summary refs log tree commit diff stats
path: root/compiler/semfold.nim
diff options
context:
space:
mode:
Diffstat (limited to 'compiler/semfold.nim')
-rw-r--r--compiler/semfold.nim793
1 files changed, 793 insertions, 0 deletions
diff --git a/compiler/semfold.nim b/compiler/semfold.nim
new file mode 100644
index 000000000..da24005c2
--- /dev/null
+++ b/compiler/semfold.nim
@@ -0,0 +1,793 @@
+#
+#
+#           The Nim Compiler
+#        (c) Copyright 2015 Andreas Rumpf
+#
+#    See the file "copying.txt", included in this
+#    distribution, for details about the copyright.
+#
+
+# this module folds constants; used by semantic checking phase
+# and evaluation phase
+
+import
+  strutils, lists, options, ast, astalgo, trees, treetab, nimsets, times,
+  nversion, platform, math, msgs, os, condsyms, idents, renderer, types,
+  commands, magicsys, saturate
+
+proc getConstExpr*(m: PSym, n: PNode): PNode
+  # evaluates the constant expression or returns nil if it is no constant
+  # expression
+proc evalOp*(m: TMagic, n, a, b, c: PNode): PNode
+proc leValueConv*(a, b: PNode): bool
+proc newIntNodeT*(intVal: BiggestInt, n: PNode): PNode
+proc newFloatNodeT(floatVal: BiggestFloat, n: PNode): PNode
+proc newStrNodeT*(strVal: string, n: PNode): PNode
+
+# implementation
+
+proc newIntNodeT(intVal: BiggestInt, n: PNode): PNode =
+  case skipTypes(n.typ, abstractVarRange).kind
+  of tyInt:
+    result = newIntNode(nkIntLit, intVal)
+    result.typ = getIntLitType(result)
+    # hrm, this is not correct: 1 + high(int) shouldn't produce tyInt64 ...
+    #setIntLitType(result)
+  of tyChar:
+    result = newIntNode(nkCharLit, intVal)
+    result.typ = n.typ
+  else:
+    result = newIntNode(nkIntLit, intVal)
+    result.typ = n.typ
+  result.info = n.info
+
+proc newFloatNodeT(floatVal: BiggestFloat, n: PNode): PNode =
+  result = newFloatNode(nkFloatLit, floatVal)
+  if skipTypes(n.typ, abstractVarRange).kind == tyFloat:
+    result.typ = getFloatLitType(result)
+  else:
+    result.typ = n.typ
+  result.info = n.info
+
+proc newStrNodeT(strVal: string, n: PNode): PNode =
+  result = newStrNode(nkStrLit, strVal)
+  result.typ = n.typ
+  result.info = n.info
+
+proc ordinalValToString*(a: PNode): string =
+  # because $ has the param ordinal[T], `a` is not necessarily an enum, but an
+  # ordinal
+  var x = getInt(a)
+
+  var t = skipTypes(a.typ, abstractRange)
+  case t.kind
+  of tyChar:
+    result = $chr(int(x) and 0xff)
+  of tyEnum:
+    var n = t.n
+    for i in countup(0, sonsLen(n) - 1):
+      if n.sons[i].kind != nkSym: internalError(a.info, "ordinalValToString")
+      var field = n.sons[i].sym
+      if field.position == x:
+        if field.ast == nil:
+          return field.name.s
+        else:
+          return field.ast.strVal
+    internalError(a.info, "no symbol for ordinal value: " & $x)
+  else:
+    result = $x
+
+proc isFloatRange(t: PType): bool {.inline.} =
+  result = t.kind == tyRange and t.sons[0].kind in {tyFloat..tyFloat128}
+
+proc isIntRange(t: PType): bool {.inline.} =
+  result = t.kind == tyRange and t.sons[0].kind in {
+      tyInt..tyInt64, tyUInt8..tyUInt32}
+
+proc pickIntRange(a, b: PType): PType =
+  if isIntRange(a): result = a
+  elif isIntRange(b): result = b
+  else: result = a
+
+proc isIntRangeOrLit(t: PType): bool =
+  result = isIntRange(t) or isIntLit(t)
+
+proc pickMinInt(n: PNode): BiggestInt =
+  if n.kind in {nkIntLit..nkUInt64Lit}:
+    result = n.intVal
+  elif isIntLit(n.typ):
+    result = n.typ.n.intVal
+  elif isIntRange(n.typ):
+    result = firstOrd(n.typ)
+  else:
+    internalError(n.info, "pickMinInt")
