Incrementalist is a .NET tool that leverages libgit2sharp and Roslyn to compute incremental build steps, to help reduce total build time, for Continuous Integration systems for large .NET solutions.
Incrementalist is available in one of two distribution methods:
- Incrementalist Library - a stand-alone .NET Standard 2.0 NuGet package that can be called from a program of your own making or
- Incrementalist.Cmd - a
dotnet tool
that can be run directly from thedotnet
CLI. Most users preferIncrementalist.Cmd
for running inside their build systems.
To install Incrementalist and try for yourself, do the following:
dotnet tool install --global Incrementalist.Cmd
The following CLI options are available on Incrementalist, which you can print out at any time via the incrementalist --help
command:
-s, --sln The name of the Solution file to be analyzed by Incrementalist.
-f, --file If specified, writes the output to the named file.
-l, --folders-only List affected folders instead of .NET projects
-b, --branch Required. (Default: dev) The git branch to compare against. i.e. the `dev` or the
`master` branch.
-d, --dir Specify the working directory explicitly. Defaults to using the current working
directory.
--verbose (Default: false) Prints out extensive debug logs during operation.
-t, --timeout (Default: 2) Specifies the load timeout for the solution in whole minutes.
Defaults to 2 minutes.
--help Display this help screen.
--version Display version information.
To run a standard Incrementalist build on a project like Akka.NET, we do the following:
PS> incrementalist -s ./src/Akka.sln -b dev --file ./bin/output/incrementalist.txt
Incrementalist works by analyizing the git diff
of each commit in your working branch, comparing it to a base branch (i.e. dev
) to determine which files have been modified in your changes, and then it uses Roslyn solution analysis to determine the graph of projects that were affected by these changes.
This graph analysis produces a text file that looks like this (when we're running it on the Akka.NET main repository):
D:\a\1\s\src\core\Akka\Akka.csproj,D:\a\1\s\src\benchmark\SerializationBenchmarks\SerializationBenchmarks.csproj
- Each line represents one graph of changes detected by
git
and Roslyn - each project will be listed using its absolute path and will be separated by comma; - If there are multiple lines in the output file, it means that multiple discrete graphs were detected (two sets of projects that don't directly relate to eachother were updated in the same set of
git
commits.)
This file can be parsed and used inside a build script, such as the FAKE file we use for running Akka.NET's build system:
//--------------------------------------------------------------------------------
// Incrementalist targets
//--------------------------------------------------------------------------------
// Pulls the set of all affected projects detected by Incrementalist from the cached file
let getAffectedProjectsTopology =
lazy(
log (sprintf "Checking inside %s for changes" incrementalistReport)
let incrementalistFoundChanges = File.Exists incrementalistReport
log (sprintf "Found changes via Incrementalist? %b - searched inside %s" incrementalistFoundChanges incrementalistReport)
if not incrementalistFoundChanges then None
else
let sortedItems = (File.ReadAllLines incrementalistReport) |> Seq.map (fun x -> (x.Split ','))
|> Seq.map (fun items -> (items.[0], items))
let d = dict sortedItems
Some(d)
)
let getAffectedProjects =
lazy(
let finalProjects = getAffectedProjectsTopology.Value
match finalProjects with
| None -> None
| Some p -> Some (p.Values |> Seq.concat)
)
Target "ComputeIncrementalChanges" (fun _ ->
if runIncrementally then
let targetBranch = match getBuildParam "targetBranch" with
| "" -> "dev"
| null -> "dev"
| b -> b
let incrementalistPath =
let incrementalistDir = toolsDir @@ "incrementalist"
let globalTool = tryFindFileOnPath "incrementalist.exe"
match globalTool with
| Some t -> t
| None -> if isWindows then findToolInSubPath "incrementalist.exe" incrementalistDir
elif isMacOS then incrementalistDir @@ "incrementalist"
else incrementalistDir @@ "incrementalist"
let args = StringBuilder()
|> append "-b"
|> append targetBranch
|> append "-s"
|> append solution
|> append "-f"
|> append incrementalistReport
|> toText
let result = ExecProcess(fun info ->
info.FileName <- incrementalistPath
info.WorkingDirectory <- __SOURCE_DIRECTORY__
info.Arguments <- args) (System.TimeSpan.FromMinutes 5.0) (* Reasonably long-running task. *)
if result <> 0 then failwithf "Incrementalist failed. %s" args
else
log "Skipping Incrementalist - not enabled for this build"
)
let filterProjects selectedProject =
if runIncrementally then
let affectedProjects = getAffectedProjects.Value
(*
if affectedProjects.IsSome then
log (sprintf "Searching for %s inside [%s]" selectedProject (String.Join(",", affectedProjects.Value)))
else
log "No affected projects found"
*)
match affectedProjects with
| None -> None
| Some x when x |> Seq.exists (fun n -> n.Contains (System.IO.Path.GetFileName(string selectedProject))) -> Some selectedProject
