tdlight/README.md
2021-07-27 23:35:50 +02:00

15 KiB

TDLight

TDLight is a fork of tdlib, focused on memory footprint and performance.

TDLight is 100% compatible with tdlib, if you don't use the sqlite database.

⚠️ Memory Cleanup remove nearly every cached value, so you must use it with caution! This function is not suggested for GUI clients, because they heavily rely on retrieving cached data! TDLib developers strongly advise against the use of this feature, since it is not an intended behavior.

Added features

Memory cleanup

TDLight can clean itself and release some ram to the OS if you want. Look at TdApi.OptimizeMemory in "Modified features" paragraph to see how.

Constant memory usage without restarting

TDLight, if used with care, doesn't grow in memory usage with time. Look at TdApi.OptimizeMemory in "Modified features" paragraph to see how

memory usage

Custom options

We added some options:

  • disable_minithumbnails (true/false) This setting removes minithumbnails everywhere. It reduces memory usage because tdlib keeps them in RAM
  • disable_document_filenames (true/false) If you don't care about having the original filenames of every file stored in RAM, you can disable them using this option. It reduces memory usage
  • disable_notifications (true/false) In TDLib pending notification updates are stored in ram until you "read" them. This option disables completely notifications and keeps the pending notifications queue empty, reducing memory usage
  • ignore_update_chat_last_message (true/false) If you don't care about have updateChatLastMessage updates enable this
  • ignore_update_chat_read_inbox (true/false) If you don't care about have updateChatReadInbox updates enable this
  • ignore_update_user_chat_action (true/false) If you don't care about have updateUserChatAction updates enable this
  • ignore_server_deletes_and_reads (true/false) If you don't care about receiving read receipts and remote deletes from other users, enable this, it will reduce memory usage
  • delete_chat_reference_after_seconds (positive number) During cleanup, free the memory of the chats that have not been touched for more than X seconds
  • delete_user_reference_after_seconds (positive number) During cleanup, free the memory of the users that have not been touched for more than X seconds
  • delete_file_reference_after_seconds (positive number) During cleanup, free the memory of the files that have not been touched for more than X seconds
  • experiment_enable_file_reference_cleanup (true/false) During cleanup, free the memory of the file references
  • experiment_enable_chat_access_hash_cleanup (true/false) During cleanup, clean chats and channels access hash
  • enable_pull_based_backpressure (true/false) Enable manual get_channel_difference execution by calling getChannelDifference(channel_difference_id). Don't modify this option unless you have a very large bot that struggles to keep up with start-up updates throughput, or you want to implement a pull-based async library.

Custom API functions

TdApi.OptimizeMemory

This method is used to optimize the memory usage, but it must be used carefully. It removes almost all cached values and releases the memory back to the OS.

You can call TdApi.OptimizeMemory normally, but removing cached values can cause problems if you don't take some precautions.

If you want to avoid receiving data with missing fields during cleanup:

  1. Before calling TdApi.OptimizeMemory you must:
    1. Read all the pending updates to empty the pending updates queue.
    2. Disable internet connection using TdApi.SetNetworkType(TdApi.NetworkTypeNone)
  2. Call TdApi.OptimizeMemory
  3. After calling TdApi.OptimizeMemory you must:
    1. NOT use the old file ids because they have been deleted! (Example: If you receive the file 12 after OptimizeMemory is not the same file 12 that you received before TdApi.OptimizeMemory, because the id 12 has been reused)
    2. Re-enable internet connection using TdApi.SetNetworkType(TdApi.NetworkTypeOther)

TdApi.GetMemoryStatistics

This method is used to read the size of all the biggest data maps inside tdlib implementation. The output contains a string that can be parsed as a JSON.

Other reccomended options

  • Options:

    • ignore_inline_thumbnails: true
    • disable_top_chats: true
    • ignore_platform_restrictions: true
    • ignore_sensitive_content_restrictions: true
  • Disable all the databases (messages_db, users_db, files_db)

    ⚠️ If you use the databases, TDLight memory cleanup feature will be automatically disabled, because databases will lose some data when cleaning up the memory.


The following text is the classic tdlib readme:

TDLib

TDLib (Telegram Database library) is a cross-platform library for building Telegram clients. It can be easily used from almost any programming language.

Table of Contents

Features

TDLib has many advantages. Notably TDLib is:

  • Cross-platform: TDLib can be used on Android, iOS, Windows, macOS, Linux, FreeBSD, OpenBSD, NetBSD, illumos, Windows Phone, WebAssembly, watchOS, tvOS, Tizen, Cygwin. It should also work on other *nix systems with or without minimal effort.
  • Multilanguage: TDLib can be easily used with any programming language that is able to execute C functions. Additionally it already has native Java (using JNI) bindings and .NET (using C++/CLI and C++/CX) bindings.
  • Easy to use: TDLib takes care of all network implementation details, encryption and local data storage.
  • High-performance: in the Telegram Bot API, each TDLib instance handles more than 24000 active bots simultaneously.
  • Well-documented: all TDLib API methods and public interfaces are fully documented.
  • Consistent: TDLib guarantees that all updates are delivered in the right order.
  • Reliable: TDLib remains stable on slow and unreliable Internet connections.
  • Secure: all local data is encrypted using a user-provided encryption key.
  • Fully-asynchronous: requests to TDLib don't block each other or anything else, responses are sent when they are available.

Examples and documentation

See our Getting Started tutorial for a description of basic TDLib concepts.

Take a look at our examples.

See a TDLib build instructions generator for detailed instructions on how to build TDLib.

See description of our JSON, C++, Java and .NET interfaces.

