Merge pull request #2890 from ozra/fix-1179-unsigned-number-literals

Fix 1179 unsigned number literals

2 files changed, 152 insertions, 47 deletions

diff --git a/compiler/lexer.nim b/compiler/lexer.nim
index 6b38ee062..5c7baf7d3 100644
--- a/compiler/lexer.nim
+++ b/compiler/lexer.nim
@@ -262,14 +262,11 @@ template eatChar(L: var TLexer, t: var TToken) =
   add(t.literal, L.buf[L.bufpos])
   inc(L.bufpos)
 
-proc getNumber(L: var TLexer): TToken =
-  var
-    startpos, endpos: int
-    xi: BiggestInt
-  const literalishChars = {   'A'..'F', 'a'..'f', '0'..'9', 'X', 'x', 'o', 'c',
-    'C', 'b', 'B', '_', '.', '\''}
-  const literalishCharsNoDot = literalishChars - {'.'}
 
+
+
+
+proc getNumber(L: var TLexer): TToken =
   proc matchUnderscoreChars(L: var TLexer, tok: var TToken, chars: set[char]) =
     var pos = L.bufpos              # use registers for pos, buf
     var buf = L.buf
@@ -318,14 +315,22 @@ proc getNumber(L: var TLexer): TToken =
     L.bufpos = msgPos
     lexMessage(L, msg, t.literal)
 
