Linked Data Store: documentation
Linked Data Store (LDS) is a server that gives REST and GraphQL access to versioned, linked JSON documents. The model of the documents is read when LDS starts. LDS is a time machine: every version of a document holds from its valid time, and an as-of read gives the document as it was at the instant asked for.
This documentation is in two parts, for two kinds of reader, and a glossary of the terms both use. Every document is written from the code as it is, and every request shown in it was run. LDS 3.0 is in development and 3.0.0 is not released: LDS is built from the repository, where its milestones are tagged. What is unfinished is said where a reader meets it, in a paragraph that starts with "Not finished:".
For those who use LDS
You run LDS, give it a model, and read and write documents. You do not need to read the code.
| Document | What it is |
|---|---|
| Getting started | From nothing to a first document: start LDS, write a document, read it, read it as it was, link two documents, ask with GraphQL. |
| Concepts | The ideas: document, entity type, namespace, version and valid time, link, model, provider, what a write goes through. |
| The model | How to write a model, as a GraphQL schema or as JSON Schema files, and what LDS checks against it. |
| The REST API | Every address, method, parameter and answer. |
| The GraphQL API | The queries LDS makes from a model, paging, links, as-of reads, search. |
| Configuration | Every key, its default and its meaning, and how keys are set. |
| Running LDS | The image and the compose file, starting and stopping, health, logs, what is stored where. |
| Glossary | The terms of this documentation, each with one meaning. |
For those who develop and maintain LDS
You change the code, add a provider, or look for a fault.
Start with the architecture: the modules, the path of a read and of a write, the ideas that hold the code together and the rules a change must keep. Then one document for each part:
| Document | The part of the code |
|---|---|
| The build | The reactor, the BOM, the rules of the build, the tests, CI. |
| The persistence interface | What a persistence provider implements: fragments, documents, reactive streams, transactions. |
| The shared suite | The tests every persistence provider passes. |
| The memory provider | Persistence in memory, for tests and for trying LDS. |
| The PostgreSQL provider | Persistence in one table of PostgreSQL. |
| The Neo4j provider | Persistence as a graph in Neo4j. |
| Search | The search interface and the Solr provider. |
| Core: the application | How LDS is put together, started and stopped. |
| Core: the model | How a model is read and how a document is checked against it. |
| Core: the REST API | How a request becomes a read or a write of a document, a part of it, or a link. |
| Core: the write path | Sagas, the saga log, the transaction log, recovery. |
| Core: GraphQL | How the GraphQL schema is built and how a query is answered. |
| The server | The main class, the defaults of the configuration, the image, the compose file. |
| The load test | What a run of the load test does and what it is made of. |