cell/regen.ce

// regen.ce — regenerate .mcode bytecode files and pre-warm .mach cache

var fd = use("fd")
var json = use("json")
var crypto = use("crypto")
var tokenize = use("tokenize")
var parse = use("parse")
var fold = use("fold")
var mcode = use("mcode")
var streamline = use("streamline")

var files = [
  {src: "tokenize.cm", name: "tokenize", out: "boot/tokenize.cm.mcode"},
  {src: "parse.cm", name: "parse", out: "boot/parse.cm.mcode"},
  {src: "fold.cm", name: "fold", out: "boot/fold.cm.mcode"},
  {src: "mcode.cm", name: "mcode", out: "boot/mcode.cm.mcode"},
  {src: "streamline.cm", name: "streamline", out: "boot/streamline.cm.mcode"},
  {src: "qbe.cm", name: "qbe", out: "boot/qbe.cm.mcode"},
  {src: "qbe_emit.cm", name: "qbe_emit", out: "boot/qbe_emit.cm.mcode"},
  {src: "verify_ir.cm", name: "verify_ir", out: "boot/verify_ir.cm.mcode"},
  {src: "internal/bootstrap.cm", name: "bootstrap", out: "boot/bootstrap.cm.mcode"},
  {src: "internal/engine.cm", name: "engine", out: "boot/engine.cm.mcode"},
  {src: "boot/seed_bootstrap.cm", name: "seed_bootstrap", out: "boot/seed_bootstrap.cm.mcode"}
]

// Resolve shop_path for cache writes
var os = use('os')
var shop = os.getenv('CELL_SHOP')
var home = null
var cache_dir = null
if (!shop) {
  home = os.getenv('HOME')
  if (home) {
    shop = home + '/.cell'
  }
}
if (shop) {
  cache_dir = shop + '/build'
  if (!fd.is_dir(cache_dir)) {
    fd.mkdir(cache_dir)
  }
}

var i = 0
var entry = null
var src = null
var tok_result = null
var ast = null
var folded = null
var mcode_blob = null
var hash = null
var compact_mcode = null
var mach_blob = null
var compiled = null
var optimized = null
var mcode_text = null
var f = null
var errs = null
var ei = 0
var e = null
var had_errors = false

// Collapse leaf arrays (instruction arrays) onto single lines
var compact_arrays = function(json_text) {
  var lines = array(json_text, "\n")
  var result = []
  var i = 0
  var line = null
  var trimmed = null
  var collecting = false
  var collected = null
  var indent = null
  var is_leaf = null
  var j = 0
  var inner = null
  var parts = null
  var trailing = null
  var chars = null
  var k = 0

  while (i < length(lines)) {
    line = lines[i]
    trimmed = trim(line)
    if (collecting == false && trimmed == "[") {
      collecting = true
      chars = array(line)
      k = 0
      while (k < length(chars) && chars[k] == " ") {
        k = k + 1
      }
      indent = text(line, 0, k)
      collected = []
      i = i + 1
      continue
    }
    if (collecting) {
      if (trimmed == "]" || trimmed == "],") {
        is_leaf = true
        j = 0
        while (j < length(collected)) {
          inner = trim(collected[j])
          if (starts_with(inner, "[") || starts_with(inner, "{")) {
            is_leaf = false
          }
          j = j + 1
        }
        if (is_leaf && length(collected) > 0) {
          parts = []
          j = 0
          while (j < length(collected)) {
            inner = trim(collected[j])
            if (ends_with(inner, ",")) {
              inner = text(inner, 0, length(inner) - 1)
            }
            parts[] = inner
            j = j + 1
          }
          trailing = ""
          if (ends_with(trimmed, ",")) {
            trailing = ","
          }
          result[] = `${indent}[${text(parts, ", ")}]${trailing}`
        } else {
          result[] = `${indent}[`
          j = 0
          while (j < length(collected)) {
            result[] = collected[j]
            j = j + 1
          }
          result[] = line
        }
        collecting = false
      } else {
        collected[] = line
      }
      i = i + 1
      continue
    }
    result[] = line
    i = i + 1
  }
  return text(result, "\n")
}

while (i < length(files)) {
  entry = files[i]
  src = text(fd.slurp(entry.src))
  tok_result = tokenize(src, entry.src)
  ast = parse(tok_result.tokens, src, entry.src, tokenize)
  // Check for parse/semantic errors
  errs = ast.errors
  if (errs != null && length(errs) > 0) {
    ei = 0
    while (ei < length(errs)) {
      e = errs[ei]
      if (e.line != null) {
        print(`${entry.src}:${text(e.line)}:${text(e.column)}: error: ${e.message}`)
      } else {
        print(`${entry.src}: error: ${e.message}`)
      }
      ei = ei + 1
    }
    had_errors = true
    i = i + 1
    continue
  }
  folded = fold(ast)
  compiled = mcode(folded)
  optimized = streamline(compiled)
  mcode_text = compact_arrays(json.encode(optimized, null, 2))
  f = fd.open(entry.out, "w")
  fd.write(f, mcode_text)
  fd.close(f)
  print(`wrote ${entry.out}`)
  // Pre-warm .mach cache
  if (cache_dir) {
    mcode_blob = stone(blob(mcode_text))
    hash = text(crypto.blake2(mcode_blob), 'h')
    compact_mcode = json.encode(optimized)
    mach_blob = mach_compile_mcode_bin(entry.name, compact_mcode)
    fd.slurpwrite(cache_dir + '/' + hash, mach_blob)
    print(`  cached ${hash}`)
  }
  i = i + 1
}
if (had_errors) {
  print("regen aborted: fix errors above")
}