+
+proc pickMaxInt(n: PNode): BiggestInt =
+  if n.kind in {nkIntLit..nkUInt64Lit}:
+    result = n.intVal
+  elif isIntLit(n.typ):
+    result = n.typ.n.intVal
+  elif isIntRange(n.typ):
+    result = lastOrd(n.typ)
+  else:
+    internalError(n.info, "pickMaxInt")
+
+proc makeRange(typ: PType, first, last: BiggestInt): PType =
+  let minA = min(first, last)
+  let maxA = max(first, last)
+  let lowerNode = newIntNode(nkIntLit, minA)
+  if typ.kind == tyInt and minA == maxA:
+    result = getIntLitType(lowerNode)
+  elif typ.kind in {tyUint, tyUInt64}:
+    # these are not ordinal types, so you get no subrange type for these:
+    result = typ
+  else:
+    var n = newNode(nkRange)
+    addSon(n, lowerNode)
+    addSon(n, newIntNode(nkIntLit, maxA))
+    result = newType(tyRange, typ.owner)
+    result.n = n
+    addSonSkipIntLit(result, skipTypes(typ, {tyRange}))
+
+proc makeRangeF(typ: PType, first, last: BiggestFloat): PType =
+  var n = newNode(nkRange)
+  addSon(n, newFloatNode(nkFloatLit, min(first.float, last.float)))
+  addSon(n, newFloatNode(nkFloatLit, max(first.float, last.float)))
+  result = newType(tyRange, typ.owner)
+  result.n = n
+  addSonSkipIntLit(result, skipTypes(typ, {tyRange}))
+
+proc getIntervalType*(m: TMagic, n: PNode): PType =
+  # Nim requires interval arithmetic for ``range`` types. Lots of tedious
+  # work but the feature is very nice for reducing explicit conversions.
+  const ordIntLit = {nkIntLit..nkUInt64Lit}
+  result = n.typ
+
+  template commutativeOp(opr: expr) {.immediate.} =
+    let a = n.sons[1]
+    let b = n.sons[2]
+    if isIntRangeOrLit(a.typ) and isIntRangeOrLit(b.typ):
+      result = makeRange(pickIntRange(a.typ, b.typ),
+                         opr(pickMinInt(a), pickMinInt(b)),
+                         opr(pickMaxInt(a), pickMaxInt(b)))
+
+  template binaryOp(opr: expr) {.immediate.} =
+    let a = n.sons[1]
+    let b = n.sons[2]
+    if isIntRange(a.typ) and b.kind in {nkIntLit..nkUInt64Lit}:
+      result = makeRange(a.typ,
+                         opr(pickMinInt(a), pickMinInt(b)),
+                         opr(pickMaxInt(a), pickMaxInt(b)))
+
+  case m
+  of mUnaryMinusI, mUnaryMinusI64:
+    let a = n.sons[1].typ
+    if isIntRange(a):
+      # (1..3) * (-1) == (-3.. -1)
+      result = makeRange(a, 0|-|lastOrd(a), 0|-|firstOrd(a))
+  of mUnaryMinusF64:
+    let a = n.sons[1].typ
+    if isFloatRange(a):
+      result = makeRangeF(a, -getFloat(a.n.sons[1]),
+                             -getFloat(a.n.sons[0]))
+  of mAbsF64:
+    let a = n.sons[1].typ
+    if isFloatRange(a):
+      # abs(-5.. 1) == (1..5)
+      if a.n[0].floatVal <= 0.0:
+        result = makeRangeF(a, 0.0, abs(getFloat(a.n.sons[0])))
+      else:
+        result = makeRangeF(a, abs(getFloat(a.n.sons[1])),
+                               abs(getFloat(a.n.sons[0])))
+  of mAbsI, mAbsI64:
+    let a = n.sons[1].typ
+    if isIntRange(a):
+      if a.n[0].intVal <= 0:
+        result = makeRange(a, 0, `|abs|`(getInt(a.n.sons[0])))
+      else:
+        result = makeRange(a, `|abs|`(getInt(a.n.sons[1])),
+                              `|abs|`(getInt(a.n.sons[0])))
+  of mSucc:
+    let a = n.sons[1].typ
+    let b = n.sons[2].typ
+    if isIntRange(a) and isIntLit(b):
+      # (-5.. 1) + 6 == (-5 + 6)..(-1 + 6)
+      result = makeRange(a, pickMinInt(n.sons[1]) |+| pickMinInt(n.sons[2]),
+                            pickMaxInt(n.sons[1]) |+| pickMaxInt(n.sons[2]))
+  of mPred:
+    let a = n.sons[1].typ
+    let b = n.sons[2].typ
+    if isIntRange(a) and isIntLit(b):
