
View File

@ -50,4 +50,5 @@ public:
using SystemLog<ProcessorProfileLogElement>::SystemLog;
};
void logProcessorProfile(ContextPtr context, const Processors & processors);
}

6
src/Interpreters/QueryNormalizer.cpp
View File

@ -161,7 +161,13 @@ void QueryNormalizer::visit(ASTTablesInSelectQueryElement & node, const ASTPtr &
{
auto & join = node.table_join->as<ASTTableJoin &>();
if (join.on_expression)
{
ASTPtr original_on_expression = join.on_expression;
visit(join.on_expression, data);
if (join.on_expression != original_on_expression)
join.children = { join.on_expression };
}
}
}

18
src/Interpreters/RewriteArrayExistsFunctionVisitor.cpp
View File

@ -6,6 +6,12 @@
namespace DB
{
namespace ErrorCode
{
extern const int LOGICAL_ERROR;
}
void RewriteArrayExistsFunctionMatcher::visit(ASTPtr & ast, Data & data)
{
if (auto * func = ast->as<ASTFunction>())
@ -20,21 +26,21 @@ void RewriteArrayExistsFunctionMatcher::visit(ASTPtr & ast, Data & data)
if (join->using_expression_list)
{
auto * it = std::find(join->children.begin(), join->children.end(), join->using_expression_list);
if (it == join->children.end())
throw Exception(ErrorCodes::LOGICAL_ERROR, "Could not find join->using_expression_list in '{}'", join->formatForLogging());
visit(join->using_expression_list, data);
if (it && *it != join->using_expression_list)
*it = join->using_expression_list;
*it = join->using_expression_list;
}
if (join->on_expression)
{
auto * it = std::find(join->children.begin(), join->children.end(), join->on_expression);
if (it == join->children.end())
throw Exception(ErrorCodes::LOGICAL_ERROR, "Could not find join->on_expression in '{}'", join->formatForLogging());
visit(join->on_expression, data);
if (it && *it != join->on_expression)
*it = join->on_expression;
*it = join->on_expression;
}
}
}

48
src/Interpreters/executeQuery.cpp
View File

@ -117,7 +117,6 @@ namespace Setting
extern const SettingsOverflowMode join_overflow_mode;
extern const SettingsString log_comment;
extern const SettingsBool log_formatted_queries;
extern const SettingsBool log_processors_profiles;
extern const SettingsBool log_profile_events;
extern const SettingsUInt64 log_queries_cut_to_length;
extern const SettingsBool log_queries;
@ -551,53 +550,8 @@ void logQueryFinish(
if (auto query_log = context->getQueryLog())
query_log->add(elem);
}
if (settings[Setting::log_processors_profiles])
{
if (auto processors_profile_log = context->getProcessorsProfileLog())
{
ProcessorProfileLogElement processor_elem;
processor_elem.event_time = elem.event_time;
processor_elem.event_time_microseconds = elem.event_time_microseconds;
processor_elem.initial_query_id = elem.client_info.initial_query_id;
processor_elem.query_id = elem.client_info.current_query_id;
auto get_proc_id = [](const IProcessor & proc) -> UInt64 { return reinterpret_cast<std::uintptr_t>(&proc); };
for (const auto & processor : query_pipeline.getProcessors())
{
std::vector<UInt64> parents;
for (const auto & port : processor->getOutputs())
{
if (!port.isConnected())
continue;
const IProcessor & next = port.getInputPort().getProcessor();
parents.push_back(get_proc_id(next));
}
processor_elem.id = get_proc_id(*processor);
processor_elem.parent_ids = std::move(parents);
processor_elem.plan_step = reinterpret_cast<std::uintptr_t>(processor->getQueryPlanStep());
processor_elem.plan_step_name = processor->getPlanStepName();
processor_elem.plan_step_description = processor->getPlanStepDescription();
processor_elem.plan_group = processor->getQueryPlanStepGroup();
processor_elem.processor_name = processor->getName();
processor_elem.elapsed_us = static_cast<UInt64>(processor->getElapsedNs() / 1000U);
processor_elem.input_wait_elapsed_us = static_cast<UInt64>(processor->getInputWaitElapsedNs() / 1000U);
processor_elem.output_wait_elapsed_us = static_cast<UInt64>(processor->getOutputWaitElapsedNs() / 1000U);
auto stats = processor->getProcessorDataStats();
processor_elem.input_rows = stats.input_rows;
processor_elem.input_bytes = stats.input_bytes;
processor_elem.output_rows = stats.output_rows;
processor_elem.output_bytes = stats.output_bytes;
processors_profile_log->add(processor_elem);
}
}
}
logProcessorProfile(context, query_pipeline.getProcessors());
logQueryMetricLogFinish(context, internal, elem.client_info.current_query_id, time_now, std::make_shared<QueryStatusInfo>(info));
}

