clickhouse-cpp

You can customize fundamental GoogleTest operations by defining the following macros in the custom directory:

• GTEST_OS_STACK_TRACE_GETTER_: Provide the name of an implementation of OsStackTraceGetterInterface to customize how stack traces are retrieved.
• GTEST_CUSTOM_TEMPDIR_FUNCTION_: Provide an override for testing::TempDir(). The override must match the signature and semantics of testing::TempDir.

When implementing retry logic for clickhouse::Client:

1. If a previous attempt threw an exception, call clickhouse::Client::ResetConnection() before retrying.
2. For clickhouse::Client::Insert(), you can reuse the existing Block from the previous attempt instead of rebuilding it.

Add clickhouse-cpp as a git submodule (e.g., in contrib/clickhouse-cpp) and include it in your CMakeLists.txt. Ensure you set the compatibility flags CH_USE_ABSEIL_FOR_BIGNUM and CH_MAP_BOOL_TO_UINT8 to OFF.

cmake_minimum_required(VERSION 3.13)
project(application-example LANGUAGES CXX)

set(CH_USE_ABSEIL_FOR_BIGNUM OFF)
set(CH_MAP_BOOL_TO_UINT8 OFF)

add_subdirectory(contrib/clickhouse-cpp)

add_executable(application-example app.cpp)
target_link_libraries(application-example PRIVATE clickhouse-cpp-lib)

Use CMake's FetchContent to automatically download and configure the library during the configuration step. Set the compatibility flags to OFF as well.

cmake_minimum_required(VERSION 3.14)
project(application-example LANGUAGES CXX)

include(FetchContent)

set(CH_USE_ABSEIL_FOR_BIGNUM OFF)
set(CH_MAP_BOOL_TO_UINT8 OFF)

FetchContent_Declare(
    clickhouse_cpp
    GIT_REPOSITORY https://github.com/ClickHouse/clickhouse-cpp.git
    GIT_TAG v2.6.2
)
FetchContent_MakeAvailable(clickhouse_cpp)

add_executable(application-example app.cpp)
target_link_libraries(application-example PRIVATE clickhouse-cpp-lib)

To build the library using CMake, use the following recommended settings. It is highly recommended to disable the legacy defaults CH_USE_ABSEIL_FOR_BIGNUM and CH_MAP_BOOL_TO_UINT8 for new projects to ensure future compatibility.

$ mkdir build .
$ cd build
$ cmake .. -DCH_USE_ABSEIL_FOR_BIGNUM=NO -DCH_MAP_BOOL_TO_UINT8=NO
$ make

To manage large datasets without high memory usage, use the batch insertion pattern. This involves starting an insertion with an INSERT statement ending in VALUES, appending data to a Block, sending it via SendInsertBlock, and finally calling EndInsert.

// Start the insertion.
auto block = client->BeginInsert("INSERT INTO foo (id, name) VALUES");

// Grab the columns from the block.
auto col1 = block[0]->As<ColumnUInt64>();
auto col2 = block[1]->As<ColumnString>();

// Add a couple of records to the block.
col1.Append(1);
col1.Append(2);
col2.Append("holden");
col2.Append("naomi");

// Send those records.
block.RefreshRowCount();
client->SendInsertBlock(block);
block.Clear();

// Add another record to the block.
col1.Append(3);

// Send it and finish.
block.RefreshRowCount();
client->EndInsert();

Add clickhouse-cpp to your MODULE.bazel file and depend on @clickhouse-cpp//:clickhouse in your BUILD.bazel.

Note: Bazel support is experimental. By default, it uses BoringSSL for TLS. You can switch to OpenSSL using --@clickhouse-cpp//:tls=openssl or disable TLS with --@clickhouse-cpp//:tls=no.

# MODULE.bazel
bazel_dep(name = "clickhouse-cpp", version = "2.6.2")

# BUILD.bazel
cc_binary(
    name = "application-example",
    srcs = ["app.cpp"],
    deps = ["@clickhouse-cpp//:clickhouse"],
)

Use ch::Client to connect to a host and iterate through results using BeginSelect() and NextBlock(). Data is retrieved as columnar Block objects.

#include <clickhouse/client.h>
#include <iostream>

namespace ch = clickhouse;

int main()
{
    ch::Client client{ch::ClientOptions{}.SetHost("localhost")};

    client.BeginSelect("SELECT 'Hello from ClickHouse :) ");
    
    while (auto block = client.NextBlock()) {
        auto col_msg = block->At(0)->AsStrict<ch::ColumnString>();
        
        for (size_t i = 0; i < block->GetRowCount(); ++i) {
            std::cout << col_msg->At(i) << "\n";
        }
    }
}