+      result = makeRange(a, pickMinInt(n.sons[1]) |-| pickMinInt(n.sons[2]),
+                            pickMaxInt(n.sons[1]) |-| pickMaxInt(n.sons[2]))
+  of mAddI, mAddI64, mAddU:
+    commutativeOp(`|+|`)
+  of mMulI, mMulI64, mMulU:
+    commutativeOp(`|*|`)
+  of mSubI, mSubI64, mSubU:
+    binaryOp(`|-|`)
+  of mBitandI, mBitandI64:
+    # since uint64 is still not even valid for 'range' (since it's no ordinal
+    # yet), we exclude it from the list (see bug #1638) for now:
+    var a = n.sons[1]
+    var b = n.sons[2]
+    # symmetrical:
+    if b.kind notin ordIntLit: swap(a, b)
+    if b.kind in ordIntLit:
+      let x = b.intVal|+|1
+      if (x and -x) == x and x >= 0:
+        result = makeRange(a.typ, 0, b.intVal)
+  of mModU:
+    let a = n.sons[1]
+    let b = n.sons[2]
+    if a.kind in ordIntLit:
+      if b.intVal >= 0:
+        result = makeRange(a.typ, 0, b.intVal-1)
+      else:
+        result = makeRange(a.typ, b.intVal+1, 0)
+  of mModI, mModI64:
+    # so ... if you ever wondered about modulo's signedness; this defines it:
+    let a = n.sons[1]
+    let b = n.sons[2]
+    if b.kind in {nkIntLit..nkUInt64Lit}:
+      if b.intVal >= 0:
+        result = makeRange(a.typ, -(b.intVal-1), b.intVal-1)
+      else:
+        result = makeRange(a.typ, b.intVal+1, -(b.intVal+1))
+  of mDivI, mDivI64, mDivU:
+    binaryOp(`|div|`)
+  of mMinI:
+    commutativeOp(min)
+  of mMaxI:
+    commutativeOp(max)
+  else: discard
+
+discard """
+  mShlI, mShlI64,
+  mShrI, mShrI64, mAddF64, mSubF64, mMulF64, mDivF64, mMaxF64, mMinF64
+"""
+
+proc evalIs(n, a: PNode): PNode =
+  # XXX: This should use the standard isOpImpl
+  internalAssert a.kind == nkSym and a.sym.kind == skType
+  internalAssert n.sonsLen == 3 and
+    n[2].kind in {nkStrLit..nkTripleStrLit, nkType}
+
+  let t1 = a.sym.typ
+
+  if n[2].kind in {nkStrLit..nkTripleStrLit}:
+    case n[2].strVal.normalize
+    of "closure":
+      let t = skipTypes(t1, abstractRange)
+      result = newIntNode(nkIntLit, ord(t.kind == tyProc and
+                                        t.callConv == ccClosure and
+                                        tfIterator notin t.flags))
+    of "iterator":
+      let t = skipTypes(t1, abstractRange)
+      result = newIntNode(nkIntLit, ord(t.kind == tyProc and
+                                        t.callConv == ccClosure and
+                                        tfIterator in t.flags))
+    else: discard
+  else:
+    # XXX semexprs.isOpImpl is slightly different and requires a context. yay.
+    let t2 = n[2].typ
+    var match = sameType(t1, t2)
+    result = newIntNode(nkIntLit, ord(match))
+  result.typ = n.typ
+
+proc evalOp(m: TMagic, n, a, b, c: PNode): PNode =
+  # b and c may be nil
+  result = nil
+  case m
+  of mOrd: result = newIntNodeT(getOrdValue(a), n)
+  of mChr: result = newIntNodeT(getInt(a), n)
+  of mUnaryMinusI, mUnaryMinusI64: result = newIntNodeT(- getInt(a), n)
+  of mUnaryMinusF64: result = newFloatNodeT(- getFloat(a), n)
+  of mNot: result = newIntNodeT(1 - getInt(a), n)
+  of mCard: result = newIntNodeT(nimsets.cardSet(a), n)
+  of mBitnotI, mBitnotI64: result = newIntNodeT(not getInt(a), n)
+  of mLengthStr, mXLenStr:
+    if a.kind == nkNilLit: result = newIntNodeT(0, n)
+    else: result = newIntNodeT(len(getStr(a)), n)
+  of mLengthArray: result = newIntNodeT(lengthOrd(a.typ), n)
+  of mLengthSeq, mLengthOpenArray, mXLenSeq:
+    if a.kind == nkNilLit: result = newIntNodeT(0, n)
+    else: result = newIntNodeT(sonsLen(a), n) # BUGFIX
+  of mUnaryPlusI, mUnaryPlusF64: result = a # throw `+` away
+  of mToFloat, mToBiggestFloat:
