@haz3y0ne/parsexl/parser/guessToken.ts

Parses Excel formulas into a clean, well-typed abstract syntax tree you can analyse or evaluate in TypeScript.

// lexer/guessTokenType.ts
import { tokenRegexMap } from "../lexer";
import type { ASTNode } from "../types";

/**
 * Map a raw token string to one or more ASTArg descriptors.
 * Leaves `args` off entirely – the parser will attach them.
 */
export function guessTokenType(token: string): ASTNode[] {
  // String literal
  if (tokenRegexMap.LITERAL.test(token)) {
    const unquoted = token.slice(1, -1).replace(/""/g, '"');
    return [
      { type: "Literal", value: unquoted, args: [], loc: { start: 0, end: 0 } },
    ];
  }

  // Boolean
  if (tokenRegexMap.BOOLEAN.test(token)) {
    return [
      {
        type: "Boolean",
        value: token.toUpperCase() === "TRUE",
        args: [],
        loc: { start: 0, end: 0 },
      },
    ];
  }

  // Dynamic errors (#SPILL!, #CALC!, etc.)
  if (tokenRegexMap.DYNAMIC_ERROR.test(token)) {
    return [
      {
        type: "DynamicError",
        value: token,
        args: [],
        loc: { start: 0, end: 0 },
      },
    ];
  }

  // Standard Excel errors (#DIV/0, #NAME?, etc.)
  if (tokenRegexMap.ERROR.test(token)) {
    return [
      { type: "Error", value: token, args: [], loc: { start: 0, end: 0 } },
    ];
  }

  // Inline array literal
  if (tokenRegexMap.ARRAY.test(token)) {
    return [
      { type: "Array", value: token, args: [], loc: { start: 0, end: 0 } },
    ];
  }

  // Three-D reference (Sheet1:Sheet3!A1)
  if (tokenRegexMap.THREE_D_REFERENCE.test(token)) {
    return [
      {
        type: "ThreeDReference",
        value: token,
        args: [],
        loc: { start: 0, end: 0 },
      },
    ];
  }

  // External workbook reference ([Book1]Sheet1!A1)
  if (tokenRegexMap.EXTERNAL_REFERENCE.test(token)) {
    return [
      {
        type: "ExternalReference",
        value: token,
        args: [],
        loc: { start: 0, end: 0 },
      },
    ];
  }

  // Structured table extension ([[#Headers]])
  if (tokenRegexMap.STRUCTURED_EXT.test(token)) {
    return [
      {
        type: "StructuredReferenceExt",
        value: token,
        args: [],
        loc: { start: 0, end: 0 },
      },
    ];
  }

  // Structured table reference ([Column] or Table[Column])
  if (
    tokenRegexMap.STRUCTURED_REFERENCE.test(token) ||
    tokenRegexMap.TABLE_COLUMN.test(token)
  ) {
    return [
      {
        type: "StructuredReference",
        value: token,
        args: [],
        loc: { start: 0, end: 0 },
      },
    ];
  }

  // R1C1 reference
  if (tokenRegexMap.R1C1.test(token)) {
    return [
      { type: "R1C1", value: token, args: [], loc: { start: 0, end: 0 } },
    ];
  }

  // Cell or range reference (A1 or A1:B2)
  if (tokenRegexMap.CELL_OR_RANGE.test(token)) {
    return token.includes(":")
      ? [{ type: "Range", value: token, args: [], loc: { start: 0, end: 0 } }]
      : [{ type: "Cell", value: token, args: [], loc: { start: 0, end: 0 } }];
  }

  // Percentage literal (e.g. 12.5%)
  if (tokenRegexMap.PERCENTAGE.test(token)) {
    return [
      {
        type: "Percentage",
        value: parseFloat(token) / 100,
        args: [],
        loc: { start: 0, end: 0 },
      },
    ];
  }

  // Single-percent operator ("%")
  if (tokenRegexMap.PERCENT_OPERATOR.test(token)) {
    return [
      {
        type: "PercentOperator",
        value: token,
        args: [],
        loc: { start: 0, end: 0 },
      },
    ];
  }

  // Number
  if (tokenRegexMap.NUMBER.test(token)) {
    return [
      {
        type: "Number",
        value: parseFloat(token),
        args: [],
        loc: { start: 0, end: 0 },
      },
    ];
  }

  // Function call (SUM(…)
  if (tokenRegexMap.FUNCTION.test(token)) {
    const name = token.slice(0, -1);
    return [
      { type: "Function", value: name, args: [], loc: { start: 0, end: 0 } },
    ];
  }

  // Operators (+, -, *, /, ^, &, =, <>, etc.)
  if (tokenRegexMap.OPERATOR.test(token)) {
    return [
      { type: "Operator", value: token, args: [], loc: { start: 0, end: 0 } },
    ];
  }

  // Parentheses
  if (tokenRegexMap.PARENTHESIS.test(token)) {
    return [
      {
        type: "Parenthesis",
        value: token,
        args: [],
        loc: { start: 0, end: 0 },
      },
    ];
  }

  // Argument separators (, or ;)
  if (tokenRegexMap.ARG_SEPARATOR.test(token)) {
    return [
      {
        type: "ArgumentSeparator",
        value: token,
        args: [],
        loc: { start: 0, end: 0 },
      },
    ];
  }

  // Implicit intersection operator (@)
  if (tokenRegexMap.IMPLICIT_INTERSECTION.test(token)) {
    return [
      {
        type: "ImplicitIntersection",
        value: token,
        args: [],
        loc: { start: 0, end: 0 },
      },
    ];
  }

  // Spill operator (#)
  if (tokenRegexMap.SPILL_OPERATOR.test(token)) {
    return [
      {
        type: "SpillOperator",
        value: token,
        args: [],
        loc: { start: 0, end: 0 },
      },
    ];
  }

  // Wildcard (* or ?)
  if (tokenRegexMap.WILDCARD.test(token)) {
    return [
      { type: "Wildcard", value: token, args: [], loc: { start: 0, end: 0 } },
    ];
  }

  // Named range (unquoted identifier)
  if (tokenRegexMap.NAMED_RANGE.test(token)) {
    return [
      { type: "NamedRange", value: token, args: [], loc: { start: 0, end: 0 } },
    ];
  }

  // Fallback: anything else as a bare expression
  return [
    { type: "Expression", value: token, args: [], loc: { start: 0, end: 0 } },
  ];
}