commit 08553befcfc976dcc78edc1d7aca909ea0262775
Author: Yiyuan Li <yy@eyuan.me>
Date:   Fri Jan 9 05:48:14 2026 +0000

    Add test.sh

diff --git a/test.sh b/test.sh
new file mode 100644
index 0000000..00c4df3
--- /dev/null
+++ b/test.sh
@@ -0,0 +1,173 @@
+#!/bin/bash
+
+# Test push_swap with multiple sizes, memory leak check, and sorting verification
+
+COUNT=${1:-5}  # Number of tests per size, default 5
+
+# Detect OS and architecture
+if [[ "$OSTYPE" == "linux-gnu"* ]]; then
+    OS="linux"
+    CHECKER="./checker/checker_linux"
+elif [[ "$OSTYPE" == "darwin"* ]]; then
+    OS="mac"
+    CHECKER="./checker/checker_Mac"
+else
+    echo "Unsupported OS: $OSTYPE"
+    exit 1
+fi
+
+ARCH=$(uname -m)
+echo "Detected OS: $OS ($ARCH)"
+echo "Using checker: $CHECKER"
+
+# Test if checker works (x86_64 checker on arm64 may hang)
+CHECKER_WORKS=true
+if [ -x "$CHECKER" ]; then
+    test_result=$(echo "" | timeout 2 $CHECKER 1 2>/dev/null; echo $?)
+    if [ "$test_result" == "124" ]; then
+        echo "⚠ Checker timeout - may have architecture mismatch (x86_64 on arm64)"
+        CHECKER_WORKS=false
+    fi
+fi
+echo ""
+
+# Test sizes and their targets (official: 5->12, 100->700, 500->5500)
+SIZES=(2 5 10 100 200 500)
+TARGETS=(1 12 50 700 2000 5500)
+
+# Generate random numbers based on OS
+generate_random() {
+    local size=$1
+    if [[ "$OS" == "mac" ]]; then
+        jot $size 1 $size | sort -R | tr '\n' ' '
+    else
+        shuf -i 1-$size | tr '\n' ' '
+    fi
+}
+
+# Run tests for each size
+overall_pass=true
+
+for idx in "${!SIZES[@]}"; do
+    SIZE=${SIZES[$idx]}
+    TARGET=${TARGETS[$idx]}
+
+    echo "============================================"
+    echo "Testing with $SIZE numbers (target: <= $TARGET ops)"
+    echo "============================================"
+
+    total=0
+    min=999999
+    max=0
+    all_sorted=true
+
+    for i in $(seq 1 $COUNT); do
+        # Generate unique random numbers
+        ARG=$(generate_random $SIZE)
+
+        # Run push_swap and capture output
+        OPS_OUTPUT=$(./push_swap $ARG 2>/dev/null)
+        ops=$(echo "$OPS_OUTPUT" | wc -l | tr -d ' ')
+
+        # Handle empty output (already sorted)
+        if [ -z "$OPS_OUTPUT" ]; then
+            ops=0
+        fi
+
+        # Verify sorting with checker
+        if [ "$CHECKER_WORKS" = true ] && [ -x "$CHECKER" ]; then
+            # Handle empty operations (already sorted case) - use printf for true empty
+            if [ -z "$OPS_OUTPUT" ]; then
+                result=$(printf '' | timeout 5 $CHECKER $ARG 2>/dev/null)
+            else
+                result=$(printf '%s\n' "$OPS_OUTPUT" | timeout 5 $CHECKER $ARG 2>/dev/null)
+            fi
+            if [ "$result" == "OK" ]; then
+                sorted_status="✓"
+            else
+                sorted_status="✗"
+                all_sorted=false
+            fi
+        else
+            sorted_status="-"  # Checker not available
+        fi
+
+        # Update stats
+        total=$((total + ops))
+        if [ $ops -lt $min ]; then min=$ops; fi
+        if [ $ops -gt $max ]; then max=$ops; fi
+
+        echo "Run $i: $ops ops $sorted_status"
+    done
+
+    avg=$((total / COUNT))
+    echo "--------------------------------------------"
+    echo "Min: $min | Max: $max | Avg: $avg | Target: $TARGET"
+
+    if [ $max -le $TARGET ]; then
+        echo "✓ PASS - Operations within benchmark"
+    else
+        echo "✗ FAIL - Max ($max) exceeds target ($TARGET)"
+        overall_pass=false
+    fi
+
+    if [ "$CHECKER_WORKS" = true ]; then
+        if [ "$all_sorted" = true ]; then
+            echo "✓ PASS - All sorts verified correct"
+        else
+            echo "✗ FAIL - Some sorts incorrect"
+            overall_pass=false
+        fi
+    else
+        echo "- SKIP - Checker not available"
+    fi
+    echo ""
+done
+
+# Memory leak check
+echo "============================================"
+echo "Memory Leak Check"
+echo "============================================"
+
+ARG=$(generate_random 100)
+
+if [[ "$OS" == "mac" ]]; then
+    # macOS: use leaks command
+    echo "Running: leaks --atExit -- ./push_swap ..."
+    leaks_output=$(leaks --atExit -- ./push_swap $ARG 2>&1)
+    if echo "$leaks_output" | grep -q "0 leaks"; then
+        echo "✓ PASS - No memory leaks detected"
+    elif echo "$leaks_output" | grep -q "cannot examine"; then
+        echo "⚠ SKIP - leaks tool cannot examine process (SIP restriction)"
+    else
+        leak_count=$(echo "$leaks_output" | grep -o "[0-9]* leak" | head -1)
+        echo "✗ FAIL - Memory leaks detected: $leak_count"
+        overall_pass=false
+    fi
+else
+    # Linux: use valgrind
+    if command -v valgrind &> /dev/null; then
+        echo "Running: valgrind --leak-check=full ./push_swap ..."
+        valgrind_output=$(valgrind --leak-check=full --error-exitcode=1 ./push_swap $ARG 2>&1 >/dev/null)
+        if [ $? -eq 0 ]; then
+            echo "✓ PASS - No memory leaks detected"
+        else
+            echo "✗ FAIL - Memory leaks detected"
+            overall_pass=false
+        fi
+    else
+        echo "⚠ SKIP - valgrind not installed"
+    fi
+fi
+
+echo ""
+echo "============================================"
+echo "Final Result"
+echo "============================================"
+if [ "$overall_pass" = true ]; then
+    echo "✓ ALL TESTS PASSED"
+    exit 0
+else
+    echo "✗ SOME TESTS FAILED"
+    exit 1
+fi