Issue 606 fix

src/arraymancer/tensor/init_cpu.nim

@@ -196,11 +196,20 @@ proc arange*[T: SomeNumber](start, stop, step: T): Tensor[T] {.noinit.} =
 
 template arange*[T: SomeNumber](stop: T): Tensor[T] =
   # Error messages of templates are very poor
-  arange(T(0), stop, T(1))
+  # While it seems that we could do `arange(T(0), stop, T(1))`
+  # that does not work well (https://github.com/mratsim/Arraymancer/issues/606)
+  # Instead we must do this:
+  type TT = typeof(stop)
+  const zero = TT(0)
+  const one = TT(1)
+  arange(zero, stop, one)
 
 template arange*[T: SomeNumber](start, stop: T): Tensor[T] =
   # Error messages of templates are very poor
-  arange(start, stop, T(1))
+  type TT = typeof(stop)
+  const one = TT(1)
+  arange(start, stop, one)
+
 
 template randomTensorCpu[T](t: Tensor[T], shape: varargs[int], max_or_range: typed): untyped =
   var size: int

tests/tensor/test_init.nim

@@ -157,6 +157,13 @@ proc main() =
       block:
         let t = arange(1.0,2.5,0.5)
         check: t == [1.0,1.5,2.0].toTensor()
+      # Test for Issue #606
+      block:
+        let numItems = 3
+        let start = 2
+        let v = @[10, 11, 12, 13, 14, 15].toTensor
+        let w = v[start +. arange(numItems)]
+        check: w == [12, 13, 14].toTensor()
 
     test "Random tensor":
       # Check that randomTensor doesn't silently convert float32 to float64