+    result = newFloatNodeT(toFloat(int(getInt(a))), n)
+  of mToInt, mToBiggestInt: result = newIntNodeT(system.toInt(getFloat(a)), n)
+  of mAbsF64: result = newFloatNodeT(abs(getFloat(a)), n)
+  of mAbsI, mAbsI64:
+    if getInt(a) >= 0: result = a
+    else: result = newIntNodeT(- getInt(a), n)
+  of mZe8ToI, mZe8ToI64, mZe16ToI, mZe16ToI64, mZe32ToI64, mZeIToI64:
+    # byte(-128) = 1...1..1000_0000'64 --> 0...0..1000_0000'64
+    result = newIntNodeT(getInt(a) and (`shl`(1, getSize(a.typ) * 8) - 1), n)
+  of mToU8: result = newIntNodeT(getInt(a) and 0x000000FF, n)
+  of mToU16: result = newIntNodeT(getInt(a) and 0x0000FFFF, n)
+  of mToU32: result = newIntNodeT(getInt(a) and 0x00000000FFFFFFFF'i64, n)
+  of mUnaryLt: result = newIntNodeT(getOrdValue(a) - 1, n)
+  of mSucc: result = newIntNodeT(getOrdValue(a) + getInt(b), n)
+  of mPred: result = newIntNodeT(getOrdValue(a) - getInt(b), n)
+  of mAddI, mAddI64: result = newIntNodeT(getInt(a) + getInt(b), n)
+  of mSubI, mSubI64: result = newIntNodeT(getInt(a) - getInt(b), n)
+  of mMulI, mMulI64: result = newIntNodeT(getInt(a) * getInt(b), n)
+  of mMinI:
+    if getInt(a) > getInt(b): result = newIntNodeT(getInt(b), n)
+    else: result = newIntNodeT(getInt(a), n)
+  of mMaxI:
+    if getInt(a) > getInt(b): result = newIntNodeT(getInt(a), n)
+    else: result = newIntNodeT(getInt(b), n)
+  of mShlI, mShlI64:
+    case skipTypes(n.typ, abstractRange).kind
+    of tyInt8: result = newIntNodeT(int8(getInt(a)) shl int8(getInt(b)), n)
+    of tyInt16: result = newIntNodeT(int16(getInt(a)) shl int16(getInt(b)), n)
+    of tyInt32: result = newIntNodeT(int32(getInt(a)) shl int32(getInt(b)), n)
+    of tyInt64, tyInt, tyUInt..tyUInt64:
+      result = newIntNodeT(`shl`(getInt(a), getInt(b)), n)
+    else: internalError(n.info, "constant folding for shl")
+  of mShrI, mShrI64:
+    case skipTypes(n.typ, abstractRange).kind
+    of tyInt8: result = newIntNodeT(int8(getInt(a)) shr int8(getInt(b)), n)
+    of tyInt16: result = newIntNodeT(int16(getInt(a)) shr int16(getInt(b)), n)
+    of tyInt32: result = newIntNodeT(int32(getInt(a)) shr int32(getInt(b)), n)
+    of tyInt64, tyInt, tyUInt..tyUInt64:
+      result = newIntNodeT(`shr`(getInt(a), getInt(b)), n)
+    else: internalError(n.info, "constant folding for shr")
+  of mDivI, mDivI64:
+    let y = getInt(b)
+    if y != 0:
+      result = newIntNodeT(getInt(a) div y, n)
+  of mModI, mModI64:
+    let y = getInt(b)
+    if y != 0:
+      result = newIntNodeT(getInt(a) mod y, n)
+  of mAddF64: result = newFloatNodeT(getFloat(a) + getFloat(b), n)
+  of mSubF64: result = newFloatNodeT(getFloat(a) - getFloat(b), n)
+  of mMulF64: result = newFloatNodeT(getFloat(a) * getFloat(b), n)
+  of mDivF64:
+    if getFloat(b) == 0.0:
+      if getFloat(a) == 0.0: result = newFloatNodeT(NaN, n)
+      else: result = newFloatNodeT(Inf, n)
+    else:
+      result = newFloatNodeT(getFloat(a) / getFloat(b), n)
+  of mMaxF64:
+    if getFloat(a) > getFloat(b): result = newFloatNodeT(getFloat(a), n)
+    else: result = newFloatNodeT(getFloat(b), n)
+  of mMinF64:
+    if getFloat(a) > getFloat(b): result = newFloatNodeT(getFloat(b), n)
+    else: result = newFloatNodeT(getFloat(a), n)
+  of mIsNil: result = newIntNodeT(ord(a.kind == nkNilLit), n)
+  of mLtI, mLtI64, mLtB, mLtEnum, mLtCh:
+    result = newIntNodeT(ord(getOrdValue(a) < getOrdValue(b)), n)
+  of mLeI, mLeI64, mLeB, mLeEnum, mLeCh:
+    result = newIntNodeT(ord(getOrdValue(a) <= getOrdValue(b)), n)
+  of mEqI, mEqI64, mEqB, mEqEnum, mEqCh:
+    result = newIntNodeT(ord(getOrdValue(a) == getOrdValue(b)), n)
