tagset-parser 0.0.0

package/lib/cjs/generated/TagSetVisitor.js

@@ -0,0 +1,68 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.TagSetVisitor = void 0;
+const antlr4ng_1 = require("antlr4ng");
+/**
+ * This interface defines a complete generic visitor for a parse tree produced
+ * by `TagSetParser`.
+ *
+ * @param <Result> The return type of the visit operation. Use `void` for
+ * operations with no return type.
+ */
+class TagSetVisitor extends antlr4ng_1.AbstractParseTreeVisitor {
+    /**
+     * Visit a parse tree produced by `TagSetParser.program`.
+     * @param ctx the parse tree
+     * @return the visitor result
+     */
+    visitProgram;
+    /**
+     * Visit a parse tree produced by `TagSetParser.line`.
+     * @param ctx the parse tree
+     * @return the visitor result
+     */
+    visitLine;
+    /**
+     * Visit a parse tree produced by `TagSetParser.setDecl`.
+     * @param ctx the parse tree
+     * @return the visitor result
+     */
+    visitSetDecl;
+    /**
+     * Visit a parse tree produced by `TagSetParser.labelPart`.
+     * @param ctx the parse tree
+     * @return the visitor result
+     */
+    visitLabelPart;
+    /**
+     * Visit a parse tree produced by `TagSetParser.itemDecl`.
+     * @param ctx the parse tree
+     * @return the visitor result
+     */
+    visitItemDecl;
+    /**
+     * Visit a parse tree produced by `TagSetParser.sugarDecl`.
+     * @param ctx the parse tree
+     * @return the visitor result
+     */
+    visitSugarDecl;
+    /**
+     * Visit a parse tree produced by `TagSetParser.pattern`.
+     * @param ctx the parse tree
+     * @return the visitor result
+     */
+    visitPattern;
+    /**
+     * Visit a parse tree produced by `TagSetParser.valueList`.
+     * @param ctx the parse tree
+     * @return the visitor result
+     */
+    visitValueList;
+    /**
+     * Visit a parse tree produced by `TagSetParser.valueItem`.
+     * @param ctx the parse tree
+     * @return the visitor result
+     */
+    visitValueItem;
+}
+exports.TagSetVisitor = TagSetVisitor;

package/lib/cjs/index.js

@@ -0,0 +1,5 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.parse = void 0;
+var parser_js_1 = require("./parser.js");
+Object.defineProperty(exports, "parse", { enumerable: true, get: function () { return parser_js_1.parse; } });

package/lib/cjs/package.json

@@ -0,0 +1 @@
+{"type": "commonjs"}

package/lib/cjs/parser.js

@@ -0,0 +1,17 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.parse = parse;
+const antlr4ng_1 = require("antlr4ng");
+const TagSetLexer_js_1 = require("./generated/TagSetLexer.js");
+const TagSetParser_js_1 = require("./generated/TagSetParser.js");
+const visitor_js_1 = require("./visitor.js");
+function parse(input) {
+    const chars = antlr4ng_1.CharStream.fromString(input);
+    const lexer = new TagSetLexer_js_1.TagSetLexer(chars);
+    const tokens = new antlr4ng_1.CommonTokenStream(lexer);
+    const parser = new TagSetParser_js_1.TagSetParser(tokens);
+    const tree = parser.program();
+    const visitor = new visitor_js_1.TagSetASTVisitor();
+    tree.accept(visitor);
+    return visitor.buildAST();
+}

package/lib/cjs/types.js

@@ -0,0 +1,2 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });

