JuliaWorkspaces.jl

The analysis engine that powers LanguageServer.jl. See also LintApp.jl a standalone command-line application that using JuliaWorkspaces to do linting.

JuliaWorkspaces.jl takes a set of files — Julia sources, Project.toml / Manifest.toml, configuration files — and answers questions about them: diagnostics, hover text, completions, go-to-definition, references, document and workspace symbols, signature help, formatting, code actions, test items, and more. The same engine can back an interactive language server, a CI linter, or a command-line tool.

Internally it is an incremental, memoized query system built on Salsa.jl: mutable inputs (files, the active project, background-indexing results) feed a graph of pure, cached derived queries, so changing one file only recomputes what depends on it.

Installation

using Pkg
Pkg.add(url="https://github.com/julia-vscode/JuliaWorkspaces.jl")

Quick start

using JuliaWorkspaces

# Build a workspace from one or more folders on disc.
jw = workspace_from_folders(["/path/to/my/project"])

# Query diagnostics for the whole workspace.
for (uri, file_diags) in get_diagnostics(jw)
    for d in file_diags
        println("$(uri): [$(d.severity)] $(d.message)")
    end
end

Documentation

Full documentation is hosted at https://julia-vscode.github.io/JuliaWorkspaces.jl/dev. Start with the architecture page if you intend to work on the package.

Repository scripts

Script	Purpose
scripts/update_vendored_packages.jl	Reconcile the vendored packages/ trees with scripts/vendored_packages.jl.
scripts/install_julia_versions.jl	Install every supported Julia via juliaup.
scripts/update_app_environments.jl	Regenerate the per-version dynamic analysis process environments.

scripts/vendored_packages.jl is the list the first of those works from: add a package there and run it with --apply to vendor it. Nothing records which version is vendored — that is read back from each tree's own Project.toml, and --verify audits it against the commit git subtree recorded.