BPOS-TMP-REPO/packages/sfeel/src/lexer.ts

/**
 * S-FEEL 子集的词法分析 —— 纯函数(编码规范 §8.3;定义域 OS 03)。
 *
 * 子集范围:比较、逻辑、算术、字符串与日期函数,**无外部调用,静态可检**。
 */

export const TOKEN_KINDS = [
  'number',
  'string',
  'name',
  'punct',
  'eof',
] as const;
export type TokenKind = (typeof TOKEN_KINDS)[number];

export type Token = {
  readonly kind: TokenKind;
  readonly text: string;
  readonly value?: number | string;
  readonly pos: number;
};

export class SfeelSyntaxError extends Error {
  constructor(
    message: string,
    readonly pos: number,
  ) {
    super(`S-FEEL 语法错误(第 ${pos} 字符):${message}`);
    this.name = 'SfeelSyntaxError';
  }
}

/** 多字符符号先于单字符匹配 */
const PUNCT = ['<=', '>=', '!=', '==', '..', '<', '>', '=', '+', '-', '*', '/', '(', ')', '[', ']', ',', '.'] as const;

const NAME_START = /[A-Za-z_一-龥]/;
const NAME_PART = /[A-Za-z0-9_一-龥]/;

export function tokenize(source: string): Token[] {
  const tokens: Token[] = [];
  let i = 0;

  while (i < source.length) {
    const ch = source[i]!;

    if (/\s/.test(ch)) {
      i += 1;
      continue;
    }

    // 字符串:单引号或双引号,支持 \" \\ \n \t
    if (ch === '"' || ch === "'") {
      const quote = ch;
      const start = i;
      i += 1;
      let value = '';
      while (i < source.length && source[i] !== quote) {
        if (source[i] === '\\') {
          const next = source[i + 1];
          value +=
            next === 'n' ? '\n' : next === 't' ? '\t' : next === '\\' ? '\\' : next === quote ? quote : (next ?? '');
          i += 2;
          continue;
        }
        value += source[i];
        i += 1;
      }
      if (i >= source.length) throw new SfeelSyntaxError('字符串未闭合', start);
      i += 1;
      tokens.push({ kind: 'string', text: source.slice(start, i), value, pos: start });
      continue;
    }

    // 数字:整数、小数;不支持科学计数(子集封顶)
    if (/[0-9]/.test(ch)) {
      const start = i;
      while (i < source.length && /[0-9]/.test(source[i]!)) i += 1;
      // `..` 是区间符号,不是小数点
      if (source[i] === '.' && source[i + 1] !== '.') {
        i += 1;
        while (i < source.length && /[0-9]/.test(source[i]!)) i += 1;
      }
      const text = source.slice(start, i);
      tokens.push({ kind: 'number', text, value: Number(text), pos: start });
      continue;
    }

    // 名字:标识符与关键字;允许空格分隔的多词函数名(如 `starts with`)在 parser 里合并
    if (NAME_START.test(ch)) {
      const start = i;
      while (i < source.length && NAME_PART.test(source[i]!)) i += 1;
      tokens.push({ kind: 'name', text: source.slice(start, i), pos: start });
      continue;
    }

    const punct = PUNCT.find((p) => source.startsWith(p, i));
    if (punct !== undefined) {
      tokens.push({ kind: 'punct', text: punct, pos: i });
      i += punct.length;
      continue;
    }

    throw new SfeelSyntaxError(`无法识别的字符 ${JSON.stringify(ch)}`, i);
  }

  tokens.push({ kind: 'eof', text: '', pos: source.length });
  return tokens;
}