package/lib/cjs/visitor.js

@@ -0,0 +1,85 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.TagSetASTVisitor = void 0;
+const TagSetVisitor_js_1 = require("./generated/TagSetVisitor.js");
+const bitmask_js_1 = require("./bitmask.js");
+function extractPatternIds(ctx) {
+    return ctx
+        .WORD()
+        .map((w) => w.getText())
+        .filter((t) => t !== '_');
+}
+function extractValues(ctx) {
+    return ctx.valueItem().map((vi) => {
+        const parts = [
+            ...vi.WORD().map((w) => w.getText()),
+            ...vi.QUOTED_STRING().map((q) => {
+                const text = q.getText();
+                return text.startsWith('"') && text.endsWith('"')
+                    ? text.slice(1, -1)
+                    : text;
+            }),
+        ];
+        return parts.join(' ');
+    });
+}
+function autoDetectSets(explicitSets, rawItems) {
+    const explicitIds = new Set(explicitSets.map((s) => s.id));
+    const discovered = [];
+    for (const item of rawItems) {
+        for (const id of item.patternIds) {
+            if (!explicitIds.has(id) && !discovered.includes(id)) {
+                discovered.push(id);
+            }
+        }
+    }
+    let nextIndex = explicitSets.length;
+    const autoSets = discovered.map((id) => ({
+        id,
+        label: id,
+        index: nextIndex++,
+    }));
+    return [...explicitSets, ...autoSets];
+}
+class TagSetASTVisitor extends TagSetVisitor_js_1.TagSetVisitor {
+    explicitSets = [];
+    rawItems = [];
+    setIndex = 0;
+    visitProgram = (ctx) => {
+        for (const line of ctx.line()) {
+            line.accept(this);
+        }
+    };
+    visitSetDecl = (ctx) => {
+        const id = ctx.WORD().getText();
+        const label = ctx
+            .labelPart()
+            .map((lp) => lp.getText())
+            .join(' ');
+        this.explicitSets.push({ id, label, index: this.setIndex++ });
+    };
+    visitItemDecl = (ctx) => {
+        const patternCtx = ctx.pattern();
+        const patternIds = extractPatternIds(patternCtx);
+        const values = extractValues(ctx.valueList());
+        this.rawItems.push({ patternIds, values });
+    };
+    visitSugarDecl = (ctx) => {
+        const patternIds = ctx
+            .WORD()
+            .map((w) => w.getText())
+            .filter((t) => t !== '_');
+        const values = extractValues(ctx.valueList());
+        this.rawItems.push({ patternIds, values });
+    };
+    buildAST() {
+        const sets = autoDetectSets(this.explicitSets, this.rawItems);
+        const items = this.rawItems.map((raw) => {
+            const pattern = raw.patternIds.join('&');
+            const bitmask = (0, bitmask_js_1.calcBitmask)(pattern, sets);
+            return { pattern, bitmask, values: raw.values };
+        });
+        return { sets, items };
+    }
+}
+exports.TagSetASTVisitor = TagSetASTVisitor;

package/lib/esm/bitmask.js

@@ -0,0 +1,15 @@
+export function calcBitmask(pattern, sets) {
+    const tokens = pattern.split('&');
+    let mask = 0;
+    for (const token of tokens) {
+        const t = token.trim();
+        if (t !== '' && t !== '_') {
+            for (const set of sets) {
+                if (set.id === t) {
+                    mask = mask | (1 << set.index);
+                }
+            }
+        }
+    }
+    return mask;
+}