| _ -> None
else
log "Not running incrementally"
Some selectedProject
And in each one of our build steps where we want to execute incremental builds only (in order to reduce the amount of work per-pull request), we call the filterProjects
method that uses the Incrementalist.Cmd
output to determine which projects need to be considered based on these changes:
//--------------------------------------------------------------------------------
// Build targets
//--------------------------------------------------------------------------------
let skipBuild =
lazy(
match getAffectedProjects.Value with
| None when runIncrementally -> true
| _ -> false
)
let headProjects =
lazy(
match getAffectedProjectsTopology.Value with
| None when runIncrementally -> [||]
| None -> [|solution|]
| Some p -> p.Keys |> Seq.toArray
)
Target "AssemblyInfo" (fun _ ->
XmlPokeInnerText "./src/common.props" "//Project/PropertyGroup/VersionPrefix" releaseNotes.AssemblyVersion
XmlPokeInnerText "./src/common.props" "//Project/PropertyGroup/PackageReleaseNotes" (releaseNotes.Notes |> String.concat "\n")
)
Target "Build" (fun _ ->
if not skipBuild.Value then
let additionalArgs = if versionSuffix.Length > 0 then [sprintf "/p:VersionSuffix=%s" versionSuffix] else []
let buildProject proj =
DotNetCli.Build
(fun p ->
{ p with
Project = proj
Configuration = configuration
AdditionalArgs = additionalArgs })
match getAffectedProjects.Value with
| Some p -> p |> Seq.iter buildProject
| None -> buildProject solution // build the entire solution if incrementalist is disabled
)
This tool is best used for large .NET projects, where the time to complete each build step can take 30+ minutes. Incrementalist can help reduce the average execution time by an order of magnitude for many pull requests.
To run the build script associated with this solution, execute the following:
Windows
c:\> build.cmd all
Linux / OS X
c:\> build.sh all
If you need any information on the supported commands, please execute the build.[cmd|sh] help
command.
This build script is powered by FAKE; please see their API documentation should you need to make any changes to the build.fsx
file.
The attached build script will automatically do the following based on the conventions of the project names added to this project:
- Any project name ending with
.Tests
will automatically be treated as a XUnit2 project and will be included during the test stages of this build script; - Any project name ending with
.Tests
will automatically be treated as a NBench project and will be included during the test stages of this build script; and - Any project meeting neither of these conventions will be treated as a NuGet packaging target and its
.nupkg
file will automatically be placed in thebin\nuget
folder upon running thebuild.[cmd|sh] all
command.
This solution also supports DocFx for generating both API documentation and articles to describe the behavior, output, and usages of your project.
All of the relevant articles you wish to write should be added to the /docs/articles/
folder and any API documentation you might need will also appear there.
All of the documentation will be statically generated and the output will be placed in the /docs/_site/
folder.
To preview the documentation for this project, execute the following command at the root of this folder:
C:\> serve-docs.cmd
This will use the built-in docfx.console
binary that is installed as part of the NuGet restore process from executing any of the usual build.cmd
or build.sh
steps to preview the fully-rendered documentation. For best results, do this immediately after calling build.cmd buildRelease
.
This project will automatically populate its release notes in all of its modules via the entries written inside RELEASE_NOTES.md
and will automatically update the versions of all assemblies and NuGet packages via the metadata included inside Directory.Build.props
.
This project uses SignService to code-sign NuGet packages prior to publication. The build.cmd
and build.sh
scripts will automatically download the SignClient
needed to execute code signing locally on the build agent, but it's still your responsibility to set up the SignService server per the instructions at the linked repository.
Once you've gone through the ropes of setting up a code-signing server, you'll need to set a few configuration options in your project in order to use the SignClient
:
- Add your Active Directory settings to
appsettings.json
and - Pass in your signature information to the
signingName
,signingDescription
, andsigningUrl
values insidebuild.fsx
.
Whenever you're ready to run code-signing on the NuGet packages published by build.fsx
, execute the following command:
C:\> build.cmd nuget SignClientSecret={your secret} SignClientUser={your username}
This will invoke the SignClient
and actually execute code signing against your .nupkg
files prior to NuGet publication.
If one of these two values isn't provided, the code signing stage will skip itself and simply produce unsigned NuGet code packages.