refactor: world generation (#17)

* refactor: use TBB for concurrent hash maps and parallelize chunk processing

* fix: tbb link fail

* refactor(chunk): remove biome check for caves in rivers and oceans

* refactor(random): replace std distributions with custom implementations

Avoid overhead and platform-dependent behavior of `<random>` distributions by using direct engine operations and integer arithmetic. This ensures deterministic, cross-platform results and improves performance.

* refactor(generation): use chunk seed for cave and river paths

- Use per-chunk seed instead of global path_id for cave and river generation.
- Remove unused m_sum variables and m_path_id members.
- Clamp river yaw within 10 degrees of initial direction.
- Fix river radius interpolation (use t instead of 1-t).
- Lower sea level from 64 to 63.

src/tools/cubed_random.cpp

@@ -5,8 +5,15 @@ namespace Cubed {
 Random::Random() {}
 Random::Random(unsigned seed) { init(seed); }
 bool Random::random_bool(double probability) {
-    std::bernoulli_distribution dist(probability);
-    return dist(m_engine);
+    if (probability <= 0.0)
+        return false;
+    if (probability >= 1.0)
+        return true;
+
+    const double MAX_VAL = 4294967295.0;
+    unsigned threshold = static_cast<unsigned>(probability * MAX_VAL);
+
+    return m_engine() <= threshold;
 }
 
 std::mt19937& Random::engine() { return m_engine; }
@@ -18,12 +25,23 @@ void Random::init(unsigned seed) {
     m_engine.seed(seed);
 }
 int Random::random_int(int min, int max) {
-    std::uniform_int_distribution<int> dist(min, max);
-    return dist(m_engine);
+    unsigned range = static_cast<unsigned>(max - min) + 1;
+
+    const unsigned LIMIT =
+        (std::numeric_limits<unsigned>::max() / range) * range;
+    unsigned r;
+    do {
+        r = m_engine();
+    } while (r >= LIMIT);
+
+    return min + static_cast<int>(r % range);
 }
 float Random::random_float(float min, float max) {
-    std::uniform_real_distribution<float> dist(min, max);
-    return dist(m_engine);
-}
 
+    unsigned r = m_engine() >> 8;
+    float t = static_cast<float>(r) * (1.0f / 16777216.0f);
+    float result = min + t * (max - min);
+
+    return result;
+}
 } // namespace Cubed