2
src/Planner/PlannerActionsVisitor.cpp
View File

@ -804,7 +804,7 @@ PlannerActionsVisitorImpl::NodeNameAndNodeMinLevel PlannerActionsVisitorImpl::vi
auto lambda_node_name = calculateActionNodeName(node, *planner_context);
auto function_capture = std::make_shared<FunctionCaptureOverloadResolver>(
lambda_actions, captured_column_names, lambda_arguments_names_and_types, lambda_node.getExpression()->getResultType(), lambda_expression_node_name);
lambda_actions, captured_column_names, lambda_arguments_names_and_types, lambda_node.getExpression()->getResultType(), lambda_expression_node_name, true);
// TODO: Pass IFunctionBase here not FunctionCaptureOverloadResolver.
const auto * actions_node = actions_stack[level].addFunctionIfNecessary(lambda_node_name, std::move(lambda_children), function_capture);

24
src/Processors/Formats/Impl/NativeORCBlockInputFormat.cpp
View File

@ -1534,15 +1534,23 @@ static ColumnWithTypeAndName readColumnWithDateData(
for (size_t i = 0; i < orc_int_column->numElements; ++i)
{
Int32 days_num = static_cast<Int32>(orc_int_column->data[i]);
if (check_date_range && (days_num > DATE_LUT_MAX_EXTEND_DAY_NUM || days_num < -DAYNUM_OFFSET_EPOCH))
throw Exception(
ErrorCodes::VALUE_IS_OUT_OF_RANGE_OF_DATA_TYPE,
"Input value {} of a column \"{}\" exceeds the range of type Date32",
days_num,
column_name);
if (!orc_int_column->hasNulls || orc_int_column->notNull[i])
{
Int32 days_num = static_cast<Int32>(orc_int_column->data[i]);
if (check_date_range && (days_num > DATE_LUT_MAX_EXTEND_DAY_NUM || days_num < -DAYNUM_OFFSET_EPOCH))
throw Exception(
ErrorCodes::VALUE_IS_OUT_OF_RANGE_OF_DATA_TYPE,
"Input value {} of a column \"{}\" exceeds the range of type Date32",
days_num,
column_name);
column_data.push_back(days_num);
column_data.push_back(days_num);
}
else
{
/// ORC library doesn't guarantee that orc_int_column->data[i] is initialized to zero when orc_int_column->notNull[i] is false since https://github.com/ClickHouse/ClickHouse/pull/69473
column_data.push_back(0);
}
}
return {std::move(internal_column), internal_type, column_name};

