find absolute value of int without branching / with branching but using BIT operations

Author: ZapDos7

README.md

@@ -57,7 +57,7 @@
 - [x] [Bit Hacks — Part 2 (Playing with k’th bit)](http://www.techiedelight.com/bit-hacks-part-2-playing-kth-bit/)
 - [x] [Bit Hacks — Part 3 (Playing with rightmost set bit of a number)](http://www.techiedelight.com/bit-hacks-part-3-playing-rightmost-set-bit-number/)
 - [x] [Bit Hacks — Part 4 (Playing with letters of English alphabet)](http://www.techiedelight.com/bit-hacks-part-4-playing-letters-english-alphabet/)
-- [ ] [Bit Hacks — Part 5 (Find absolute value of an integer without branching)](http://www.techiedelight.com/bit-hacks-part-5-find-absolute-value-integer-without-branching/)
+- [x] [Bit Hacks — Part 5 (Find absolute value of an integer without branching)](http://www.techiedelight.com/bit-hacks-part-5-find-absolute-value-integer-without-branching/)
 - [ ] [Bit Hacks — Part 6 (Random Problems)](http://www.techiedelight.com/bit-hacks-part-6-random-problems/)
 
 ## Algorithms

src/binary/absoluteCalculator/AbsolutNumCalculator.java

@@ -0,0 +1,39 @@
+package binary.absoluteCalculator;
+
+import utils.InputReader;
+
+// Given an integer, compute its absolute value (abs) without branching.
+public class AbsolutNumCalculator {
+
+    public static void main(String[] args) {
+        System.out.println("Provide input number: ");
+        int input = InputReader.readInt();
+        System.out.println("No  branching: " + absoluteNoBranch(input));
+        System.out.println("Yes branching: " + absoluteYesBranch(input));
+    }
+
+    private static int absoluteNoBranch(int x) {
+        final int mask = x >> 31; // correct for 32-bit int
+        return ((x + mask) ^ mask);
+        // Explanation:
+        // mask = x >> 31 gives:
+        //      0 if x ≥ 0
+        //      -1 (all 1s) if x < 0 (in two's complement)
+        // Then:
+        //      x + mask subtracts 1 from x if mask == -1
+        //      XOR with mask flips bits if mask == -1
+        // Equivalent to x >= 0 ? x : -x
+    }
+
+    private static int absoluteYesBranch(int n) {
+        return ((n >> 31) & 1) == 1 ? -n : n; // ((n >> 31) & 1) ? (~n + 1) : n;
+        // Explanation:
+        // 1. shift 31 bits: keep sign bit (leftmost)
+        // that's 0 for even nums and 1 for odd ones
+        // 2. masking with 1
+        // (0 for positives, 1s from two's complement for negatives)
+        // 3. AND 1 to extract the least significant bit
+        // If n >= 0, (n >> 31) is 0, so (n >> 31) & 1 = 0.
+        // If n < 0, (n >> 31) is -1 (binary 111...111), so (n >> 31) & 1 = 1.
+    }
+}