A Java implementation of a parser for the KDL Document Language.

This project is beta-quality. It's been extensively tested, but the spec it implements is still in flux.

final KDLParser parser = new KDLParser();

final KDLDocument documentFromString = parser.parse("node_name \"arg\"")
// OR
final KDLDocument documentFromReader = parser.parse(new FileReader("some/file.kdl"))

KDLDocument objects, and all descendants of KDLObject, are immutable and threadsafe, though that is not true of their Builder objects. If you need to make changes to a KDLDocument, use the filter() and mutate() functions explained below.

Several utilities are provided for finding nodes in documents. Each presents the same interface, but the way they search the document differs. There are three search types:

• RootSearch - Searches entirely at the root, primarily used for mutations to the root as discussed below
• GeneralSearch - Searches for nodes anywhere in the tree matching a single, possibly compound, node predicate
• PathedSearch - Searches for nodes down a specified path. At each level a different node predicate can be specified

Each provides four methods for searching or mutating documents:

• anyMatch(document) - Returns true if any node matches the search, false otherwise
• filter(document, trim) - Removes all nodes from the tree not on a branch that matches the predicates of the search. if trim is set, removes all their non-matching children
• list(document, trim) - Produces a new document with all matching nodes at the root. If trim is set, removes all their non-matching children
• mutate(document, mutation) - Applies a provided Mutation to every matching node in the tree, depth first.

There are 3 types of Mutations provided, and users may provide custom mutations. Provided are AddMutation, SubtractMutation, and SetMutation. Each performs functions hinted at by the name. See individual javadocs for details.

By default, calling document.toKDL() or document.writeKDL(writer) will print the structure with:

• 4 space indents
• No semicolons
• Printable ASCII characters which can be escaped, escaped
• Empty children printed
• null arguments and properties with null values printed
• \n (unicode \u{0a}) for newlines

Any of these can be changed by creating a new PrintConfig object and passing it into the print method. See the javadocs on PrintConfig for more information.

Please read the Code of Conduct before opening any issues or pull requests.

Besides code fixes, the easiest way to contribute is by generating test cases. Check out the test cases directory to see the existing ones. See the README there for more details.