ARC Core User Guide

ProcessCore is the in-memory F# library for building, decorating, reading, querying, and editing ARC Core process graphs and ARC package metadata.

The ARC Core model is intentionally compact:

• ARC is the top-level package wrapper around a dataset graph and its administrative metadata.
• A Dataset groups process graphs, nested datasets, data files, data contexts, and administrative information.
• A Process connects Sample and Data input/output nodes.
• A Recipe describes what the process executes.
• An Annotation annotates datasets, process parameters, input/output nodes, and protocol components.
• A DataContext describes data files and selected data fragments with additional structural and contextual information.
• Agent, Organization, and ScholarlyArticle carry administrative profile metadata, while ARC adds package-level persistence, project-configured storage, and metadata handling.
• additionalType, additionalProperty, and DynamicObj carry domain-specific extensions without changing the shared graph shape.

The pages in this section are user-facing implementation guides.

For normative field definitions, use the specification index. These project pages focus on using the F# library.

For API reference, see the API docs.

Loading the library

The ProcessCore library is available as a NuGet package for .NET and currently a NuGet package for JavaScript transpilation.

.NET

dotnet add package ProcessCore

#r "nuget: ProcessCore" // in scripts

open ProcessCore

let d = Dataset("MyDataset")

JavaScript

dotnet add package ProcessCore.Javascript

#r "nuget: ProcessCore.Javascript" // in scripts

open ProcessCore

let d = Dataset("MyDataset")

Recommended Path

1. ARC Layer shows package-level metadata, explicit YAML/XLSX persistence, and .arc/project.yml-configured workspaces.
2. Creating A Dataset builds a graph from F# objects.
3. Decorations shows how to add domain specificity through typed annotations and DynamicObj overflow fields.
4. Reading And Writing YAML loads profile-shaped examples and writes inline or indexed YAML.
5. Querying Process Graphs traverses upstream, downstream, and connected context.
6. Fragment Selector Providers makes file fragments first-class in traversal.
7. Using DataContext describes Datamap entries and shows how to combine them with process annotations.
8. Tabular Views edits process graphs through ISA-like table projections.
9. Graph Identity, Back-Edges, And Scope explains the invariants behind shared nodes and scoped traversal.

What To Use When