Fix: Copy-less insertions in dense indexes

Closes #628

Co-authored-by: Robin Winter <45263519+jrwnter@users.noreply.github.com>

Author: ashvardanian

include/usearch/index_dense.hpp

@@ -757,12 +757,12 @@ class index_dense_gt {
     };
 
     // clang-format off
-    add_result_t add(vector_key_t key, b1x8_t const* vector, std::size_t thread = any_thread(), bool force_vector_copy = true) { return add_(key, vector, thread, force_vector_copy, casts_.from.b1x8); }
-    add_result_t add(vector_key_t key, i8_t const* vector, std::size_t thread = any_thread(), bool force_vector_copy = true) { return add_(key, vector, thread, force_vector_copy, casts_.from.i8); }
-    add_result_t add(vector_key_t key, f16_t const* vector, std::size_t thread = any_thread(), bool force_vector_copy = true) { return add_(key, vector, thread, force_vector_copy, casts_.from.f16); }
-    add_result_t add(vector_key_t key, bf16_t const* vector, std::size_t thread = any_thread(), bool force_vector_copy = true) { return add_(key, vector, thread, force_vector_copy, casts_.from.bf16); }
-    add_result_t add(vector_key_t key, f32_t const* vector, std::size_t thread = any_thread(), bool force_vector_copy = true) { return add_(key, vector, thread, force_vector_copy, casts_.from.f32); }
-    add_result_t add(vector_key_t key, f64_t const* vector, std::size_t thread = any_thread(), bool force_vector_copy = true) { return add_(key, vector, thread, force_vector_copy, casts_.from.f64); }
+    add_result_t add(vector_key_t key, b1x8_t const* vector, std::size_t thread = any_thread(), bool copy_vector = true) { return add_(key, vector, thread, copy_vector, casts_.from.b1x8); }
+    add_result_t add(vector_key_t key, i8_t const* vector, std::size_t thread = any_thread(), bool copy_vector = true) { return add_(key, vector, thread, copy_vector, casts_.from.i8); }
+    add_result_t add(vector_key_t key, f16_t const* vector, std::size_t thread = any_thread(), bool copy_vector = true) { return add_(key, vector, thread, copy_vector, casts_.from.f16); }
+    add_result_t add(vector_key_t key, bf16_t const* vector, std::size_t thread = any_thread(), bool copy_vector = true) { return add_(key, vector, thread, copy_vector, casts_.from.bf16); }
+    add_result_t add(vector_key_t key, f32_t const* vector, std::size_t thread = any_thread(), bool copy_vector = true) { return add_(key, vector, thread, copy_vector, casts_.from.f32); }
+    add_result_t add(vector_key_t key, f64_t const* vector, std::size_t thread = any_thread(), bool copy_vector = true) { return add_(key, vector, thread, copy_vector, casts_.from.f64); }
 
     search_result_t search(b1x8_t const* vector, std::size_t wanted, std::size_t thread = any_thread(), bool exact = false) const { return search_(vector, wanted, dummy_predicate_t {}, thread, exact, casts_.from.b1x8); }
     search_result_t search(i8_t const* vector, std::size_t wanted, std::size_t thread = any_thread(), bool exact = false) const { return search_(vector, wanted, dummy_predicate_t {}, thread, exact, casts_.from.i8); }
@@ -2002,14 +2002,13 @@ class index_dense_gt {
     template <typename scalar_at>
     add_result_t add_(                             //
         vector_key_t key, scalar_at const* vector, //
-        std::size_t thread, bool force_vector_copy, cast_punned_t const& cast) {
+        std::size_t thread, bool copy_vector, cast_punned_t const& cast) {
 
         if (!multi() && config().enable_key_lookups && contains(key))
             return add_result_t{}.failed("Duplicate keys not allowed in high-level wrappers");
 
         // Cast the vector, if needed for compatibility with `metric_`
         thread_lock_t lock = thread_lock_(thread);
-        bool copy_vector = !config_.exclude_vectors || force_vector_copy;
         byte_t const* vector_data = reinterpret_cast<byte_t const*>(vector);
         {
             byte_t* casted_data = cast_buffer_.data() + metric_.bytes_per_vector() * lock.thread_id;