fix: split-numbers test solutions were wrong #76

members/libpt-bintols/src/split_numbers.rs

@@ -3,9 +3,24 @@
 //! Sometimes, you need a large integer in the form of many bytes, so split into [u8].
 //! Rust provides
 
-/// split an integer into it's bytes, ignoring those bytes that would be all zero.
+
+/// Split unsigned integers into a [Vec] of [u8]s
 ///
-/// If the integer is zero, the Vec contains a single null byte.
+/// Say you have the [u32] 1717 (binary: `00000000 00000000 00000110 10110101 `). This number would
+/// be splitted to `vec![0b00000110, 0b10110101]`.
+///
+/// The 0 bytes of the numbers will be discarded (unless the number is 0, then the Vec contains a
+/// single Null byte.) and the remaining parts of the numbers are inserted into a Vec as [u8].
+///
+/// # Examples
+///
+/// ```
+/// # use libpt_bintols::split_numbers::*;
+///
+/// let x: u32 = 1717;
+///
+/// assert_eq!(unsigned_to_vec(x), vec![0b00000110, 0b10110101]);
+/// ```
 pub fn unsigned_to_vec<T>(num: T) -> Vec<u8>
 where
     u128: std::convert::From<T>
@@ -19,5 +34,6 @@ where
         buf.push(num as u8);
         num >>= 8;
     }
+    buf.reverse();
     buf
 }

members/libpt-bintols/tests/split_numbers.rs

@@ -2,16 +2,25 @@ use libpt_bintols::split_numbers::*;
 
 #[test]
 fn split_u128() {
-    let source = [16, 255, 256, 0, u128::MAX, u64::MAX as u128];
+    let source = [
+        16,
+        255,
+        256,
+        0,
+        u128::MAX,
+        u64::MAX as u128,
+        u64::MAX as u128 + 1,
+    ];
     let correct = [
         vec![16],
         vec![255],
-        vec![255, 1],
+        vec![1, 0],
         vec![0],
         vec![
             255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255,
         ],
         vec![255, 255, 255, 255, 255, 255, 255, 255],
+        vec![1, 0, 0, 0, 0, 0, 0, 0, 0],
     ];
     for (i, n) in source.iter().enumerate() {
         assert_eq!(unsigned_to_vec(*n), correct[i]);
@@ -20,14 +29,25 @@ fn split_u128() {
 
 #[test]
 fn split_u64() {
-    let source = [16, 255, 256, 0, u64::MAX, u32::MAX as u64];
+    let source = [
+        16,
+        255,
+        256,
+        0,
+        u64::MAX,
+        u32::MAX as u64,
+        0b1_00000001,
+        0b10011011_10110101_11110000_00110011,
+    ];
     let correct = [
         vec![16],
         vec![255],
-        vec![255, 1],
+        vec![1, 0],
         vec![0],
         vec![255, 255, 255, 255, 255, 255, 255, 255],
         vec![255, 255, 255, 255],
+        vec![1, 1],
+        vec![0b10011011, 0b10110101, 0b11110000, 0b00110011],
     ];
     for (i, n) in source.iter().enumerate() {
         assert_eq!(unsigned_to_vec(*n), correct[i]);