fixed tests for addition/subtraction

Author: pietroppeter

src/bigints.nim

@@ -310,7 +310,9 @@ proc unsignedSubtraction(a: var BigInt, b, c: BigInt) =
       negate(a)
 
 proc additionInt(a: var BigInt, b: BigInt, c: int32) =
-  if Negative in b.flags:
+  if b.isZero:
+    a = c.initBigInt
+  elif Negative in b.flags:
     if c < 0:
       unsignedAdditionInt(a, b, c)
       a.flags.incl(Negative)
@@ -358,7 +360,9 @@ template optAddInt*{x = y + z}(x,y: BigInt, z: int32) = additionInt(x, y, z)
 template optAdd*{x = y + z}(x,y,z: BigInt) = addition(x, y, z)
 
 proc subtractionInt(a: var BigInt, b: BigInt, c: int32) =
-  if Negative in b.flags:
+  if b.isZero:
+    a = (-c).initBigInt
+  elif Negative in b.flags:
     if c < 0:
       # TODO: is this right?
       unsignedSubtractionInt(a, b, c)

tests/tester.nim

@@ -118,7 +118,6 @@ test "empty limbs when uninitialized (https://github.com/def-/nim-bigints/issues
   # logic around sign might also play a role
   var
     zeroEmpty: BigInt # should be treated as zero, same with -zeroEmpty
-    result: BigInt
   let
     zeroInt32: int32 = 0
     oneInt32: int32 = 1
@@ -156,10 +155,10 @@ test "empty limbs when uninitialized (https://github.com/def-/nim-bigints/issues
   check -zeroEmpty == zero # error: fixed
 
   # proc unsignedAdditionInt(a: var BigInt, b: BigInt, c: int32)
-  check zeroEmpty + 1.int32 == one # fatal[IndexError] in bigints.nim(181) unsignedAdditionInt
-  check -zeroEmpty + 1.int32 == one # fatal[IndexError] in bigints.nim(245) unsignedSubtractionInt
-  check zeroEmpty + (-1).int32 == -one # fatal[IndexError] in bigints.nim(245) unsignedSubtractionInt
-  check -zeroEmpty + (-1).int32 == -one # fatal[IndexError] in bigints.nim(181) unsignedAdditionInt
+  check zeroEmpty + 1.int32 == one # fixed: fatal[IndexError] in bigints.nim(181) unsignedAdditionInt
+  check -zeroEmpty + 1.int32 == one # fixed: fatal[IndexError] in bigints.nim(245) unsignedSubtractionInt
+  check zeroEmpty + (-1).int32 == -one # fixed: fatal[IndexError] in bigints.nim(245) unsignedSubtractionInt
+  check -zeroEmpty + (-1).int32 == -one # fixed: fatal[IndexError] in bigints.nim(181) unsignedAdditionInt
 
   # proc unsignedAddition(a: var BigInt, b, c: BigInt)
   check zeroEmpty + one == one # ok
@@ -180,10 +179,10 @@ test "empty limbs when uninitialized (https://github.com/def-/nim-bigints/issues
   check -zeroEmpty + -bigOne == -bigOne # ok
 
   # proc unsignedSubtractionInt(a: var BigInt, b: BigInt, c: int32)
-  check zeroEmpty - 1.int32 == -one # fatal[IndexError] in bigints.nim(245) unsignedSubtractionInt
-  check -zeroEmpty - 1.int32 == -one # fatal[IndexError] in bigints.nim(181) unsignedAdditionInt
-  check zeroEmpty - (-1).int32 == one # fatal[IndexError] in bigints.nim(181) unsignedAdditionInt
-  check -zeroEmpty - (-1).int32 == one # fatal[IndexError] in bigints.nim(245) unsignedSubtractionInt
+  check zeroEmpty - 1.int32 == -one # fixed: fatal[IndexError] in bigints.nim(245) unsignedSubtractionInt
+  check -zeroEmpty - 1.int32 == -one # fixed: fatal[IndexError] in bigints.nim(181) unsignedAdditionInt
+  check zeroEmpty - (-1).int32 == one # fixed: fatal[IndexError] in bigints.nim(181) unsignedAdditionInt
+  check -zeroEmpty - (-1).int32 == one # fixed: fatal[IndexError] in bigints.nim(245) unsignedSubtractionInt
 
   # proc unsignedSubtraction(a: var BigInt, b, c: BigInt)
   check zeroEmpty - one == -one # ok