fetch_ml/native/tests/test_dataset_hash.cpp

// Simple test suite for dataset_hash library (no external dependencies)

#include <cstdio>
#include <cstdlib>
#include <cstring>
#include <cassert>
#include <string>
#include <vector>
#include <chrono>
#include <filesystem>
#include <fstream>
#include <iostream>

#include "dataset_hash/dataset_hash.h"

namespace fs = std::filesystem;
using namespace std::chrono;

// Simple test macros
#define TEST_ASSERT(cond) \
    do { \
        if (!(cond)) { \
            fprintf(stderr, "ASSERTION FAILED: %s at line %d\n", #cond, __LINE__); \
            return 1; \
        } \
    } while(0)

#define TEST_ASSERT_EQ(a, b) TEST_ASSERT((a) == (b))
#define TEST_ASSERT_NE(a, b) TEST_ASSERT((a) != (b))
#define TEST_ASSERT_STR_EQ(a, b) TEST_ASSERT(strcmp((a), (b)) == 0)

// Helper functions
fs::path create_temp_dir() {
    fs::path temp = fs::temp_directory_path() / "dataset_hash_test_XXXXXX";
    fs::create_directories(temp);
    return temp;
}

void cleanup_temp_dir(const fs::path& dir) {
    fs::remove_all(dir);
}

void create_test_file(const fs::path& dir, const std::string& name, const std::string& content) {
    std::ofstream file(dir / name);
    file << content;
    file.close();
}

// Test 1: Context creation
int test_context_creation() {
    printf("Testing context creation...\n");
    
    // Auto-detect threads
    fh_context_t* ctx = fh_init(0);
    TEST_ASSERT_NE(ctx, nullptr);
    fh_cleanup(ctx);
    
    // Specific thread count
    ctx = fh_init(4);
    TEST_ASSERT_NE(ctx, nullptr);
    fh_cleanup(ctx);
    
    printf("  PASSED\n");
    return 0;
}

// Test 2: SIMD detection
int test_simd_detection() {
    printf("Testing SIMD detection...\n");
    
    int has_simd = fh_has_simd_sha256();
    const char* impl_name = fh_get_simd_impl_name();
    
    printf("  SIMD available: %s\n", has_simd ? "yes" : "no");
    printf("  Implementation: %s\n", impl_name);
    
    TEST_ASSERT_NE(impl_name, nullptr);
    TEST_ASSERT(strlen(impl_name) > 0);
    
    printf("  PASSED\n");
    return 0;
}

// Test 3: Hash single file
int test_hash_single_file() {
    printf("Testing single file hash...\n");
    
    fs::path temp = create_temp_dir();
    
    fh_context_t* ctx = fh_init(1);
    TEST_ASSERT_NE(ctx, nullptr);
    
    // Create test file
    create_test_file(temp, "test.txt", "Hello, World!");
    
    // Hash it
    char* hash = fh_hash_file(ctx, (temp / "test.txt").string().c_str());
    TEST_ASSERT_NE(hash, nullptr);
    
    // Verify hash format (64 hex characters + null)
    TEST_ASSERT_EQ(strlen(hash), 64);
    
    // Hash should be deterministic
    char* hash2 = fh_hash_file(ctx, (temp / "test.txt").string().c_str());
    TEST_ASSERT_NE(hash2, nullptr);
    TEST_ASSERT_STR_EQ(hash, hash2);
    
    fh_free_string(hash);
    fh_free_string(hash2);
    fh_cleanup(ctx);
    cleanup_temp_dir(temp);
    
    printf("  PASSED\n");
    return 0;
}

// Test 4: Hash empty file (known hash)
int test_hash_empty_file() {
    printf("Testing empty file hash...\n");
    
    fs::path temp = create_temp_dir();
    
    fh_context_t* ctx = fh_init(1);
    TEST_ASSERT_NE(ctx, nullptr);
    
    // Create empty file
    create_test_file(temp, "empty.txt", "");
    
    char* hash = fh_hash_file(ctx, (temp / "empty.txt").string().c_str());
    TEST_ASSERT_NE(hash, nullptr);
    TEST_ASSERT_EQ(strlen(hash), 64);
    
    // Debug: print actual hash
    printf("  Empty file hash: %s\n", hash);
    