+  var
+    startpos, endpos: int
+    xi: BiggestInt
+    isBase10 = true
+  const
+    baseCodeChars = {'X', 'x', 'o', 'c', 'C', 'b', 'B'}
+    literalishChars = baseCodeChars + {'A'..'F', 'a'..'f', '0'..'9', '_', '\''}
+    floatTypes = {tkFloatLit, tkFloat32Lit, tkFloat64Lit, tkFloat128Lit}
   result.tokType = tkIntLit   # int literal until we know better
   result.literal = ""
   result.base = base10
   startpos = L.bufpos
-  var isAFloatLiteral = false
+
   # First stage: find out base, make verifications, build token literal string
-  if  L.buf[L.bufpos] == '0' and 
-      L.buf[L.bufpos + 1] in {'X', 'x', 'o', 'O', 'c', 'C', 'b', 'B'}:
+  if L.buf[L.bufpos] == '0' and L.buf[L.bufpos + 1] in baseCodeChars + {'O'}:
+    isBase10 = false
     eatChar(L, result, '0')
     case L.buf[L.bufpos]
     of 'O':
@@ -344,21 +349,23 @@ proc getNumber(L: var TLexer): TToken =
   else:
     matchUnderscoreChars(L, result, {'0'..'9'})
     if (L.buf[L.bufpos] == '.') and (L.buf[L.bufpos + 1] in {'0'..'9'}):
-      isAFloatLiteral = true
+      result.tokType = tkFloat64Lit
       eatChar(L, result, '.')
       matchUnderscoreChars(L, result, {'0'..'9'})
     if L.buf[L.bufpos] in {'e', 'E'}:
-      isAFloatLiteral = true
+      result.tokType = tkFloat64Lit
       eatChar(L, result, 'e')
       if L.buf[L.bufpos] in {'+', '-'}:
         eatChar(L, result)
       matchUnderscoreChars(L, result, {'0'..'9'})
   endpos = L.bufpos
-  # Second stage, find out if there's a datatype postfix and handle it
+
+  # Second stage, find out if there's a datatype suffix and handle it
   var postPos = endpos
   if L.buf[postPos] in {'\'', 'f', 'F', 'd', 'D', 'i', 'I', 'u', 'U'}:
-    if L.buf[postPos] == '\'': 
+    if L.buf[postPos] == '\'':
       inc(postPos)
+
     case L.buf[postPos]
     of 'f', 'F':
       inc(postPos)
@@ -410,21 +417,23 @@ proc getNumber(L: var TLexer): TToken =
         inc(postPos)
       else:
         result.tokType = tkUIntLit
-    else: 
+    else:
       lexMessageLitNum(L, errInvalidNumber, startpos)
+
   # Is there still a literalish char awaiting? Then it's an error!
-  if  L.buf[postPos] in literalishCharsNoDot or 
+  if  L.buf[postPos] in literalishChars or
      (L.buf[postPos] == '.' and L.buf[postPos + 1] in {'0'..'9'}):
     lexMessageLitNum(L, errInvalidNumber, startpos)
+
   # Third stage, extract actual number
   L.bufpos = startpos            # restore position
   var pos: int = startpos
   try:
-    if (L.buf[pos] == '0') and
-        (L.buf[pos + 1] in {'x', 'X', 'b', 'B', 'o', 'O', 'c', 'C'}):
+    if (L.buf[pos] == '0') and (L.buf[pos + 1] in baseCodeChars):
       inc(pos, 2)
       xi = 0                  # it is a base prefix
-      case L.buf[pos - 1]     # now look at the optional type suffix:
+
+      case L.buf[pos - 1]
       of 'b', 'B':
         result.base = base2
         while pos < endpos:
@@ -452,48 +461,76 @@ proc getNumber(L: var TLexer): TToken =
           of 'A'..'F':
             xi = `shl`(xi, 4) or (ord(L.buf[pos]) - ord('A') + 10)
             inc(pos)
-          else: 
+          else:
             break
-      else: 
+      else:
         internalError(getLineInfo(L), "getNumber")
+
       case result.tokType
       of tkIntLit, tkInt64Lit: result.iNumber = xi
       of tkInt8Lit: result.iNumber = BiggestInt(int8(toU8(int(xi))))
-      of tkInt16Lit: result.iNumber = BiggestInt(toU16(int(xi)))
-      of tkInt32Lit: result.iNumber = BiggestInt(toU32(xi))
+      of tkInt16Lit: result.iNumber = BiggestInt(int16(toU16(int(xi))))
+      of tkInt32Lit: result.iNumber = BiggestInt(int32(toU32(int64(xi))))
       of tkUIntLit, tkUInt64Lit: result.iNumber = xi
-      of tkUInt8Lit: result.iNumber = BiggestInt(int8(toU8(int(xi))))
-      of tkUInt16Lit: result.iNumber = BiggestInt(toU16(int(xi)))
-      of tkUInt32Lit: result.iNumber = BiggestInt(toU32(xi))
+      of tkUInt8Lit: result.iNumber = BiggestInt(uint8(toU8(int(xi))))
+      of tkUInt16Lit: result.iNumber = BiggestInt(uint16(toU16(int(xi))))
+      of tkUInt32Lit: result.iNumber = BiggestInt(uint32(toU32(int64(xi))))
       of tkFloat32Lit:
         result.fNumber = (cast[PFloat32](addr(xi)))[]
         # note: this code is endian neutral!
         # XXX: Test this on big endian machine!
       of tkFloat64Lit: result.fNumber = (cast[PFloat64](addr(xi)))[]
       else: internalError(getLineInfo(L), "getNumber")
-    elif isAFloatLiteral or (result.tokType == tkFloat32Lit) or
-        (result.tokType == tkFloat64Lit):
-      result.fNumber = parseFloat(result.literal)
-      if result.tokType == tkIntLit: result.tokType = tkFloatLit
-    elif result.tokType == tkUint64Lit:
-      xi = 0
-      let len = unsafeParseUInt(result.literal, xi)
-      if len != result.literal.len or len == 0:
-        raise newException(ValueError, "invalid integer: " & $xi)
