Convert fill values correctly from negative to unsigned integers (#165)

* Convert fill values correctly from negative to unsigned integers

Positive integers are unaffected by this change anyway.

* Only reinterpret negative integers when decoding fill values to unsigned

---------

Co-authored-by: Fabian Gans <[email protected]>

src/metadata.jl

@@ -222,6 +222,13 @@ Base.eltype(::Metadata{T}) where T = T
 fill_value_decoding(v::AbstractString, T::Type{<:Number}) = parse(T, v)
 fill_value_decoding(v::Nothing, ::Any) = v
 fill_value_decoding(v, T) = T(v)
-fill_value_decoding(v::Integer, T::Type{<: Unsigned}) = reinterpret(T, signed(T)(v))
 fill_value_decoding(v::Number, T::Type{String}) = v == 0 ? "" : T(UInt8[v])
 fill_value_decoding(v, ::Type{ASCIIChar}) = v == "" ? nothing : v
+# Sometimes when translating between CF (climate and forecast) convention data
+# and Zarr groups, fill values are left as "negative integers" to encode unsigned
+# integers.  So, we have to convert to the signed type with the same number of bytes
+# as the unsigned integer, then reinterpret as unsigned.  That's how a fill value 
+# of -1 can have a realistic meaning with an unsigned dtype.
+# However, we have to apply this correction only if the integer is negative.  
+# If it's positive, then the value might be out of range of the signed integer type.
+fill_value_decoding(v::Integer, T::Type{<: Unsigned}) = sign(v) < 0 ? reinterpret(T, signed(T)(v)) : T(v)