@plurnk/plurnk-parser

The parser for the PLURNK model language: the ANTLR grammars, the generated lexer and parser, the AST builder, and path decomposition. It implements the language that @plurnk/plurnk-contracts specifies and teaches; the AST and wire types it produces are the contracts package's.

Install

npm install @plurnk/plurnk-parser

The service's execution path is the consumer. A client that validates or presents the wire needs only @plurnk/plurnk-contracts, which carries no parser runtime.

Parser

import { PlurnkParser } from "@plurnk/plurnk-parser";

const result = PlurnkParser.parse(input);

for (const item of result.items) {
    if (item.kind === "statement") {
        console.log(item.statement.op);
    }
}

Parse items are ordered and discriminate as statement or error. Outside-block text is ignored in every tier; literal bodies remain exact. The parser entry points deliberately accept different document tiers:

| Entry point | Accepted input | |--------------------------------|-------------------------------------------------------| | PlurnkParser.parse | One operation-bearing model turn; omitted lifecycle declaration continues silently | | PlurnkParser.parseStatements | A sequence of protocol statements | | PlurnkParser.parseClient | Protocol statements plus the client-only LOOK | | parsePath | One path or URI using parser-equivalent decomposition |

See the contracts SPEC {§turn-shape} and {§tier-entrypoints} for the tier boundaries. The AST, parse-result, and wire types come from @plurnk/plurnk-contracts.

CLI

plurnk-parser [file]    parse a file, or standard input when omitted
plurnk-parser --help    show usage

The CLI prints the parse result as JSON and exits 0 for a clean parse or 1 when the result contains an error or unparsed tail.

Development

npm run build
npm test
npm run test:installation

npm run build:grammar regenerates the parser from plurnkLexer.g4 and plurnkParser.g4. Change the grammar rather than editing generated output.

License

MIT