package/lib/esm/generated/TagSetLexer.js

@@ -0,0 +1,75 @@
+import * as antlr from "antlr4ng";
+export class TagSetLexer extends antlr.Lexer {
+    static SET = 1;
+    static ITEM = 2;
+    static AMP = 3;
+    static COMMA = 4;
+    static COLON = 5;
+    static QUOTED_STRING = 6;
+    static WORD = 7;
+    static NL = 8;
+    static WS = 9;
+    static channelNames = [
+        "DEFAULT_TOKEN_CHANNEL", "HIDDEN"
+    ];
+    static literalNames = [
+        null, "'set'", "'item'", "'&'", "','", "':'"
+    ];
+    static symbolicNames = [
+        null, "SET", "ITEM", "AMP", "COMMA", "COLON", "QUOTED_STRING", "WORD",
+        "NL", "WS"
+    ];
+    static modeNames = [
+        "DEFAULT_MODE",
+    ];
+    static ruleNames = [
+        "SET", "ITEM", "AMP", "COMMA", "COLON", "QUOTED_STRING", "WORD",
+        "NL", "WS",
+    ];
+    constructor(input) {
+        super(input);
+        this.interpreter = new antlr.LexerATNSimulator(this, TagSetLexer._ATN, TagSetLexer.decisionsToDFA, new antlr.PredictionContextCache());
+    }
+    get grammarFileName() { return "TagSet.g4"; }
+    get literalNames() { return TagSetLexer.literalNames; }
+    get symbolicNames() { return TagSetLexer.symbolicNames; }
+    get ruleNames() { return TagSetLexer.ruleNames; }
+    get serializedATN() { return TagSetLexer._serializedATN; }
+    get channelNames() { return TagSetLexer.channelNames; }
+    get modeNames() { return TagSetLexer.modeNames; }
+    static _serializedATN = [
+        4, 0, 9, 60, 6, -1, 2, 0, 7, 0, 2, 1, 7, 1, 2, 2, 7, 2, 2, 3, 7, 3, 2, 4, 7, 4, 2, 5, 7, 5, 2,
+        6, 7, 6, 2, 7, 7, 7, 2, 8, 7, 8, 1, 0, 1, 0, 1, 0, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1,
+        2, 1, 3, 1, 3, 1, 4, 1, 4, 1, 5, 1, 5, 5, 5, 37, 8, 5, 10, 5, 12, 5, 40, 9, 5, 1, 5, 1, 5, 1,
+        6, 4, 6, 45, 8, 6, 11, 6, 12, 6, 46, 1, 7, 4, 7, 50, 8, 7, 11, 7, 12, 7, 51, 1, 8, 4, 8, 55,
+        8, 8, 11, 8, 12, 8, 56, 1, 8, 1, 8, 0, 0, 9, 1, 1, 3, 2, 5, 3, 7, 4, 9, 5, 11, 6, 13, 7, 15,
+        8, 17, 9, 1, 0, 4, 3, 0, 10, 10, 13, 13, 34, 34, 7, 0, 9, 10, 13, 13, 32, 32, 34, 34, 38,
+        38, 44, 44, 58, 58, 2, 0, 10, 10, 13, 13, 2, 0, 9, 9, 32, 32, 63, 0, 1, 1, 0, 0, 0, 0, 3,
+        1, 0, 0, 0, 0, 5, 1, 0, 0, 0, 0, 7, 1, 0, 0, 0, 0, 9, 1, 0, 0, 0, 0, 11, 1, 0, 0, 0, 0, 13, 1,
+        0, 0, 0, 0, 15, 1, 0, 0, 0, 0, 17, 1, 0, 0, 0, 1, 19, 1, 0, 0, 0, 3, 23, 1, 0, 0, 0, 5, 28, 1,
+        0, 0, 0, 7, 30, 1, 0, 0, 0, 9, 32, 1, 0, 0, 0, 11, 34, 1, 0, 0, 0, 13, 44, 1, 0, 0, 0, 15, 49,
+        1, 0, 0, 0, 17, 54, 1, 0, 0, 0, 19, 20, 5, 115, 0, 0, 20, 21, 5, 101, 0, 0, 21, 22, 5, 116,
+        0, 0, 22, 2, 1, 0, 0, 0, 23, 24, 5, 105, 0, 0, 24, 25, 5, 116, 0, 0, 25, 26, 5, 101, 0, 0,
+        26, 27, 5, 109, 0, 0, 27, 4, 1, 0, 0, 0, 28, 29, 5, 38, 0, 0, 29, 6, 1, 0, 0, 0, 30, 31, 5,
+        44, 0, 0, 31, 8, 1, 0, 0, 0, 32, 33, 5, 58, 0, 0, 33, 10, 1, 0, 0, 0, 34, 38, 5, 34, 0, 0,
+        35, 37, 8, 0, 0, 0, 36, 35, 1, 0, 0, 0, 37, 40, 1, 0, 0, 0, 38, 36, 1, 0, 0, 0, 38, 39, 1,
+        0, 0, 0, 39, 41, 1, 0, 0, 0, 40, 38, 1, 0, 0, 0, 41, 42, 5, 34, 0, 0, 42, 12, 1, 0, 0, 0, 43,
+        45, 8, 1, 0, 0, 44, 43, 1, 0, 0, 0, 45, 46, 1, 0, 0, 0, 46, 44, 1, 0, 0, 0, 46, 47, 1, 0, 0,
+        0, 47, 14, 1, 0, 0, 0, 48, 50, 7, 2, 0, 0, 49, 48, 1, 0, 0, 0, 50, 51, 1, 0, 0, 0, 51, 49,
+        1, 0, 0, 0, 51, 52, 1, 0, 0, 0, 52, 16, 1, 0, 0, 0, 53, 55, 7, 3, 0, 0, 54, 53, 1, 0, 0, 0,
+        55, 56, 1, 0, 0, 0, 56, 54, 1, 0, 0, 0, 56, 57, 1, 0, 0, 0, 57, 58, 1, 0, 0, 0, 58, 59, 6,
+        8, 0, 0, 59, 18, 1, 0, 0, 0, 5, 0, 38, 46, 51, 56, 1, 6, 0, 0
+    ];
+    static __ATN;
+    static get _ATN() {
+        if (!TagSetLexer.__ATN) {
+            TagSetLexer.__ATN = new antlr.ATNDeserializer().deserialize(TagSetLexer._serializedATN);
+        }
+        return TagSetLexer.__ATN;
+    }
+    static vocabulary = new antlr.Vocabulary(TagSetLexer.literalNames, TagSetLexer.symbolicNames, []);
+    get vocabulary() {
+        return TagSetLexer.vocabulary;
+    }
+    static decisionsToDFA = TagSetLexer._ATN.decisionToState.map((ds, index) => new antlr.DFA(ds, index));
+}