94
src/Processors/Formats/Impl/PrettyBlockOutputFormat.cpp
View File

@ -1,5 +1,6 @@
#include <Processors/Formats/Impl/PrettyBlockOutputFormat.h>
#include <Formats/FormatFactory.h>
#include <Formats/PrettyFormatHelpers.h>
#include <IO/WriteBuffer.h>
#include <IO/WriteHelpers.h>
#include <IO/WriteBufferFromString.h>
@ -352,7 +353,8 @@ void PrettyBlockOutputFormat::writeChunk(const Chunk & chunk, PortKind port_kind
}
writeCString(grid_symbols.bar, out);
writeReadableNumberTip(chunk);
if (readable_number_tip)
writeReadableNumberTipIfSingleValue(out, chunk, format_settings, color);
writeCString("\n", out);
}
@ -392,72 +394,6 @@ void PrettyBlockOutputFormat::writeChunk(const Chunk & chunk, PortKind port_kind
}
static String highlightDigitGroups(String source)
{
if (source.size() <= 4)
return source;
bool is_regular_number = true;
size_t num_digits_before_decimal = 0;
for (auto c : source)
{
if (c == '-' || c == ' ')
continue;
if (c == '.')
break;
if (c >= '0' && c <= '9')
{
++num_digits_before_decimal;
}
else
{
is_regular_number = false;
break;
}
}
if (!is_regular_number || num_digits_before_decimal <= 4)
return source;
String result;
size_t size = source.size();
result.reserve(2 * size);
bool before_decimal = true;
size_t digit_num = 0;
for (size_t i = 0; i < size; ++i)
{
auto c = source[i];
if (before_decimal && c >= '0' && c <= '9')
{
++digit_num;
size_t offset = num_digits_before_decimal - digit_num;
if (offset && offset % 3 == 0)
{
result += "\033[4m";
result += c;
result += "\033[0m";
}
else
{
result += c;
}
}
else if (c == '.')
{
before_decimal = false;
result += c;
}
else
{
result += c;
}
}
return result;
}
void PrettyBlockOutputFormat::writeValueWithPadding(
const IColumn & column, const ISerialization & serialization, size_t row_num,
size_t value_width, size_t pad_to_width, size_t cut_to_width, bool align_right, bool is_number)
@ -553,30 +489,6 @@ void PrettyBlockOutputFormat::writeSuffix()
}
}
void PrettyBlockOutputFormat::writeReadableNumberTip(const Chunk & chunk)
{
const auto & columns = chunk.getColumns();
auto is_single_number = readable_number_tip && chunk.getNumRows() == 1 && chunk.getNumColumns() == 1;
if (!is_single_number)
return;
if (columns[0]->isNullAt(0))
return;
auto value = columns[0]->getFloat64(0);
auto threshold = format_settings.pretty.output_format_pretty_single_large_number_tip_threshold;
if (threshold && isFinite(value) && abs(value) > threshold)
{
if (color)
writeCString("\033[90m", out);
writeCString(" -- ", out);
formatReadableQuantity(value, out, 2);
if (color)
writeCString("\033[0m", out);
}
}
void registerOutputFormatPretty(FormatFactory & factory)
{
registerPrettyFormatWithNoEscapesAndMonoBlock<PrettyBlockOutputFormat>(factory, "Pretty");

3
src/Processors/Formats/Impl/PrettyBlockOutputFormat.h
View File

@ -38,7 +38,6 @@ protected:
virtual void writeChunk(const Chunk & chunk, PortKind port_kind);
void writeMonoChunkIfNeeded();
void writeSuffix() override;
void writeReadableNumberTip(const Chunk & chunk);
void onRowsReadBeforeUpdate() override { total_rows = getRowsReadBefore(); }
@ -56,9 +55,9 @@ protected:
}
bool color;
bool readable_number_tip = false;
private:
bool readable_number_tip = false;
bool mono_block;
/// For mono_block == true only
Chunk mono_chunk;

4
src/Processors/Formats/Impl/PrettyCompactBlockOutputFormat.cpp
View File

@ -4,6 +4,7 @@
#include <IO/WriteHelpers.h>
#include <IO/Operators.h>
#include <Formats/FormatFactory.h>
#include <Formats/PrettyFormatHelpers.h>
#include <Processors/Formats/Impl/PrettyCompactBlockOutputFormat.h>
@ -189,7 +190,8 @@ void PrettyCompactBlockOutputFormat::writeRow(
}
writeCString(grid_symbols.bar, out);
writeReadableNumberTip(chunk);
if (readable_number_tip)
writeReadableNumberTipIfSingleValue(out, chunk, format_settings, color);
writeCString("\n", out);
}

4
src/Processors/Formats/Impl/PrettySpaceBlockOutputFormat.cpp
View File

@ -1,4 +1,5 @@
#include <Formats/FormatFactory.h>
#include <Formats/PrettyFormatHelpers.h>
#include <IO/WriteBuffer.h>
#include <IO/WriteHelpers.h>
#include <Processors/Formats/Impl/PrettySpaceBlockOutputFormat.h>
@ -102,7 +103,8 @@ void PrettySpaceBlockOutputFormat::writeChunk(const Chunk & chunk, PortKind port
writeValueWithPadding(
*columns[column], *serializations[column], row, cur_width, max_widths[column], cut_to_width, type.shouldAlignRightInPrettyFormats(), isNumber(type));
}
writeReadableNumberTip(chunk);
if (readable_number_tip)
writeReadableNumberTipIfSingleValue(out, chunk, format_settings, color);
writeChar('\n', out);
}