+  of mLtF64: result = newIntNodeT(ord(getFloat(a) < getFloat(b)), n)
+  of mLeF64: result = newIntNodeT(ord(getFloat(a) <= getFloat(b)), n)
+  of mEqF64: result = newIntNodeT(ord(getFloat(a) == getFloat(b)), n)
+  of mLtStr: result = newIntNodeT(ord(getStr(a) < getStr(b)), n)
+  of mLeStr: result = newIntNodeT(ord(getStr(a) <= getStr(b)), n)
+  of mEqStr: result = newIntNodeT(ord(getStr(a) == getStr(b)), n)
+  of mLtU, mLtU64:
+    result = newIntNodeT(ord(`<%`(getOrdValue(a), getOrdValue(b))), n)
+  of mLeU, mLeU64:
+    result = newIntNodeT(ord(`<=%`(getOrdValue(a), getOrdValue(b))), n)
+  of mBitandI, mBitandI64, mAnd: result = newIntNodeT(a.getInt and b.getInt, n)
+  of mBitorI, mBitorI64, mOr: result = newIntNodeT(getInt(a) or getInt(b), n)
+  of mBitxorI, mBitxorI64, mXor: result = newIntNodeT(a.getInt xor b.getInt, n)
+  of mAddU: result = newIntNodeT(`+%`(getInt(a), getInt(b)), n)
+  of mSubU: result = newIntNodeT(`-%`(getInt(a), getInt(b)), n)
+  of mMulU: result = newIntNodeT(`*%`(getInt(a), getInt(b)), n)
+  of mModU:
+    let y = getInt(b)
+    if y != 0:
+      result = newIntNodeT(`%%`(getInt(a), y), n)
+  of mDivU:
+    let y = getInt(b)
+    if y != 0:
+      result = newIntNodeT(`/%`(getInt(a), y), n)
+  of mLeSet: result = newIntNodeT(ord(containsSets(a, b)), n)
+  of mEqSet: result = newIntNodeT(ord(equalSets(a, b)), n)
+  of mLtSet:
+    result = newIntNodeT(ord(containsSets(a, b) and not equalSets(a, b)), n)
+  of mMulSet:
+    result = nimsets.intersectSets(a, b)
+    result.info = n.info
+  of mPlusSet:
+    result = nimsets.unionSets(a, b)
+    result.info = n.info
+  of mMinusSet:
+    result = nimsets.diffSets(a, b)
+    result.info = n.info
+  of mSymDiffSet:
+    result = nimsets.symdiffSets(a, b)
+    result.info = n.info
+  of mConStrStr: result = newStrNodeT(getStrOrChar(a) & getStrOrChar(b), n)
+  of mInSet: result = newIntNodeT(ord(inSet(a, b)), n)
+  of mRepr:
+    # BUGFIX: we cannot eval mRepr here for reasons that I forgot.
+    discard
+  of mIntToStr, mInt64ToStr: result = newStrNodeT($(getOrdValue(a)), n)
+  of mBoolToStr:
+    if getOrdValue(a) == 0: result = newStrNodeT("false", n)
+    else: result = newStrNodeT("true", n)
+  of mCopyStr: result = newStrNodeT(substr(getStr(a), int(getOrdValue(b))), n)
+  of mCopyStrLast:
+    result = newStrNodeT(substr(getStr(a), int(getOrdValue(b)),
+                                           int(getOrdValue(c))), n)
+  of mFloatToStr: result = newStrNodeT($getFloat(a), n)
+  of mCStrToStr, mCharToStr: result = newStrNodeT(getStrOrChar(a), n)
+  of mStrToStr: result = a
+  of mEnumToStr: result = newStrNodeT(ordinalValToString(a), n)
+  of mArrToSeq:
+    result = copyTree(a)
+    result.typ = n.typ
+  of mCompileOption:
+    result = newIntNodeT(ord(commands.testCompileOption(a.getStr, n.info)), n)
+  of mCompileOptionArg:
+    result = newIntNodeT(ord(
+      testCompileOptionArg(getStr(a), getStr(b), n.info)), n)
+  of mNewString, mNewStringOfCap,
+     mExit, mInc, ast.mDec, mEcho, mSwap, mAppendStrCh,
+     mAppendStrStr, mAppendSeqElem, mSetLengthStr, mSetLengthSeq,
+     mParseExprToAst, mParseStmtToAst, mExpandToAst, mTypeTrait, mDotDot,
+     mNLen..mNError, mEqRef, mSlurp, mStaticExec, mNGenSym, mSpawn,
+     mParallel, mPlugin:
+    discard
+  else: internalError(a.info, "evalOp(" & $m & ')')
+
+proc getConstIfExpr(c: PSym, n: PNode): PNode =
+  result = nil
+  for i in countup(0, sonsLen(n) - 1):
+    var it = n.sons[i]
+    if it.len == 2:
+      var e = getConstExpr(c, it.sons[0])
+      if e == nil: return nil