package/lib/esm/generated/TagSetListener.js

@@ -0,0 +1,100 @@
+/**
+ * This interface defines a complete listener for a parse tree produced by
+ * `TagSetParser`.
+ */
+export class TagSetListener {
+    /**
+     * Enter a parse tree produced by `TagSetParser.program`.
+     * @param ctx the parse tree
+     */
+    enterProgram;
+    /**
+     * Exit a parse tree produced by `TagSetParser.program`.
+     * @param ctx the parse tree
+     */
+    exitProgram;
+    /**
+     * Enter a parse tree produced by `TagSetParser.line`.
+     * @param ctx the parse tree
+     */
+    enterLine;
+    /**
+     * Exit a parse tree produced by `TagSetParser.line`.
+     * @param ctx the parse tree
+     */
+    exitLine;
+    /**
+     * Enter a parse tree produced by `TagSetParser.setDecl`.
+     * @param ctx the parse tree
+     */
+    enterSetDecl;
+    /**
+     * Exit a parse tree produced by `TagSetParser.setDecl`.
+     * @param ctx the parse tree
+     */
+    exitSetDecl;
+    /**
+     * Enter a parse tree produced by `TagSetParser.labelPart`.
+     * @param ctx the parse tree
+     */
+    enterLabelPart;
+    /**
+     * Exit a parse tree produced by `TagSetParser.labelPart`.
+     * @param ctx the parse tree
+     */
+    exitLabelPart;
+    /**
+     * Enter a parse tree produced by `TagSetParser.itemDecl`.
+     * @param ctx the parse tree
+     */
+    enterItemDecl;
+    /**
+     * Exit a parse tree produced by `TagSetParser.itemDecl`.
+     * @param ctx the parse tree
+     */
+    exitItemDecl;
+    /**
+     * Enter a parse tree produced by `TagSetParser.sugarDecl`.
+     * @param ctx the parse tree
+     */
+    enterSugarDecl;
+    /**
+     * Exit a parse tree produced by `TagSetParser.sugarDecl`.
+     * @param ctx the parse tree
+     */
+    exitSugarDecl;
+    /**
+     * Enter a parse tree produced by `TagSetParser.pattern`.
+     * @param ctx the parse tree
+     */
+    enterPattern;
+    /**
+     * Exit a parse tree produced by `TagSetParser.pattern`.
+     * @param ctx the parse tree
+     */
+    exitPattern;
+    /**
+     * Enter a parse tree produced by `TagSetParser.valueList`.
+     * @param ctx the parse tree
+     */
+    enterValueList;
+    /**
+     * Exit a parse tree produced by `TagSetParser.valueList`.
+     * @param ctx the parse tree
+     */
+    exitValueList;
+    /**
+     * Enter a parse tree produced by `TagSetParser.valueItem`.
+     * @param ctx the parse tree
+     */
+    enterValueItem;
+    /**
+     * Exit a parse tree produced by `TagSetParser.valueItem`.
+     * @param ctx the parse tree
+     */
+    exitValueItem;
+    visitTerminal(node) { }
+    visitErrorNode(node) { }
+    enterEveryRule(node) { }
+    exitEveryRule(node) { }
+}