Improve simple binary search according to the standard binary search definition

Author: kanilsz

searches/simple_binary_search.py

@@ -11,50 +11,61 @@
 from __future__ import annotations
 
 
-def binary_search(a_list: list[int], item: int) -> bool:
+def binary_search(a_list: list[int], item: int) -> int:
     """
+    Returns the leftmost index of `item` in `a_list` if found,
+    otherwise returns -1.
+
     >>> test_list = [0, 1, 2, 8, 13, 17, 19, 32, 42]
     >>> binary_search(test_list, 3)
-    False
+    -1
     >>> binary_search(test_list, 13)
-    True
+    4
     >>> binary_search([4, 4, 5, 6, 7], 4)
-    True
+    0
     >>> binary_search([4, 4, 5, 6, 7], -10)
-    False
+    -1
     >>> binary_search([-18, 2], -18)
-    True
+    0
     >>> binary_search([5], 5)
-    True
+    0
     >>> binary_search(['a', 'c', 'd'], 'c')
-    True
+    1
     >>> binary_search(['a', 'c', 'd'], 'f')
-    False
+    -1
     >>> binary_search([], 1)
-    False
+    -1
     >>> binary_search([-.1, .1 , .8], .1)
-    True
+    1
     >>> binary_search(range(-5000, 5000, 10), 80)
-    True
+    508
     >>> binary_search(range(-5000, 5000, 10), 1255)
-    False
+    -1
     >>> binary_search(range(0, 10000, 5), 2)
-    False
+    -1
     """
-    if len(a_list) == 0:
-        return False
-    midpoint = len(a_list) // 2
-    if a_list[midpoint] == item:
-        return True
-    if item < a_list[midpoint]:
-        return binary_search(a_list[:midpoint], item)
-    else:
-        return binary_search(a_list[midpoint + 1 :], item)
+    low, high = 0, len(a_list) - 1
+    result = -1
+
+    while low <= high:
+        mid = (low + high) // 2
+        if a_list[mid] == item:
+            result = mid
+            high = mid - 1  # keep searching left for duplicates
+        elif item < a_list[mid]:
+            high = mid - 1
+        else:
+            low = mid + 1
+
+    return result
 
 
 if __name__ == "__main__":
     user_input = input("Enter numbers separated by comma:\n").strip()
     sequence = [int(item.strip()) for item in user_input.split(",")]
     target = int(input("Enter the number to be found in the list:\n").strip())
-    not_str = "" if binary_search(sequence, target) else "not "
-    print(f"{target} was {not_str}found in {sequence}")
+    index = binary_search(sequence, target)
+    if index != -1:
+        print(f"{target} found at index {index} in {sequence}")
+    else:
+        print(f"{target} was not found in {sequence}")

test.patch

@@ -0,0 +1,18 @@
+#!/usr/bin/env bash
+set -uo pipefail
+cd /app
+
+MODE="${1:-new}"
+
+case "$MODE" in
+  base)
+    pytest tests  # run existing repo tests
+    ;;
+  new)
+    pytest tests/test_binary_search_duplicates.py
+    ;;
+  *)
+    echo "unknown mode: $MODE (expected base or new)" >&2
+    exit 2
+    ;;
+esac

test.sh

@@ -0,0 +1,8 @@
+def test_binary_search_leftmost_duplicate():
+    assert binary_search([1, 2, 2, 2, 3], 2) == 1
+
+def test_binary_search_all_duplicates():
+    assert binary_search([1, 1, 1, 1], 1) == 0
+
+def test_binary_search_not_found():
+    assert binary_search([1, 2, 3], 4) == -1