28
src/Processors/Formats/Impl/VerticalRowOutputFormat.cpp
View File

@ -4,7 +4,10 @@
#include <IO/WriteBufferFromString.h>
#include <Processors/Formats/Impl/VerticalRowOutputFormat.h>
#include <Formats/FormatFactory.h>
#include <Formats/PrettyFormatHelpers.h>
#include <Common/UTF8Helpers.h>
#include <DataTypes/DataTypeLowCardinality.h>
#include <DataTypes/DataTypeNullable.h>
namespace DB
@ -14,6 +17,8 @@ VerticalRowOutputFormat::VerticalRowOutputFormat(
WriteBuffer & out_, const Block & header_, const FormatSettings & format_settings_)
: IRowOutputFormat(header_, out_), format_settings(format_settings_)
{
color = format_settings.pretty.color == 1 || (format_settings.pretty.color == 2 && format_settings.is_writing_to_terminal);
const auto & sample = getPort(PortKind::Main).getHeader();
size_t columns = sample.columns();
@ -31,6 +36,7 @@ VerticalRowOutputFormat::VerticalRowOutputFormat(
}
names_and_paddings.resize(columns);
is_number.resize(columns);
for (size_t i = 0; i < columns; ++i)
{
WriteBufferFromString buf(names_and_paddings[i]);
@ -42,6 +48,7 @@ VerticalRowOutputFormat::VerticalRowOutputFormat(
{
size_t new_size = max_name_width - name_widths[i] + names_and_paddings[i].size();
names_and_paddings[i].resize(new_size, ' ');
is_number[i] = isNumber(removeNullable(recursiveRemoveLowCardinality(sample.getByPosition(i).type)));
}
}
@ -61,7 +68,26 @@ void VerticalRowOutputFormat::writeField(const IColumn & column, const ISerializ
void VerticalRowOutputFormat::writeValue(const IColumn & column, const ISerialization & serialization, size_t row_num) const
{
serialization.serializeText(column, row_num, out, format_settings);
if (color && format_settings.pretty.highlight_digit_groups && is_number[field_number])
{
String serialized_value;
{
WriteBufferFromString buf(serialized_value);
serialization.serializeText(column, row_num, buf, format_settings);
}
/// Highlight groups of thousands.
serialized_value = highlightDigitGroups(serialized_value);
out.write(serialized_value.data(), serialized_value.size());
}
else
{
serialization.serializeText(column, row_num, out, format_settings);
}
/// Write a tip.
if (is_number[field_number])
writeReadableNumberTip(out, column, row_num, format_settings, color);
}

3
src/Processors/Formats/Impl/VerticalRowOutputFormat.h
View File

@ -56,6 +56,9 @@ private:
using NamesAndPaddings = std::vector<String>;
NamesAndPaddings names_and_paddings;
std::vector<UInt8> is_number;
bool color;
};
}

2
src/Processors/Merges/Algorithms/IMergingAlgorithmWithDelayedChunk.cpp
View File

@ -1,5 +1,5 @@
#include <Processors/Merges/Algorithms/IMergingAlgorithmWithDelayedChunk.h>
#include <Processors/Merges/Algorithms/MergeTreePartLevelInfo.h>
#include <Processors/Merges/Algorithms/MergeTreeReadInfo.h>
namespace DB

2
src/Processors/Merges/Algorithms/IMergingAlgorithmWithSharedChunks.cpp
View File

@ -1,5 +1,5 @@
#include <Processors/Merges/Algorithms/IMergingAlgorithmWithSharedChunks.h>
#include <Processors/Merges/Algorithms/MergeTreePartLevelInfo.h>
#include <Processors/Merges/Algorithms/MergeTreeReadInfo.h>
namespace DB
{

29
src/Processors/Merges/Algorithms/MergeTreePartLevelInfo.h
View File

@ -1,29 +0,0 @@
#pragma once
#include <Processors/Chunk.h>
namespace DB
{
/// To carry part level if chunk is produced by a merge tree source
class MergeTreePartLevelInfo : public ChunkInfoCloneable<MergeTreePartLevelInfo>
{
public:
MergeTreePartLevelInfo() = delete;
explicit MergeTreePartLevelInfo(ssize_t part_level)
: origin_merge_tree_part_level(part_level)
{ }
MergeTreePartLevelInfo(const MergeTreePartLevelInfo & other) = default;
size_t origin_merge_tree_part_level = 0;
};
inline size_t getPartLevelFromChunk(const Chunk & chunk)
{
const auto part_level_info = chunk.getChunkInfos().get<MergeTreePartLevelInfo>();
if (part_level_info)
return part_level_info->origin_merge_tree_part_level;
return 0;
}
}

