51Degrees Device Detection Engines

Pipeline API

Developer Documentation

Introduction

This repository contains the device detection engines for the .NET implementation of the Pipeline API.

The specification is also available on GitHub and is recommended reading if you wish to understand the concepts and design of this API.

Dependencies

Visual Studio 2022 or later is recommended. Although Visual Studio Code can be used for working with most of the projects.

The core device detection projects are written in C and C++. The Pipeline engines are written in C# and target .NET Standard 2.0.3. Example and test projects mostly target .NET 6.0 though in some cases, projects are available targeting other frameworks.

For runtime dependencies, see our dependencies page. The ci/options.json file lists the tested and packaged .NET versions and operating systems automatic tests are performed with. The solution will likely operate with other versions.

Data

The API can either use our cloud service to get its data or it can use a local (on-premise) copy of the data.

Cloud

You will require a resource key to use the Cloud API. You can create resource keys using our configurator, see our documentation on how to use this.

On-Premise

In order to perform device detection on-premise, you will need to use a 51Degrees data file. This repository includes a free, 'lite' file in the 'device-detection-data' sub-module that has a significantly reduced set of properties. To obtain a file with a more complete set of device properties see the 51Degrees website. If you want to use the lite file, you will need to install GitLFS.

On Linux:

sudo apt-get install git-lfs
git lfs install

Then, navigate to 'device-detection-cxx/device-detection-data' and execute:

git lfs pull

Solutions and projects

• FiftyOne.DeviceDetection - Device detection engines and related projects.
  • FiftyOne.DeviceDetection - Contains device detection engine builders.
  • FiftyOne.DeviceDetection.Cloud - A .NET engine which retrieves device detection results by consuming the 51Degrees cloud service. This can be swapped out with either the hash or pattern engines seamlessly.
  • FiftyOne.DeviceDetection.Hash.Engine.OnPremise - .NET implementation of the device detection hash engine. CMake is used to build the native binaries.
  • FiftyOne.DeviceDetection.Shared - Shared classes used by the device detection engines.

Installation

Nuget

The easiest way to install is to use NuGet to add the reference to the package:

Install-Package FiftyOne.DeviceDetection

Build from Source

Device detection on-premise uses a native binary (i.e. compiled from C code to target a specific platform/architecture). The NuGet package contains several binaries for common platforms. However, in some cases, you'll need to build the native binaries yourself for your target platform. This section explains how to do this.

Pre-requisites

• Install C build tools:
  • Windows:
    • You will need either Visual Studio 2022 or the C++ Build Tools installed.
      • Minimum platform toolset version is v143
      • Minimum Windows SDK version is 10.0.18362.0
  • Linux/MacOS:
    • sudo apt-get install g++ make libatomic1
• If you have not already done so, pull the git submodules that contain the native code:
  • git submodule update --init --recursive

Visual studio should now be able to build the native binaries as part of its normal build process.

Packaging

You can package a project into NuGet *.nupkg file by running a command like:

dotnet pack [Project] -o "[PackagesFolder]" /p:PackageVersion=0.0.0 -c [Configuration] /p:Platform=[Architecture]

⚠️ Notes on packaging FiftyOne.DeviceDetection.Hash.Engine.OnPremise

📝 Using AnyCPU might prevent the unmanaged (C++) code from being built into .Native.dll library. Use x86/x64/arm64 specifically.

📝 If creating cross-platform package from multiple native dlls, put all 6x FiftyOne.DeviceDetection.Hash.Engine.OnPremise.Native.dll into respective folders:

../
    macos/
        arm64/
        x64/
    linux/
        x64/
        x86/
    windows/
        x64/
        x86/

and add to the packaging command:

/p:BuiltOnCI=true

related CI scripts:

• BuiltOnCI var:
  • [https://github.com/51Degrees/common-ci/blob/main/dotnet/build-project-core.ps1]
  • [https://github.com/51Degrees/common-ci/blob/main/dotnet/build-package-nuget.ps1]
  • [https://github.com/51Degrees/common-ci/blob/main/dotnet/build-project-framework.ps1]
  • [https://github.com/51Degrees/device-detection-dotnet/blob/main/ci/run-performance-tests-console.ps1]
• Copying native binaries:
  • [https://github.com/51Degrees/device-detection-dotnet/blob/main/ci/build-package.ps1]

Strong naming

We currently do not strong name assemblies due to downsides for developers. The main of which is that .NET Framework on Windows enables strict loading of assemblies once an assembly is strong named. A strong-named assembly reference must exactly match the version of the loaded assembly, forcing developers to configure binding redirects when using the assembly.

If it is absolutely critical for your use case to integrate a strong-named assembly - please create a feature request issue.

Examples

Examples can be found in device-detection-dotnet-examples repository.

Tests

Tests can be found in the Tests/ folder. These can all be run from within Visual Studio or by using the dotnet test command line tool.

Some tests require additional resources to run. These will either fail or return an 'inconclusive' result if these resources are not provided.

• Some tests require an 'Enterprise' data file. This can be obtained by purchasing a license.
  • Once available, the full path to this data file must be specified in the DEVICEDETECTIONDATAFILE environment variable.
• Tests using the cloud service require resource keys with specific properties to be provided using environment variables:
  • The SUPER_RESOURCE_KEY environment variable should be populated with a key that includes all properties. A license is required in order to access some properties.

Project documentation

For complete documentation on the Pipeline API and associated engines, see the 51Degrees documentation site.