+      if getOrdValue(e) != 0:
+        if result == nil:
+          result = getConstExpr(c, it.sons[1])
+          if result == nil: return
+    elif it.len == 1:
+      if result == nil: result = getConstExpr(c, it.sons[0])
+    else: internalError(it.info, "getConstIfExpr()")
+
+proc partialAndExpr(c: PSym, n: PNode): PNode =
+  # partial evaluation
+  result = n
+  var a = getConstExpr(c, n.sons[1])
+  var b = getConstExpr(c, n.sons[2])
+  if a != nil:
+    if getInt(a) == 0: result = a
+    elif b != nil: result = b
+    else: result = n.sons[2]
+  elif b != nil:
+    if getInt(b) == 0: result = b
+    else: result = n.sons[1]
+
+proc partialOrExpr(c: PSym, n: PNode): PNode =
+  # partial evaluation
+  result = n
+  var a = getConstExpr(c, n.sons[1])
+  var b = getConstExpr(c, n.sons[2])
+  if a != nil:
+    if getInt(a) != 0: result = a
+    elif b != nil: result = b
+    else: result = n.sons[2]
+  elif b != nil:
+    if getInt(b) != 0: result = b
+    else: result = n.sons[1]
+
+proc leValueConv(a, b: PNode): bool =
+  result = false
+  case a.kind
+  of nkCharLit..nkUInt64Lit:
+    case b.kind
+    of nkCharLit..nkUInt64Lit: result = a.intVal <= b.intVal
+    of nkFloatLit..nkFloat128Lit: result = a.intVal <= round(b.floatVal)
+    else: internalError(a.info, "leValueConv")
+  of nkFloatLit..nkFloat128Lit:
+    case b.kind
+    of nkFloatLit..nkFloat128Lit: result = a.floatVal <= b.floatVal
+    of nkCharLit..nkUInt64Lit: result = a.floatVal <= toFloat(int(b.intVal))
+    else: internalError(a.info, "leValueConv")
+  else: internalError(a.info, "leValueConv")
+
+proc magicCall(m: PSym, n: PNode): PNode =
+  if sonsLen(n) <= 1: return
+
+  var s = n.sons[0].sym
+  var a = getConstExpr(m, n.sons[1])
+  var b, c: PNode
+  if a == nil: return
+  if sonsLen(n) > 2:
+    b = getConstExpr(m, n.sons[2])
+    if b == nil: return
+    if sonsLen(n) > 3:
+      c = getConstExpr(m, n.sons[3])
+      if c == nil: return
+  result = evalOp(s.magic, n, a, b, c)
+
+proc getAppType(n: PNode): PNode =
+  if gGlobalOptions.contains(optGenDynLib):
+    result = newStrNodeT("lib", n)
+  elif gGlobalOptions.contains(optGenStaticLib):
+    result = newStrNodeT("staticlib", n)
+  elif gGlobalOptions.contains(optGenGuiApp):
+    result = newStrNodeT("gui", n)
+  else:
+    result = newStrNodeT("console", n)
+
+proc rangeCheck(n: PNode, value: BiggestInt) =
+  if value < firstOrd(n.typ) or value > lastOrd(n.typ):
+    localError(n.info, errGenerated, "cannot convert " & $value &
+                                     " to " & typeToString(n.typ))
+
+proc foldConv*(n, a: PNode; check = false): PNode =
+  # XXX range checks?
+  case skipTypes(n.typ, abstractRange).kind
+  of tyInt..tyInt64, tyUInt..tyUInt64:
+    case skipTypes(a.typ, abstractRange).kind
+    of tyFloat..tyFloat64:
+      result = newIntNodeT(int(getFloat(a)), n)
+    of tyChar: result = newIntNodeT(getOrdValue(a), n)
+    else:
+      result = a
+      result.typ = n.typ
+    if check: rangeCheck(n, result.intVal)
+  of tyFloat..tyFloat64:
+    case skipTypes(a.typ, abstractRange).kind
+    of tyInt..tyInt64, tyEnum, tyBool, tyChar:
+      result = newFloatNodeT(toFloat(int(getOrdValue(a))), n)
+    else:
+      result = a
+      result.typ = n.typ
+  of tyOpenArray, tyVarargs, tyProc:
+    discard
+  else:
+    result = a
+    result.typ = n.typ
+
+proc getArrayConstr(m: PSym, n: PNode): PNode =
+  if n.kind == nkBracket:
+    result = n
+  else:
+    result = getConstExpr(m, n)
+    if result == nil: result = n
+
+proc foldArrayAccess(m: PSym, n: PNode): PNode =
+  var x = getConstExpr(m, n.sons[0])