86
src/Processors/Merges/Algorithms/MergeTreeReadInfo.h Normal file
View File

@ -0,0 +1,86 @@
#pragma once
#include <Processors/Chunk.h>
#include <Core/Block.h>
#include <Interpreters/ExpressionActions.h>
namespace DB
{
namespace ErrorCodes
{
extern const int LOGICAL_ERROR;
}
/// To carry part level and virtual row if chunk is produced by a merge tree source
class MergeTreeReadInfo : public ChunkInfoCloneable<MergeTreeReadInfo>
{
public:
MergeTreeReadInfo() = delete;
explicit MergeTreeReadInfo(size_t part_level) :
origin_merge_tree_part_level(part_level) {}
explicit MergeTreeReadInfo(size_t part_level, const Block & pk_block_, ExpressionActionsPtr virtual_row_conversions_) :
origin_merge_tree_part_level(part_level), pk_block(pk_block_), virtual_row_conversions(std::move(virtual_row_conversions_)) {}
MergeTreeReadInfo(const MergeTreeReadInfo & other) = default;
size_t origin_merge_tree_part_level = 0;
/// If is virtual_row, block should not be empty.
Block pk_block;
ExpressionActionsPtr virtual_row_conversions;
};
inline size_t getPartLevelFromChunk(const Chunk & chunk)
{
const auto read_info = chunk.getChunkInfos().get<MergeTreeReadInfo>();
if (read_info)
return read_info->origin_merge_tree_part_level;
return 0;
}
inline bool isVirtualRow(const Chunk & chunk)
{
const auto read_info = chunk.getChunkInfos().get<MergeTreeReadInfo>();
if (read_info)
return read_info->pk_block.columns() > 0;
return false;
}
inline void setVirtualRow(Chunk & chunk, const Block & header, bool apply_virtual_row_conversions)
{
auto read_info = chunk.getChunkInfos().get<MergeTreeReadInfo>();
chassert(read_info);
Block & pk_block = read_info->pk_block;
// std::cerr << apply_virtual_row_conversions << std::endl;
// std::cerr << read_info->virtual_row_conversions->dumpActions() << std::endl;
if (apply_virtual_row_conversions)
read_info->virtual_row_conversions->execute(pk_block);
// std::cerr << "++++" << pk_block.dumpStructure() << std::endl;
Columns ordered_columns;
ordered_columns.reserve(pk_block.columns());
for (size_t i = 0; i < header.columns(); ++i)
{
const ColumnWithTypeAndName & col = header.getByPosition(i);
if (const auto * pk_col = pk_block.findByName(col.name))
{
if (!col.type->equals(*pk_col->type))
throw Exception(ErrorCodes::LOGICAL_ERROR,
"Virtual row has different type for {}. Expected {}, got {}",
col.name, col.dumpStructure(), pk_col->dumpStructure());
ordered_columns.push_back(pk_col->column);
}
else
ordered_columns.push_back(col.type->createColumnConstWithDefaultValue(1));
}
chunk.setColumns(ordered_columns, 1);
}
}

14
src/Processors/Merges/Algorithms/MergingSortedAlgorithm.cpp
View File

@ -1,3 +1,4 @@
#include <Processors/Merges/Algorithms/MergeTreeReadInfo.h>
#include <Processors/Merges/Algorithms/MergingSortedAlgorithm.h>
#include <Processors/Transforms/ColumnGathererTransform.h>
#include <IO/WriteBuffer.h>
@ -16,12 +17,14 @@ MergingSortedAlgorithm::MergingSortedAlgorithm(
SortingQueueStrategy sorting_queue_strategy_,
UInt64 limit_,
WriteBuffer * out_row_sources_buf_,
bool use_average_block_sizes)
bool use_average_block_sizes,
bool apply_virtual_row_conversions_)
: header(std::move(header_))
, merged_data(use_average_block_sizes, max_block_size_, max_block_size_bytes_)
, description(description_)
, limit(limit_)
, out_row_sources_buf(out_row_sources_buf_)
, apply_virtual_row_conversions(apply_virtual_row_conversions_)
, current_inputs(num_inputs)
, sorting_queue_strategy(sorting_queue_strategy_)
, cursors(num_inputs)
@ -49,6 +52,15 @@ void MergingSortedAlgorithm::addInput()
void MergingSortedAlgorithm::initialize(Inputs inputs)
{
for (auto & input : inputs)
{
if (!isVirtualRow(input.chunk))
continue;
setVirtualRow(input.chunk, header, apply_virtual_row_conversions);
input.skip_last_row = true;
}
removeConstAndSparse(inputs);
merged_data.initialize(header, inputs);
current_inputs = std::move(inputs);

Some files were not shown because too many files have changed in this diff Show More