mirror of
https://github.com/kanaka/mal.git
synced 2024-11-13 11:23:59 +03:00
d5b81cc072
Related to issue #142.
602 lines
20 KiB
Swift
602 lines
20 KiB
Swift
//******************************************************************************
|
|
// MAL - step 8 - macros
|
|
//******************************************************************************
|
|
// This file is automatically generated from templates/step.swift. Rather than
|
|
// editing it directly, it's probably better to edit templates/step.swift and
|
|
// regenerate this file. Otherwise, your change might be lost if/when someone
|
|
// else performs that process.
|
|
//******************************************************************************
|
|
|
|
import Foundation
|
|
|
|
// The number of times EVAL has been entered recursively. We keep track of this
|
|
// so that we can protect against overrunning the stack.
|
|
//
|
|
private var EVAL_level = 0
|
|
|
|
// The maximum number of times we let EVAL recurse before throwing an exception.
|
|
// Testing puts this at some place between 1800 and 1900. Let's keep it at 500
|
|
// for safety's sake.
|
|
//
|
|
private let EVAL_leval_max = 500
|
|
|
|
// Control whether or not tail-call optimization (TCO) is enabled. We want it
|
|
// `true` most of the time, but may disable it for debugging purposes (it's
|
|
// easier to get a meaningful backtrace that way).
|
|
//
|
|
private let TCO = true
|
|
|
|
// Control whether or not we emit debugging statements in EVAL.
|
|
//
|
|
private let DEBUG_EVAL = false
|
|
|
|
// String used to prefix information logged in EVAL. Increasing lengths of the
|
|
// string are used the more EVAL is recursed.
|
|
//
|
|
private let INDENT_TEMPLATE = "|----|----|----|----|----|----|----|----|" +
|
|
"----|----|----|----|----|----|----|----|----|----|----|" +
|
|
"----|----|----|----|----|----|----|----|----|----|----|" +
|
|
"----|----|----|----|----|----|----|----|----|----|----|" +
|
|
"----|----|----|----|----|----|----|----|----|----|----|" +
|
|
"----|----|----|----|----|----|----|----|----|----|----|" +
|
|
"----|----|----|----|----|----|----|----|----|----|----|" +
|
|
"----|----|----|----|----|----|----|----|----|----|----|" +
|
|
"----|----|----|----|----|----|----|----|----|----|----|" +
|
|
"----|----|----|----|----|----|----|----|----|----|----|" +
|
|
"----|----|----|----|----|----|----|----|----|----|----|"
|
|
|
|
// Holds the prefix of INDENT_TEMPLATE used for actual logging.
|
|
//
|
|
private var indent = String()
|
|
|
|
// Symbols used in this module.
|
|
//
|
|
private let kValArgv = make_symbol("*ARGV*")
|
|
private let kValConcat = make_symbol("concat")
|
|
private let kValCons = make_symbol("cons")
|
|
private let kValDef = make_symbol("def!")
|
|
private let kValDefMacro = make_symbol("defmacro!")
|
|
private let kValDo = make_symbol("do")
|
|
private let kValEval = make_symbol("eval")
|
|
private let kValFn = make_symbol("fn*")
|
|
private let kValIf = make_symbol("if")
|
|
private let kValLet = make_symbol("let*")
|
|
private let kValMacroExpand = make_symbol("macroexpand")
|
|
private let kValQuasiQuote = make_symbol("quasiquote")
|
|
private let kValQuote = make_symbol("quote")
|
|
private let kValSpliceUnquote = make_symbol("splice-unquote")
|
|
private let kValUnquote = make_symbol("unquote")
|
|
private let kValTry = make_symbol("try*")
|
|
|
|
private let kSymbolArgv = as_symbol(kValArgv)
|
|
private let kSymbolConcat = as_symbol(kValConcat)
|
|
private let kSymbolCons = as_symbol(kValCons)
|
|
private let kSymbolDef = as_symbol(kValDef)
|
|
private let kSymbolDefMacro = as_symbol(kValDefMacro)
|
|
private let kSymbolDo = as_symbol(kValDo)
|
|
private let kSymbolEval = as_symbol(kValEval)
|
|
private let kSymbolFn = as_symbol(kValFn)
|
|
private let kSymbolIf = as_symbol(kValIf)
|
|
private let kSymbolLet = as_symbol(kValLet)
|
|
private let kSymbolMacroExpand = as_symbol(kValMacroExpand)
|
|
private let kSymbolQuasiQuote = as_symbol(kValQuasiQuote)
|
|
private let kSymbolQuote = as_symbol(kValQuote)
|
|
private let kSymbolSpliceUnquote = as_symbol(kValSpliceUnquote)
|
|
private let kSymbolUnquote = as_symbol(kValUnquote)
|
|
|
|
func substring(s: String, _ begin: Int, _ end: Int) -> String {
|
|
return s[s.startIndex.advancedBy(begin) ..< s.startIndex.advancedBy(end)]
|
|
}
|
|
|
|
// Parse the string into an AST.