    // Known SHA-256 of empty string
    const char* expected = "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855";
    printf("  Expected hash: %s\n", expected);
    TEST_ASSERT_STR_EQ(hash, expected);
    
    fh_free_string(hash);
    fh_cleanup(ctx);
    cleanup_temp_dir(temp);
    
    printf("  PASSED\n");
    return 0;
}

// Test 5: Hash directory
int test_hash_directory() {
    printf("Testing directory hash...\n");
    
    fs::path temp = create_temp_dir();
    
    fh_context_t* ctx = fh_init(4);
    TEST_ASSERT_NE(ctx, nullptr);
    
    // Create directory structure
    create_test_file(temp, "root.txt", "root");
    fs::create_directories(temp / "subdir");
    create_test_file(temp, "subdir/file1.txt", "file1");
    create_test_file(temp, "subdir/file2.txt", "file2");
    
    // Hash directory
    char* hash = fh_hash_directory(ctx, temp.string().c_str());
    TEST_ASSERT_NE(hash, nullptr);
    TEST_ASSERT_EQ(strlen(hash), 64);
    
    // Hash should be deterministic
    char* hash2 = fh_hash_directory(ctx, temp.string().c_str());
    TEST_ASSERT_NE(hash2, nullptr);
    TEST_ASSERT_STR_EQ(hash, hash2);
    
    fh_free_string(hash);
    fh_free_string(hash2);
    fh_cleanup(ctx);
    cleanup_temp_dir(temp);
    
    printf("  PASSED\n");
    return 0;
}

// Test 6: Batch hash
int test_batch_hash() {
    printf("Testing batch hash...\n");
    
    fs::path temp = create_temp_dir();
    
    fh_context_t* ctx = fh_init(4);
    TEST_ASSERT_NE(ctx, nullptr);
    
    // Create test files
    const int num_files = 10;
    std::vector<std::string> paths;
    std::vector<const char*> c_paths;
    
    for (int i = 0; i < num_files; i++) {
        std::string name = "file_" + std::to_string(i) + ".txt";
        create_test_file(temp, name, "Content " + std::to_string(i));
        paths.push_back((temp / name).string());
        c_paths.push_back(paths.back().c_str());
    }
    
    // Hash batch
    std::vector<char*> results(num_files, nullptr);
    int ret = fh_hash_batch(ctx, c_paths.data(), num_files, results.data());
    TEST_ASSERT_EQ(ret, 0);
    
    // Verify all hashes
    for (int i = 0; i < num_files; i++) {
        TEST_ASSERT_NE(results[i], nullptr);
        TEST_ASSERT_EQ(strlen(results[i]), 64);
        fh_free_string(results[i]);
    }
    
    fh_cleanup(ctx);
    cleanup_temp_dir(temp);
    
    printf("  PASSED\n");
    return 0;
}

// Test 7: Performance test
int test_performance() {
    printf("Testing performance...\n");
    
    fs::path temp = create_temp_dir();
    
    fh_context_t* ctx = fh_init(4);
    TEST_ASSERT_NE(ctx, nullptr);
    
    // Create 1000 small files
    const int num_files = 1000;
    auto start = high_resolution_clock::now();
    
    for (int i = 0; i < num_files; i++) {
        create_test_file(temp, "perf_" + std::to_string(i) + ".txt", "content");
    }
    
    auto create_end = high_resolution_clock::now();
    
    // Hash all files
    char* hash = fh_hash_directory(ctx, temp.string().c_str());
    TEST_ASSERT_NE(hash, nullptr);
    
    auto hash_end = high_resolution_clock::now();
    
    auto create_time = duration_cast<milliseconds>(create_end - start);
    auto hash_time = duration_cast<milliseconds>(hash_end - create_end);
    
    printf("  Created %d files in %ld ms\n", num_files, create_time.count());
    printf("  Hashed %d files in %ld ms\n", num_files, hash_time.count());
    printf("  Throughput: %.1f files/sec\n", num_files * 1000.0 / hash_time.count());
    
    fh_free_string(hash);
    fh_cleanup(ctx);
    cleanup_temp_dir(temp);
    
    printf("  PASSED\n");
    return 0;
}

// Main test runner
int main() {
    printf("\n=== Dataset Hash Library Test Suite ===\n\n");
    
    int failed = 0;
    
    failed += test_context_creation();
    failed += test_simd_detection();
    failed += test_hash_single_file();
    failed += test_hash_empty_file();
    failed += test_hash_directory();
    failed += test_batch_hash();
    failed += test_performance();
    
    printf("\n=== Test Results ===\n");
    if (failed == 0) {
        printf("All tests PASSED!\n");
        return 0;
    } else {
        printf("%d test(s) FAILED\n", failed);
        return 1;
    }
}