Best prompts for ChatGPT for coding using debugging complex algorithms

You are a debugging assistant. Analyze the following C++ binary search implementation that intermittently returns wrong indices. Explain the root cause (off-by-one or loop invariant), provide a corrected implementation, a short explanation of why it now works, and at least five unit tests including edge cases. Code: int binary_search(const vector<int>& a, int target) { int l = 0, r = a.size() - 1; while (l <= r) { int m = (l + r) / 2; if (a[m] == target) return m; if (a[m] < target) l = m + 1; else r = m - 1; } return -1; } Observed failure: searching for the first or last element on some inputs returns incorrect index. Explain why and provide a corrected, robust version (handle empty array, integer overflow in mid calculation) and tests.

You are an expert in dynamic programming debugging. Given the following Python implementation for computing the longest increasing subsequence (LIS) length, find why it returns incorrect answers on some inputs. Explain the bug (wrong memoization key or recursion state), provide a corrected version (iterative or recursive + memo) with complexity analysis, and add unit tests that cover tricky patterns (equal elements, strictly increasing requirement, long sequences). Code: def lis_length(arr): memo = {} def helper(i, prev): if i == len(arr): return 0 key = (i, prev) if key in memo: return memo[key] taken = 0 if prev < arr[i]: taken = 1 + helper(i+1, arr[i]) not_taken = helper(i+1, prev) memo[key] = max(taken, not_taken) return memo[key] return helper(0, float('-inf')) Observed: returns wrong length on arrays with negative numbers and equal elements. Provide a corrected version and 6 unit tests.

You are a C++ memory debugging assistant. Inspect the code below for memory leaks and lifetime issues. Identify where allocations are not freed, show how to reproduce the leak with a small test, provide a fixed version (prefer modern C++ RAII smart pointers or std::vector), and show sample valgrind or sanitizer output before and after the fix. Code: void process(int n) { for (int i = 0; i < n; ++i) { int* buf = new int[1024]; // do some work with buf if (i % 100 == 0) break; // early exit } } Observed: memory usage grows when process is called repeatedly. Explain and fix.

You are a numerical debugging assistant. Given the following Java code that compares doubles and produces incorrect branch decisions, identify precision pitfalls, propose and implement a stable solution (epsilon strategy, relative tolerance, or use BigDecimal where appropriate), and produce unit tests demonstrating correctness. Code: public boolean isZero(double x) { return x == 0.0; } // used in: if (isZero(a/b - 1.0)) { // assume a==b } Observed: occasionally branch not taken due to tiny floating error. Provide corrected implementation, explanation of chosen tolerance, and tests showing edge cases.

You are a concurrent-systems debugger. Analyze the multithreaded Java code below where a shared counter produces incorrect totals under concurrency. Explain the race, show how to reproduce it with a stress test, provide a thread-safe fix (e.g., AtomicInteger, synchronized, or Lock), compare performance implications, and add a concurrent unit test that reliably demonstrates correctness. Code: public class Counter { private int count = 0; public void inc() { count++; } public int get() { return count; } } Multiple threads call inc() heavily and final count is sometimes less than expected. Fix and test.

You are a C systems debugging assistant. The following C function sometimes causes a segmentation fault. Explain why (pointer arithmetic or off-by-one), produce a minimized reproducer, fix the code and recommend sanitizer/compilation flags. Provide corrected code and a small test that confirms no segfault. Code: char *concat(const char *a, const char *b) { size_t la = strlen(a); size_t lb = strlen(b); char *s = malloc(la + lb); strcpy(s, a); strcat(s, b); return s; } Observed: segmentation fault on certain inputs. Fix and explain.