|
|
//
|
|
private func READ(str: String) throws -> MalVal {
|
|
return try read_str(str)
|
|
}
|
|
|
|
// Return whether or not `val` is a non-empty list.
|
|
//
|
|
private func is_pair(val: MalVal) -> Bool {
|
|
if let seq = as_sequenceQ(val) {
|
|
return !seq.isEmpty
|
|
}
|
|
return false
|
|
}
|
|
|
|
// Expand macros for as long as the expression looks like a macro invocation.
|
|
//
|
|
private func macroexpand(var ast: MalVal, _ env: Environment) throws -> MalVal {
|
|
while true {
|
|
if let ast_as_list = as_listQ(ast) where !ast_as_list.isEmpty,
|
|
let macro_name = as_symbolQ(ast_as_list.first()),
|
|
let obj = env.get(macro_name),
|
|
let macro = as_macroQ(obj)
|
|
{
|
|
let new_env = Environment(outer: macro.env)
|
|
let rest = as_sequence(ast_as_list.rest())
|
|
let _ = try new_env.set_bindings(macro.args, with_exprs: rest)
|
|
ast = try EVAL(macro.body, new_env)
|
|
continue
|
|
}
|
|
return ast
|
|
}
|
|
}
|
|
|
|
// Evaluate `quasiquote`, possibly recursing in the process.
|
|
//
|
|
// As with quote, unquote, and splice-unquote, quasiquote takes a single
|
|
// parameter, typically a list. In the general case, this list is processed
|
|
// recursively as:
|
|
//
|
|
// (quasiquote (first rest...)) -> (cons (quasiquote first) (quasiquote rest))
|
|
//
|
|
// In the processing of the parameter passed to it, quasiquote handles three
|
|
// special cases:
|
|
//
|
|
// * If the parameter is an atom or an empty list, the following expression
|
|
// is formed and returned for evaluation:
|
|
//
|
|
// (quasiquote atom-or-empty-list) -> (quote atom-or-empty-list)
|
|
//
|
|
// * If the first element of the non-empty list is the symbol "unquote"
|
|
// followed by a second item, the second item is returned as-is:
|
|
//
|
|
// (quasiquote (unquote fred)) -> fred
|
|
//
|
|
// * If the first element of the non-empty list is another list containing
|
|
// the symbol "splice-unquote" followed by a list, that list is catenated
|
|
// with the quasiquoted result of the remaining items in the non-empty
|
|
// parent list:
|
|
//
|
|
// (quasiquote (splice-unquote list) rest...) -> (items-from-list items-from-quasiquote(rest...))
|
|
//
|
|
// Note the inconsistent handling between "quote" and "splice-quote". The former
|
|
// is handled when this function is handed a list that starts with "quote",
|
|
// whereas the latter is handled when this function is handled a list whose
|
|
// first element is a list that starts with "splice-quote". The handling of the
|
|
// latter is forced by the need to incorporate the results of (splice-quote
|
|
// list) with the remaining items of the list containing that splice-quote
|
|
// expression. However, it's not clear to me why the handling of "unquote" is
|
|
// not handled similarly, for consistency's sake.
|
|
//
|
|
private func quasiquote(qq_arg: MalVal) throws -> MalVal {
|
|
|
|
// If the argument is an atom or empty list:
|
|
//
|
|
// Return: (quote <argument>)
|
|
|
|
if !is_pair(qq_arg) {
|
|
return make_list_from(kValQuote, qq_arg)
|
|
}
|
|
|
|
// The argument is a non-empty list -- that is (item rest...)