-      result.iNumber = xi
+
+      # Bounds checks. Non decimal literals are allowed to overflow the range of
+      # the datatype as long as their pattern don't overflow _bitwise_, hence 
+      # below checks of signed sizes against uint*.high is deliberate:
+      # (0x80'u8 = 128, 0x80'i8 = -128, etc == OK)
+      if result.tokType notin floatTypes:
+        let outOfRange = case result.tokType:
+        of tkUInt8Lit, tkUInt16Lit, tkUInt32Lit: result.iNumber != xi
+        of tkInt8Lit: (xi > BiggestInt(uint8.high))
+        of tkInt16Lit: (xi > BiggestInt(uint16.high))
+        of tkInt32Lit: (xi > BiggestInt(uint32.high))
+        else: false
+
+        if outOfRange:
+          echo "out of range num: ", result.iNumber, " vs ", xi
+          lexMessageLitNum(L, errNumberOutOfRange, startpos)
+
     else:
-      result.iNumber = parseBiggestInt(result.literal)
-      if (result.iNumber < low(int32)) or (result.iNumber > high(int32)):
-        if result.tokType == tkIntLit:
+      case result.tokType
+      of floatTypes:
+        result.fNumber = parseFloat(result.literal)
+      of tkUint64Lit:
+        xi = 0
+        let len = unsafeParseUInt(result.literal, xi)
+        if len != result.literal.len or len == 0:
+          raise newException(ValueError, "invalid integer: " & $xi)
+        result.iNumber = xi
+      else:
+        result.iNumber = parseBiggestInt(result.literal)
+
+      # Explicit bounds checks
+      let outOfRange = case result.tokType:
+      of tkInt8Lit: (result.iNumber < int8.low or result.iNumber > int8.high)
+      of tkUInt8Lit: (result.iNumber < BiggestInt(uint8.low) or
+                      result.iNumber > BiggestInt(uint8.high))
+      of tkInt16Lit: (result.iNumber < int16.low or result.iNumber > int16.high)
+      of tkUInt16Lit: (result.iNumber < BiggestInt(uint16.low) or
+                      result.iNumber > BiggestInt(uint16.high))
+      of tkInt32Lit: (result.iNumber < int32.low or result.iNumber > int32.high)
+      of tkUInt32Lit: (result.iNumber < BiggestInt(uint32.low) or
+                      result.iNumber > BiggestInt(uint32.high))
+      else: false
+
+      if outOfRange: lexMessageLitNum(L, errNumberOutOfRange, startpos)
+
+    # Promote int literal to int64? Not always necessary, but more consistent
+    if result.tokType == tkIntLit:
+        if (result.iNumber < low(int32)) or (result.iNumber > high(int32)):
           result.tokType = tkInt64Lit
-        elif result.tokType in {tkInt8Lit, tkInt16Lit, tkInt32Lit}:
-          lexMessageLitNum(L, errNumberOutOfRange, startpos)
-      elif result.tokType == tkInt8Lit and
-          (result.iNumber < int8.low or result.iNumber > int8.high):
-        lexMessageLitNum(L, errNumberOutOfRange, startpos)
-      elif result.tokType == tkInt16Lit and
-          (result.iNumber < int16.low or result.iNumber > int16.high):
-        lexMessageLitNum(L, errNumberOutOfRange, startpos)
+
   except ValueError:
     lexMessageLitNum(L, errInvalidNumber, startpos)
   except OverflowError, RangeError:
diff --git a/tests/misc/tunsignedmisc.nim b/tests/misc/tunsignedmisc.nim
new file mode 100644
index 000000000..e6a497a3d
--- /dev/null
+++ b/tests/misc/tunsignedmisc.nim
@@ -0,0 +1,68 @@
+import unsigned
+
+discard """
+  errormsg: "number 0x123'u8 out of valid range"
+"""
+
+# Bug #1179
+
+# Unsigneds
+
+# 8 bit
+let ref1 = 128'u8 shr 7
+let hex1 = 0x80'u8 shr 7
+let oct1 = 0c200'u8 shr 7
+let dig1 = 0b10000000'u8 shr 7
+
+doAssert(ref1 == 1)
+doAssert(ref1 == hex1)
+doAssert(ref1 == oct1)
+doAssert(ref1 == dig1)
+
+# 16 bit
+let ref2 = 32768'u16 shr 15
+let hex2 = 0x8000'u16 shr 15
+let oct2 = 0c100000'u16 shr 15
+let dig2 = 0b1000000000000000'u16 shr 15
+
+doAssert(ref2 == 1)
+doAssert(ref2 == hex2)
+doAssert(ref2 == oct2)
+doAssert(ref2 == dig2)
+
+# 32 bit
+let ref3 = 2147483648'u32 shr 31
+let hex3 = 0x80000000'u32 shr 31
+let oct3 = 0c20000000000'u32 shr 31
+let dig3 = 0b10000000000000000000000000000000'u32 shr 31
+
+doAssert(ref3 == 1)
+doAssert(ref3 == hex3)
+doAssert(ref3 == oct3)
+doAssert(ref3 == dig3)
+
+# Below doesn't work for lexer stage errors...
+# doAssert(compiles(0xFF'u8) == true)
+# doAssert(compiles(0xFFF'u16) == true)
+# doAssert(compiles(0x7FFF'i16) == true)
+
+# doAssert(compiles(0x123'u8) == false)
+# doAssert(compiles(0x123'i8) == false)
+# doAssert(compiles(0x123123'u16) == false)
+# doAssert(compiles(0x123123'i16) == false)
+
+# Should compile #
+let boundOkHex1 = 0xFF'u8
+let boundOkHex2 = 0xFFFF'u16
+let boundOkHex3 = 0x7FFF'i16
+
+let boundOkHex4 = 0x80'i8
+let boundOkHex5 = 0xFF'i8
+let boundOkHex6 = 0xFFFF'i16
+let boundOkHex7 = 0x7FFF'i16
+
+# Should _not_ compile #
+let boundBreakingHex1 = 0x123'u8
+let boundBreakingHex2 = 0x123'i8
+let boundBreakingHex3 = 0x123123'u16
+let boundBreakingHex4 = 0x123123'i16