File size: 5,691 Bytes
6efa67a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
import { chevrotain } from '../../../lib.js';
import { MacroLexer } from './MacroLexer.js';

const { CstParser } = chevrotain;

/** @typedef {import('chevrotain').TokenType} TokenType */
/** @typedef {import('chevrotain').CstNode} CstNode */
/** @typedef {import('chevrotain').ILexingError} ILexingError */
/** @typedef {import('chevrotain').IRecognitionException} IRecognitionException */

/**
 * The singleton instance of the MacroParser.
 *
 * @type {MacroParser}
 */
let instance;
export { instance as MacroParser };

class MacroParser extends CstParser {
    /** @type {MacroParser} */ static #instance;
    /** @type {MacroParser} */ static get instance() { return MacroParser.#instance ?? (MacroParser.#instance = new MacroParser()); }

    /** @private */
    constructor() {
        super(MacroLexer.def, {
            traceInitPerf: false,
            nodeLocationTracking: 'full',
            recoveryEnabled: true,
        });
        const Tokens = MacroLexer.tokens;

        const $ = this;

        // Top-level document rule that can handle both plaintext and macros
        $.document = $.RULE('document', () => {
            $.MANY(() => {
                $.OR([
                    { ALT: () => $.CONSUME(Tokens.Plaintext, { LABEL: 'plaintext' }) },
                    { ALT: () => $.CONSUME(Tokens.PlaintextOpenBrace, { LABEL: 'plaintext' }) },
                    { ALT: () => $.SUBRULE($.macro) },
                    { ALT: () => $.CONSUME(Tokens.Macro.Start, { LABEL: 'plaintext' }) },
                ]);
            });
        });

        // Basic Macro Structure
        $.macro = $.RULE('macro', () => {
            $.CONSUME(Tokens.Macro.Start);
            $.OR([
                { ALT: () => $.CONSUME(Tokens.Macro.DoubleSlash, { LABEL: 'Macro.identifier' }) },
                { ALT: () => $.CONSUME(Tokens.Macro.Identifier, { LABEL: 'Macro.identifier' }) },
            ]);
            $.OPTION(() => $.SUBRULE($.arguments));
            $.CONSUME(Tokens.Macro.End);
        });

        // Arguments Parsing
        $.arguments = $.RULE('arguments', () => {
            $.OR([
                {
                    ALT: () => {
                        $.CONSUME(Tokens.Args.DoubleColon, { LABEL: 'separator' });
                        $.AT_LEAST_ONE_SEP({
                            SEP: Tokens.Args.DoubleColon,
                            DEF: () => $.SUBRULE($.argument, { LABEL: 'argument' }),
                        });
                    },
                },
                {
                    ALT: () => {
                        $.OPTION(() => {
                            $.CONSUME(Tokens.Args.Colon, { LABEL: 'separator' });
                        });
                        $.SUBRULE($.argumentAllowingColons, { LABEL: 'argument' });
                    },
                    // So, this is a bit hacky. But implemented below, the argument capture does explicitly exclude double colons
                    // from being captured as the first token. The potential ambiguity chevrotain claims here is not possible.
                    // It says stuff like <Args.DoubleColon, Identifier/Macro/Unknown> is possible in both branches, but it is not.
                    IGNORE_AMBIGUITIES: true,
                },
            ]);
        });

        // List the argument tokens here, as we need two rules, one to be able to parse with double colons and one without
        const validArgumentTokens = [
            { ALT: () => $.SUBRULE($.macro) }, // Nested Macros
            { ALT: () => $.CONSUME(Tokens.Identifier) },
            { ALT: () => $.CONSUME(Tokens.Unknown) },
            { ALT: () => $.CONSUME(Tokens.Args.Colon) },
            { ALT: () => $.CONSUME(Tokens.Args.Equals) },
            { ALT: () => $.CONSUME(Tokens.Args.Quote) },
        ];

        $.argument = $.RULE('argument', () => {
            $.MANY(() => {
                $.OR([...validArgumentTokens]);
            });
        });
        $.argumentAllowingColons = $.RULE('argumentAllowingColons', () => {
            $.AT_LEAST_ONE(() => {
                $.OR([
                    ...validArgumentTokens,
                    { ALT: () => $.CONSUME(Tokens.Args.DoubleColon) },
                ]);
            });
        });

        this.performSelfAnalysis();
    }

    /**
     * Parses a document into a CST.
     *
     * @param {string} input
     * @returns {{ cst: CstNode|null, errors: ({ message: string }|ILexingError|IRecognitionException)[] , lexingErrors: ILexingError[], parserErrors: IRecognitionException[] }}
     */
    parseDocument(input) {
        if (!input) {
            return { cst: null, errors: [{ message: 'Input is empty' }], lexingErrors: [], parserErrors: [] };
        }

        const lexingResult = MacroLexer.tokenize(input);

        this.input = lexingResult.tokens;
        const cst = this.document();

        const errors = [
            ...lexingResult.errors,
            ...this.errors,
        ];

        return { cst, errors, lexingErrors: lexingResult.errors, parserErrors: this.errors };
    }

    test(input) {
        const lexingResult = MacroLexer.tokenize(input);
        // "input" is a setter which will reset the parser's state.
        this.input = lexingResult.tokens;
        const cst = this.macro();

        // For testing purposes we need to actually persist the error messages in the object,
        // otherwise the test cases cannot read those, as they don't have access to the exception object type.
        const errors = this.errors.map(x => ({ message: x.message, ...x, stack: x.stack }));

        return { cst, errors: errors };
    }
}

instance = MacroParser.instance;