|
|
|
|
// If the first item from the list is a symbol and it's "unquote" -- that
|
|
// is, (unquote item ignored...):
|
|
//
|
|
// Return: item
|
|
|
|
let qq_list = as_sequence(qq_arg)
|
|
if let sym = as_symbolQ(qq_list.first()) where sym == kSymbolUnquote {
|
|
return qq_list.count >= 2 ? try! qq_list.nth(1) : make_nil()
|
|
}
|
|
|
|
// If the first item from the list is itself a non-empty list starting with
|
|
// "splice-unquote"-- that is, ((splice-unquote item ignored...) rest...):
|
|
//
|
|
// Return: (concat item quasiquote(rest...))
|
|
|
|
if is_pair(qq_list.first()) {
|
|
let qq_list_item0 = as_sequence(qq_list.first())
|
|
if let sym = as_symbolQ(qq_list_item0.first()) where sym == kSymbolSpliceUnquote {
|
|
let result = try quasiquote(qq_list.rest())
|
|
return make_list_from(kValConcat, try! qq_list_item0.nth(1), result)
|
|
}
|
|
}
|
|
|
|
// General case: (item rest...):
|
|
//
|
|
// Return: (cons (quasiquote item) (quasiquote (rest...))
|
|
|
|
let first = try quasiquote(qq_list.first())
|
|
let rest = try quasiquote(qq_list.rest())
|
|
return make_list_from(kValCons, first, rest)
|
|
}
|
|
|
|
// Perform a simple evaluation of the `ast` object. If it's a symbol,
|
|
// dereference it and return its value. If it's a collection, call EVAL on all
|
|
// elements (or just the values, in the case of the hashmap). Otherwise, return
|
|
// the object unchanged.
|
|
//
|
|
private func eval_ast(ast: MalVal, _ env: Environment) throws -> MalVal {
|
|
if let symbol = as_symbolQ(ast) {
|
|
guard let val = env.get(symbol) else {
|
|
try throw_error("'\(symbol)' not found") // Specific text needed to match MAL unit tests
|
|
}
|
|
return val
|
|
}
|
|
if let list = as_listQ(ast) {
|
|
var result = [MalVal]()
|
|
result.reserveCapacity(Int(list.count))
|
|
for item in list {
|
|
let eval = try EVAL(item, env)
|
|
result.append(eval)
|
|
}
|
|
return make_list(result)
|
|
}
|
|
if let vec = as_vectorQ(ast) {
|
|
var result = [MalVal]()
|
|
result.reserveCapacity(Int(vec.count))
|
|
for item in vec {
|
|
let eval = try EVAL(item, env)
|
|
result.append(eval)
|
|
}
|
|
return make_vector(result)
|
|
}
|
|
if let hash = as_hashmapQ(ast) {
|
|
var result = [MalVal]()
|
|
result.reserveCapacity(Int(hash.count) * 2)
|
|
for (k, v) in hash {
|
|
let new_v = try EVAL(v, env)
|
|
result.append(k)
|
|
result.append(new_v)
|
|
}
|
|
return make_hashmap(result)
|
|
}
|
|
return ast
|
|
}
|
|
|
|
private enum TCOVal {
|
|
case NoResult
|
|
case Return(MalVal)
|
|
case Continue(MalVal, Environment)
|
|
|
|
init() { self = .NoResult }
|
|
init(_ result: MalVal) { self = .Return(result) }
|
|
init(_ ast: MalVal, _ env: Environment) { self = .Continue(ast, env) }
|
|
}
|
|
|
|
// EVALuate "def!" and "defmacro!".
|
|
//
|
|
private func eval_def(list: MalSequence, _ env: Environment) throws -> TCOVal {
|
|
guard list.count == 3 else {
|
|
try throw_error("expected 2 arguments to def!, got \(list.count - 1)")
|
|
}
|
|
let arg0 = try! list.nth(0)
|
|
let arg1 = try! list.nth(1)
|
|
let arg2 = try! list.nth(2)
|
|
guard let sym = as_symbolQ(arg1) else {
|
|
try throw_error("expected symbol for first argument to def!")
|
|
}
|
|
var value = try EVAL(arg2, env)
|
|
if as_symbol(arg0) == kSymbolDefMacro {
|
|
guard let closure = as_closureQ(value) else {
|
|
try throw_error("expected closure, got \(value)")
|
|
}
|
|
value = make_macro(closure)
|
|
}
|
|
return TCOVal(env.set(sym, value))
|
|
}
|
|
|
|
// EVALuate "let*".