+  if x == nil or x.typ.skipTypes({tyGenericInst}).kind == tyTypeDesc: return
+
+  var y = getConstExpr(m, n.sons[1])
+  if y == nil: return
+
+  var idx = getOrdValue(y)
+  case x.kind
+  of nkPar:
+    if idx >= 0 and idx < sonsLen(x):
+      result = x.sons[int(idx)]
+      if result.kind == nkExprColonExpr: result = result.sons[1]
+    else:
+      localError(n.info, errIndexOutOfBounds)
+  of nkBracket:
+    idx = idx - x.typ.firstOrd
+    if idx >= 0 and idx < x.len: result = x.sons[int(idx)]
+    else: localError(n.info, errIndexOutOfBounds)
+  of nkStrLit..nkTripleStrLit:
+    result = newNodeIT(nkCharLit, x.info, n.typ)
+    if idx >= 0 and idx < len(x.strVal):
+      result.intVal = ord(x.strVal[int(idx)])
+    elif idx == len(x.strVal):
+      discard
+    else:
+      localError(n.info, errIndexOutOfBounds)
+  else: discard
+
+proc foldFieldAccess(m: PSym, n: PNode): PNode =
+  # a real field access; proc calls have already been transformed
+  var x = getConstExpr(m, n.sons[0])
+  if x == nil or x.kind notin {nkObjConstr, nkPar}: return
+
+  var field = n.sons[1].sym
+  for i in countup(ord(x.kind == nkObjConstr), sonsLen(x) - 1):
+    var it = x.sons[i]
+    if it.kind != nkExprColonExpr:
+      # lookup per index:
+      result = x.sons[field.position]
+      if result.kind == nkExprColonExpr: result = result.sons[1]
+      return
+    if it.sons[0].sym.name.id == field.name.id:
+      result = x.sons[i].sons[1]
+      return
+  localError(n.info, errFieldXNotFound, field.name.s)
+
+proc foldConStrStr(m: PSym, n: PNode): PNode =
+  result = newNodeIT(nkStrLit, n.info, n.typ)
+  result.strVal = ""
+  for i in countup(1, sonsLen(n) - 1):
+    let a = getConstExpr(m, n.sons[i])
+    if a == nil: return nil
+    result.strVal.add(getStrOrChar(a))
+
+proc newSymNodeTypeDesc*(s: PSym; info: TLineInfo): PNode =
+  result = newSymNode(s, info)
+  result.typ = newType(tyTypeDesc, s.owner)
+  result.typ.addSonSkipIntLit(s.typ)
+
+proc getConstExpr(m: PSym, n: PNode): PNode =
+  result = nil
+  case n.kind
+  of nkSym:
+    var s = n.sym
+    case s.kind
+    of skEnumField:
+      result = newIntNodeT(s.position, n)
+    of skConst:
+      case s.magic
+      of mIsMainModule: result = newIntNodeT(ord(sfMainModule in m.flags), n)
+      of mCompileDate: result = newStrNodeT(times.getDateStr(), n)
+      of mCompileTime: result = newStrNodeT(times.getClockStr(), n)
+      of mCpuEndian: result = newIntNodeT(ord(CPU[targetCPU].endian), n)
+      of mHostOS: result = newStrNodeT(toLower(platform.OS[targetOS].name), n)
+      of mHostCPU: result = newStrNodeT(platform.CPU[targetCPU].name.toLower, n)
+      of mAppType: result = getAppType(n)
+      of mNaN: result = newFloatNodeT(NaN, n)
+      of mInf: result = newFloatNodeT(Inf, n)
+      of mNegInf: result = newFloatNodeT(NegInf, n)
+      else:
+        if sfFakeConst notin s.flags: result = copyTree(s.ast)
+    of {skProc, skMethod}:
+      result = n
+    of skType:
+      result = newSymNodeTypeDesc(s, n.info)
+    of skGenericParam:
+      if s.typ.kind == tyStatic:
+        if s.typ.n != nil:
+          result = s.typ.n
+          result.typ = s.typ.sons[0]
+      else:
+        result = newSymNodeTypeDesc(s, n.info)
+    else: discard
+  of nkCharLit..nkNilLit:
+    result = copyNode(n)
+  of nkIfExpr:
+    result = getConstIfExpr(m, n)
+  of nkCallKinds:
+    if n.sons[0].kind != nkSym: return
+    var s = n.sons[0].sym
+    if s.kind != skProc: return
+    try:
+      case s.magic
+      of mNone:
+        # If it has no sideEffect, it should be evaluated. But not here.