See the td_api.tl scheme or the automatically generated HTML documentation for a list of all available TDLib methods and classes.

Dependencies

TDLib depends on:

  • C++14 compatible compiler (Clang 3.4+, GCC 4.9+, MSVC 19.0+ (Visual Studio 2015+), Intel C++ Compiler 17+)
  • OpenSSL
  • zlib
  • gperf (build only)
  • CMake (3.0.2+, build only)
  • PHP (optional, for documentation generation)

Building

The simplest way to build TDLib is to use our TDLib build instructions generator. You need only to choose your programming language and target operating system to receive complete build instructions.

In general, you need to install all TDLib dependencies as described in Installing dependencies. Then enter directory containing TDLib sources and compile them using CMake:

mkdir build
cd build
cmake -DCMAKE_BUILD_TYPE=Release ..
cmake --build .

To build TDLib on low memory devices you can run SplitSource.php script before compiling main TDLib source code and compile only needed targets:

mkdir build
cd build
cmake -DCMAKE_BUILD_TYPE=Release ..
cmake --build . --target prepare_cross_compiling
cd ..
php SplitSource.php
cd build
cmake --build . --target tdjson
cmake --build . --target tdjson_static
cd ..
php SplitSource.php --undo

In our tests clang 6.0 with libc++ required less than 500 MB of RAM per file and GCC 4.9/6.3 used less than 1 GB of RAM per file.

Installing dependencies

macOS

  • Install the latest Xcode command line tools, for example, via xcode-select --install.
  • Install other dependencies, for example, using Homebrew:
brew install gperf cmake openssl
  • Build TDLib with CMake as explained in building. You will likely need to manually specify path to the installed OpenSSL to CMake, e.g.,
cmake -DCMAKE_BUILD_TYPE=Release -DOPENSSL_ROOT_DIR=/usr/local/opt/openssl/ ..

Windows

  • Download and install Microsoft Visual Studio 2015 or later.
  • Download and install gperf. Add the path to gperf.exe to the PATH environment variable.
  • Install vcpkg.
  • Run the following commands to install TDLib dependencies using vcpkg:
cd <path to vcpkg>
.\vcpkg.exe install openssl:x64-windows openssl:x86-windows zlib:x64-windows zlib:x86-windows
  • Download and install CMake; choose "Add CMake to the system PATH" option while installing.
  • Build TDLib with CMake as explained in building, but instead of cmake -DCMAKE_BUILD_TYPE=Release .. use
cmake -DCMAKE_TOOLCHAIN_FILE=<path to vcpkg>/scripts/buildsystems/vcpkg.cmake ..

To build 32-bit/64-bit TDLib using MSVC, you will need to additionally specify parameter -A Win32/-A x64 to CMake. To build TDLib in Release mode using MSVC, you will need to additionally specify parameter --config Release to the cmake --build . command.

Linux

Using in CMake C++ projects

For C++ projects that use CMake, the best approach is to build TDLib as part of your project or to install it system-wide.

There are several libraries that you could use in your CMake project:

  • Td::TdJson, Td::TdJsonStatic — dynamic and static version of a JSON interface. This has a simple C interface, so it can be easily used with any programming language that is able to execute C functions. See td_json_client and td_log documentation for more information.
  • Td::TdStatic — static library with C++ interface for general usage. See Client and Log documentation for more information.
  • Td::TdCoreStatic — static library with low-level C++ interface intended mostly for internal usage. See ClientActor and Log documentation for more information.

For example, part of your CMakeLists.txt may look like this:

add_subdirectory(td)
target_link_libraries(YourTarget PRIVATE Td::TdStatic)

Or you could install TDLib and then reference it in your CMakeLists.txt like this:

find_package(Td 1.7.6 REQUIRED)
target_link_libraries(YourTarget PRIVATE Td::TdStatic)

See example/cpp/CMakeLists.txt.

Using in Java projects

TDLib provides native Java interface through JNI. To enable it, specify option -DTD_ENABLE_JNI=ON to CMake.

See example/java for example of using TDLib from Java and detailed build and usage instructions.

Using in .NET projects

TDLib provides native .NET interface through C++/CLI and C++/CX. To enable it, specify option -DTD_ENABLE_DOTNET=ON to CMake. .NET Core supports C++/CLI only since version 3.1 and only on Windows, so if older .NET Core is used or portability is needed, then TDLib JSON interface should be used through P/Invoke instead.

See example/csharp for example of using TDLib from C# and detailed build and usage instructions. See example/uwp for example of using TDLib from C# UWP application and detailed build and usage instructions for Visual Studio Extension "TDLib for Universal Windows Platform".

When TDLib is built with TD_ENABLE_DOTNET option enabled, C++ documentation is removed from some files. You need to checkout these files to return C++ documentation back:

git checkout td/telegram/Client.h td/telegram/Log.h td/tl/TlObject.h

Using from other programming languages

TDLib provides efficient native C++, Java, and .NET interfaces. But for most use cases we suggest to use the JSON interface, which can be easily used with any programming language that is able to execute C functions. See td_json_client and td_log documentation for detailed JSON interface description, the td_api.tl scheme or the automatically generated HTML documentation for a list of all available TDLib methods and classes.

TDLib JSON interface adheres to semantic versioning and versions with the same major version number are binary and backward compatible, but the underlying TDLib API can be different for different minor and even patch versions. If you need to support different TDLib versions, then you can use a value of the version option to find exact TDLib version to use appropriate API methods.

See example/python/tdjson_example.py for an example of such usage.

License

TDLib is licensed under the terms of the Boost Software License. See LICENSE_1_0.txt for more information.