|
|
//
|
|
private func eval_let(list: MalSequence, _ env: Environment) throws -> TCOVal {
|
|
guard list.count == 3 else {
|
|
try throw_error("expected 2 arguments to let*, got \(list.count - 1)")
|
|
}
|
|
let arg1 = try! list.nth(1)
|
|
let arg2 = try! list.nth(2)
|
|
guard let bindings = as_sequenceQ(arg1) else {
|
|
try throw_error("expected list for first argument to let*")
|
|
}
|
|
guard bindings.count % 2 == 0 else {
|
|
try throw_error("expected even number of elements in bindings to let*, got \(bindings.count)")
|
|
}
|
|
let new_env = Environment(outer: env)
|
|
for var index: MalIntType = 0; index < bindings.count; index += 2 {
|
|
let binding_name = try! bindings.nth(index)
|
|
let binding_value = try! bindings.nth(index + 1)
|
|
guard let binding_symbol = as_symbolQ(binding_name) else {
|
|
try throw_error("expected symbol for first element in binding pair")
|
|
}
|
|
let evaluated_value = try EVAL(binding_value, new_env)
|
|
new_env.set(binding_symbol, evaluated_value)
|
|
}
|
|
if TCO {
|
|
return TCOVal(arg2, new_env)
|
|
}
|
|
return TCOVal(try EVAL(arg2, new_env))
|
|
}
|
|
|
|
// EVALuate "do".
|
|
//
|
|
private func eval_do(list: MalSequence, _ env: Environment) throws -> TCOVal {
|
|
if TCO {
|
|
let _ = try eval_ast(list.range_from(1, to: list.count-1), env)
|
|
return TCOVal(list.last(), env)
|
|
}
|
|
|
|
let evaluated_ast = try eval_ast(list.rest(), env)
|
|
let evaluated_seq = as_sequence(evaluated_ast)
|
|
return TCOVal(evaluated_seq.last())
|
|
}
|
|
|
|
// EVALuate "if".
|
|
//
|
|
private func eval_if(list: MalSequence, _ env: Environment) throws -> TCOVal {
|
|
guard list.count >= 3 else {
|
|
try throw_error("expected at least 2 arguments to if, got \(list.count - 1)")
|
|
}
|
|
let cond_result = try EVAL(try! list.nth(1), env)
|
|
var new_ast: MalVal
|
|
if is_truthy(cond_result) {
|
|
new_ast = try! list.nth(2)
|
|
} else if list.count == 4 {
|
|
new_ast = try! list.nth(3)
|
|
} else {
|
|
return TCOVal(make_nil())
|
|
}
|
|
if TCO {
|
|
return TCOVal(new_ast, env)
|
|
}
|
|
return TCOVal(try EVAL(new_ast, env))
|
|
}
|
|
|
|
// EVALuate "fn*".
|
|
//
|
|
private func eval_fn(list: MalSequence, _ env: Environment) throws -> TCOVal {
|
|
guard list.count == 3 else {
|
|
try throw_error("expected 2 arguments to fn*, got \(list.count - 1)")
|
|
}
|
|
guard let seq = as_sequenceQ(try! list.nth(1)) else {
|
|
try throw_error("expected list or vector for first argument to fn*")
|
|
}
|
|
return TCOVal(make_closure((eval: EVAL, args: seq, body: try! list.nth(2), env: env)))
|
|
}
|
|
|
|
// EVALuate "quote".
|
|
//
|
|
private func eval_quote(list: MalSequence, _ env: Environment) throws -> TCOVal {
|
|
if list.count >= 2 {
|
|
return TCOVal(try! list.nth(1))
|
|
}
|
|
return TCOVal(make_nil())
|
|
}
|
|
|
|
// EVALuate "quasiquote".
|
|
//
|
|
private func eval_quasiquote(list: MalSequence, _ env: Environment) throws -> TCOVal {
|
|
guard list.count >= 2 else {
|
|
try throw_error("Expected non-nil parameter to 'quasiquote'")
|
|
}
|
|
if TCO {
|
|
return TCOVal(try quasiquote(try! list.nth(1)), env)
|
|
}
|
|
return TCOVal(try EVAL(try quasiquote(try! list.nth(1)), env))
|
|
}
|
|
|
|
// EVALuate "macroexpand".