+        return
+      of mSizeOf:
+        var a = n.sons[1]
+        if computeSize(a.typ) < 0:
+          localError(a.info, errCannotEvalXBecauseIncompletelyDefined,
+                     "sizeof")
+          result = nil
+        elif skipTypes(a.typ, typedescInst).kind in
+             IntegralTypes+NilableTypes+{tySet}:
+          #{tyArray,tyObject,tyTuple}:
+          result = newIntNodeT(getSize(a.typ), n)
+        else:
+          result = nil
+          # XXX: size computation for complex types is still wrong
+      of mLow:
+        result = newIntNodeT(firstOrd(n.sons[1].typ), n)
+      of mHigh:
+        if skipTypes(n.sons[1].typ, abstractVar).kind notin
+            {tySequence, tyString, tyCString, tyOpenArray, tyVarargs}:
+          result = newIntNodeT(lastOrd(skipTypes(n[1].typ, abstractVar)), n)
+        else:
+          var a = getArrayConstr(m, n.sons[1])
+          if a.kind == nkBracket:
+            # we can optimize it away:
+            result = newIntNodeT(sonsLen(a)-1, n)
+      of mLengthOpenArray:
+        var a = getArrayConstr(m, n.sons[1])
+        if a.kind == nkBracket:
+          # we can optimize it away! This fixes the bug ``len(134)``.
+          result = newIntNodeT(sonsLen(a), n)
+        else:
+          result = magicCall(m, n)
+      of mLengthArray:
+        # It doesn't matter if the argument is const or not for mLengthArray.
+        # This fixes bug #544.
+        result = newIntNodeT(lengthOrd(n.sons[1].typ), n)
+      of mAstToStr:
+        result = newStrNodeT(renderTree(n[1], {renderNoComments}), n)
+      of mConStrStr:
+        result = foldConStrStr(m, n)
+      of mIs:
+        let a = getConstExpr(m, n[1])
+        if a != nil and a.kind == nkSym and a.sym.kind == skType:
+          result = evalIs(n, a)
+      else:
+        result = magicCall(m, n)
+    except OverflowError:
+      localError(n.info, errOverOrUnderflow)
+    except DivByZeroError:
+      localError(n.info, errConstantDivisionByZero)
+  of nkAddr:
+    var a = getConstExpr(m, n.sons[0])
+    if a != nil:
+      result = n
+      n.sons[0] = a
+  of nkBracket:
+    result = copyTree(n)
+    for i in countup(0, sonsLen(n) - 1):
+      var a = getConstExpr(m, n.sons[i])
+      if a == nil: return nil
+      result.sons[i] = a
+    incl(result.flags, nfAllConst)
+  of nkRange:
+    var a = getConstExpr(m, n.sons[0])
+    if a == nil: return
+    var b = getConstExpr(m, n.sons[1])
+    if b == nil: return
+    result = copyNode(n)
+    addSon(result, a)
+    addSon(result, b)
+  of nkCurly:
+    result = copyTree(n)
+    for i in countup(0, sonsLen(n) - 1):
+      var a = getConstExpr(m, n.sons[i])
+      if a == nil: return nil
+      result.sons[i] = a
+    incl(result.flags, nfAllConst)
+  of nkObjConstr:
+    result = copyTree(n)
+    for i in countup(1, sonsLen(n) - 1):
+      var a = getConstExpr(m, n.sons[i].sons[1])
+      if a == nil: return nil
+      result.sons[i].sons[1] = a
+    incl(result.flags, nfAllConst)
+  of nkPar:
+    # tuple constructor
+    result = copyTree(n)
+    if (sonsLen(n) > 0) and (n.sons[0].kind == nkExprColonExpr):
+      for i in countup(0, sonsLen(n) - 1):
+        var a = getConstExpr(m, n.sons[i].sons[1])
+        if a == nil: return nil
+        result.sons[i].sons[1] = a
+    else:
+      for i in countup(0, sonsLen(n) - 1):
+        var a = getConstExpr(m, n.sons[i])
+        if a == nil: return nil
+        result.sons[i] = a
+    incl(result.flags, nfAllConst)
+  of nkChckRangeF, nkChckRange64, nkChckRange:
+    var a = getConstExpr(m, n.sons[0])
+    if a == nil: return
+    if leValueConv(n.sons[1], a) and leValueConv(a, n.sons[2]):
+      result = a              # a <= x and x <= b
+      result.typ = n.typ
+    else:
+      localError(n.info, errGenerated, `%`(
+          msgKindToString(errIllegalConvFromXtoY),
+          [typeToString(n.sons[0].typ), typeToString(n.typ)]))
+  of nkStringToCString, nkCStringToString:
+    var a = getConstExpr(m, n.sons[0])
+    if a == nil: return
+    result = a
+    result.typ = n.typ
+  of nkHiddenStdConv, nkHiddenSubConv, nkConv:
+    var a = getConstExpr(m, n.sons[1])
+    if a == nil: return
+    result = foldConv(n, a, check=n.kind == nkHiddenStdConv)
+  of nkCast:
+    var a = getConstExpr(m, n.sons[1])
+    if a == nil: return
+    if n.typ.kind in NilableTypes:
+      # we allow compile-time 'cast' for pointer types:
+      result = a
+      result.typ = n.typ
+  of nkBracketExpr: result = foldArrayAccess(m, n)
+  of nkDotExpr: result = foldFieldAccess(m, n)
+  else:
+    discard