ReefVM/src/bytecode.ts

import { type Value, type FunctionDef, toValue } from "./value"
import { OpCode } from "./opcode"

export type Bytecode = {
  instructions: Instruction[]
  constants: Constant[]
}

export type Instruction = {
  op: OpCode
  operand?: number | string
}

export type Constant =
  | Value
  | FunctionDef

//
// Parse bytecode from human-readable string format.
// Operand types are determined by prefix/literal:
//   #42      -> immediate number (e.g., JUMP #5, MAKE_ARRAY #3)
//   name     -> variable/function name (e.g., LOAD x, CALL_NATIVE add)
//   42       -> number constant (e.g., PUSH 42)
//   "str"    -> string constant (e.g., PUSH "hello")
//   'str'    -> string constant (e.g., PUSH 'hello')
//   true     -> boolean constant (e.g., PUSH true)
//   false    -> boolean constant (e.g., PUSH false)
//   null     -> null constant (e.g., PUSH null)
//
// Function definitions:
//   MAKE_FUNCTION (x y) #7           -> basic function
//   MAKE_FUNCTION (x y=42) #7        -> with defaults
//   MAKE_FUNCTION (x ...rest) #7     -> variadic
//   MAKE_FUNCTION (x @named) #7      -> kwargs
//

function parseFunctionParams(paramStr: string, constants: Constant[]): {
  params: string[]
  defaults: Record<string, number>
  variadic: boolean
  named: boolean
} {
  const params: string[] = []
  const defaults: Record<string, number> = {}
  let variadic = false
  let kwargs = false

  // Remove parens and split by whitespace
  const paramList = paramStr.slice(1, -1).trim()
  if (!paramList) {
    return { params, defaults, variadic, named: kwargs }
  }

  const parts = paramList.split(/\s+/)

  for (const part of parts) {
    // Check for named args (@name)
    if (part.startsWith('@')) {
      kwargs = true
      params.push(part.slice(1))

    } else if (part.startsWith('...')) {
      // Check for variadic (...name)
      variadic = true
      params.push(part.slice(3))

    } else if (part.includes('=')) {
      // Check for default value (name=value)
      const [name, defaultValue] = part.split('=').map(s => s.trim())
      params.push(name!)

      // Parse default value and add to constants
      if (/^-?\d+(\.\d+)?$/.test(defaultValue!)) {
        constants.push(toValue(parseFloat(defaultValue!)))
      } else if (/^['\"].*['\"]$/.test(defaultValue!)) {
        constants.push(toValue(defaultValue!.slice(1, -1)))
      } else if (defaultValue === 'true') {
        constants.push(toValue(true))
      } else if (defaultValue === 'false') {
        constants.push(toValue(false))
      } else if (defaultValue === 'null') {
        constants.push(toValue(null))
      } else {
        throw new Error(`Invalid default value: ${defaultValue}`)
      }

      defaults[name!] = constants.length - 1

    } else {
      params.push(part)
    }
  }

  return { params, defaults, variadic, named: kwargs }
}

export function toBytecode(str: string): Bytecode /* throws */ {
  const lines = str.trim().split("\n")

  const bytecode: Bytecode = {
    instructions: [],
    constants: []
  }

  for (let line of lines) {
    const trimmed = line.trim()
    if (!trimmed) continue

    const [op, ...rest] = trimmed.split(/\s+/)
    const opCode = OpCode[op as keyof typeof OpCode]

    if (opCode === undefined) {
      throw new Error(`Unknown opcode: ${op}`)
    }

    let operandValue: number | string | undefined = undefined

    if (rest.length > 0) {
      const operand = rest.join(' ')

      // Special handling for MAKE_FUNCTION with paren syntax
      if (opCode === OpCode.MAKE_FUNCTION && operand.startsWith('(')) {
        // Parse: MAKE_FUNCTION (params) #body
        const match = operand.match(/^(\(.*?\))\s+(#\d+)$/)
        if (!match) {
          throw new Error(`Invalid MAKE_FUNCTION syntax: ${operand}`)
        }

        const paramStr = match[1]!
        const bodyStr = match[2]!
        const body = parseInt(bodyStr.slice(1))

        const { params, defaults, variadic, named: kwargs } = parseFunctionParams(paramStr, bytecode.constants)

        // Add function definition to constants
        bytecode.constants.push({
          type: 'function_def',
          params,
          defaults,
          body,
          variadic,
          kwargs
        })

        operandValue = bytecode.constants.length - 1
      }
      else if (operand.startsWith('#')) {
        // immediate number
        operandValue = parseInt(operand.slice(1))

      } else if (/^['"].*['"]$/.test(operand)) {
        // string
        const stringValue = operand.slice(1, operand.length - 1)
        bytecode.constants.push(toValue(stringValue))
        operandValue = bytecode.constants.length - 1

      } else if (/^-?\d+(\.\d+)?$/.test(operand)) {
        // number
        bytecode.constants.push(toValue(parseFloat(operand)))
        operandValue = bytecode.constants.length - 1

      } else if (operand === 'true' || operand === 'false') {
        // boolean
        bytecode.constants.push(toValue(operand === 'true'))
        operandValue = bytecode.constants.length - 1

      } else if (operand === 'null') {
        // null
        bytecode.constants.push(toValue(null))
        operandValue = bytecode.constants.length - 1

      } else if (/^[a-zA-Z_].*$/.test(operand)) {
        // variable
        operandValue = operand

      } else {
        throw new Error(`Invalid operand: ${operand}`)
      }
    }

    bytecode.instructions.push({
      op: opCode,
      operand: operandValue
    })
  }

  return bytecode
}