|
|
//
|
|
private func eval_macroexpand(list: MalSequence, _ env: Environment) throws -> TCOVal {
|
|
guard list.count >= 2 else {
|
|
try throw_error("Expected parameter to 'macroexpand'")
|
|
}
|
|
return TCOVal(try macroexpand(try! list.nth(1), env))
|
|
}
|
|
|
|
// Walk the AST and completely evaluate it, handling macro expansions, special
|
|
// forms and function calls.
|
|
//
|
|
private func EVAL(var ast: MalVal, var _ env: Environment) throws -> MalVal {
|
|
EVAL_level++
|
|
defer { EVAL_level-- }
|
|
guard EVAL_level <= EVAL_leval_max else {
|
|
try throw_error("Recursing too many levels (> \(EVAL_leval_max))")
|
|
}
|
|
|
|
if DEBUG_EVAL {
|
|
indent = substring(INDENT_TEMPLATE, 0, EVAL_level)
|
|
}
|
|
|
|
while true {
|
|
if DEBUG_EVAL { print("\(indent)> \(ast)") }
|
|
|
|
if !is_list(ast) {
|
|
|
|
// Not a list -- just evaluate and return.
|
|
|
|
let answer = try eval_ast(ast, env)
|
|
if DEBUG_EVAL { print("\(indent)>>> \(answer)") }
|
|
return answer
|
|
}
|
|
|
|
// Special handling if it's a list.
|
|
|
|
var list = as_list(ast)
|
|
ast = try macroexpand(ast, env)
|
|
if !is_list(ast) {
|
|
|
|
// Not a list -- just evaluate and return.
|
|
|
|
let answer = try eval_ast(ast, env)
|
|
if DEBUG_EVAL { print("\(indent)>>> \(answer)") }
|
|
return answer
|
|
}
|
|
list = as_list(ast)
|
|
|
|
if DEBUG_EVAL { print("\(indent)>. \(list)") }
|
|
|
|
if list.isEmpty {
|
|
return ast
|
|
}
|
|
|
|
// Check for special forms, where we want to check the operation
|
|
// before evaluating all of the parameters.
|
|
|
|
let arg0 = list.first()
|
|
if let fn_symbol = as_symbolQ(arg0) {
|
|
let res: TCOVal
|
|
|
|
switch fn_symbol {
|
|
case kSymbolDef: res = try eval_def(list, env)
|
|
case kSymbolDefMacro: res = try eval_def(list, env)
|
|
case kSymbolLet: res = try eval_let(list, env)
|
|
case kSymbolDo: res = try eval_do(list, env)
|
|
case kSymbolIf: res = try eval_if(list, env)
|
|
case kSymbolFn: res = try eval_fn(list, env)
|
|
case kSymbolQuote: res = try eval_quote(list, env)
|
|
case kSymbolQuasiQuote: res = try eval_quasiquote(list, env)
|
|
case kSymbolMacroExpand: res = try eval_macroexpand(list, env)
|
|
default: res = TCOVal()
|
|
}
|
|
switch res {
|
|
case let .Return(result): return result
|
|
case let .Continue(new_ast, new_env): ast = new_ast; env = new_env; continue
|
|
case .NoResult: break
|
|
}
|
|
}
|
|
|
|
// Standard list to be applied. Evaluate all the elements first.
|
|
|
|
let eval = try eval_ast(ast, env)
|
|
|
|
// The result had better be a list and better be non-empty.
|
|
|
|
let eval_list = as_list(eval)
|
|
if eval_list.isEmpty {
|
|
return eval
|
|
}
|
|
|
|
if DEBUG_EVAL { print("\(indent)>> \(eval)") }
|
|
|
|
// Get the first element of the list and execute it.
|
|
|
|
let first = eval_list.first()
|
|
let rest = as_sequence(eval_list.rest())
|
|
|
|
if let fn = as_builtinQ(first) {
|
|
let answer = try fn.apply(rest)
|
|
if DEBUG_EVAL { print("\(indent)>>> \(answer)") }
|
|
return answer
|
|
} else if let fn = as_closureQ(first) {
|
|
let new_env = Environment(outer: fn.env)
|
|
let _ = try new_env.set_bindings(fn.args, with_exprs: rest)
|
|
if TCO {
|
|
env = new_env
|
|
ast = fn.body
|
|
continue
|
|
}
|
|
let answer = try EVAL(fn.body, new_env)
|
|
if DEBUG_EVAL { print("\(indent)>>> \(answer)") }
|
|
return answer
|
|
}
|
|
|
|
// The first element wasn't a function to be executed. Return an
|
|
// error saying so.
|
|
|
|
try throw_error("first list item does not evaluate to a function: \(first)")
|
|
}
|
|
}
|
|
|
|
// Convert the value into a human-readable string for printing.
|
|
//
|
|
private func PRINT(exp: MalVal) -> String {
|
|
return pr_str(exp, true)
|
|
}
|
|
|
|
// Perform the READ and EVAL steps. Useful for when you don't care about the
|
|
// printable result.
|
|
//
|
|
private func RE(text: String, _ env: Environment) -> MalVal? {
|
|
if !text.isEmpty {
|
|
do {
|
|
let ast = try READ(text)
|
|
do {
|
|
return try EVAL(ast, env)
|
|
} catch let error as MalException {
|
|
print("Error evaluating input: \(error)")
|
|
} catch {
|
|
print("Error evaluating input: \(error)")
|
|
}
|
|
} catch let error as MalException {
|
|
print("Error parsing input: \(error)")
|
|
} catch {
|
|
print("Error parsing input: \(error)")
|
|
}
|
|
}
|
|
return nil
|
|
}
|
|
|
|
// Perform the full READ/EVAL/PRINT, returning a printable string.
|
|
//
|
|
private func REP(text: String, _ env: Environment) -> String? {
|
|
let exp = RE(text, env)
|
|
if exp == nil { return nil }
|
|
return PRINT(exp!)
|
|
}
|
|
|
|
// Perform the full REPL.
|
|
//
|
|
private func REPL(env: Environment) {
|
|
while true {
|
|
if let text = _readline("user> ") {
|
|
if let output = REP(text, env) {
|
|
print("\(output)")
|
|
}
|
|
} else {
|
|
print("")
|
|
break
|
|
}
|
|
}
|
|
}
|
|
|
|
// Process any command line arguments. Any trailing arguments are incorporated
|
|
// into the environment. Any argument immediately after the process name is
|
|
// taken as a script to execute. If one exists, it is executed in lieu of
|
|
// running the REPL.
|
|
//
|
|
private func process_command_line(args: [String], _ env: Environment) -> Bool {
|
|
var argv = make_list()
|
|
if args.count > 2 {
|
|
let args1 = args[2..<args.count]
|
|
let args2 = args1.map { make_string($0) }
|
|
let args3 = [MalVal](args2)
|
|
argv = make_list(args3)
|
|
}
|
|
env.set(kSymbolArgv, argv)
|
|
|
|
if args.count > 1 {
|
|
RE("(load-file \"\(args[1])\")", env)
|
|
return false
|
|
}
|
|
|
|
return true
|
|
}
|
|
|
|
func main() {
|
|
let env = Environment(outer: nil)
|
|
|
|
load_history_file()
|
|
load_builtins(env)
|
|
|
|
RE("(def! not (fn* (a) (if a false true)))", env)
|
|
RE("(def! load-file (fn* (f) (eval (read-string (str \"(do \" (slurp f) \")\")))))", env)
|
|
RE("(defmacro! cond (fn* (& xs) (if (> (count xs) 0) (list 'if (first xs) (if (> (count xs) 1) (nth xs 1) " +
|
|
"(throw \"odd number of forms to cond\")) (cons 'cond (rest (rest xs)))))))", env)
|
|
RE("(defmacro! or (fn* (& xs) (if (empty? xs) nil (if (= 1 (count xs)) (first xs) " +
|
|
"`(let* (or_FIXME ~(first xs)) (if or_FIXME or_FIXME (or ~@(rest xs))))))))", env)
|
|
|
|
env.set(kSymbolEval, make_builtin({
|
|
try! unwrap_args($0) {
|
|
(ast: MalVal) -> MalVal in
|
|
try EVAL(ast, env)
|
|
}
|
|
}))
|
|
|
|
if process_command_line(Process.arguments, env) {
|
|
REPL(env)
|
|
}
|
|
|
|
save_history_file()
|
|
}
|