Reworking the approach to enabling/disabling exceptions

Now there is no dependency on Abseil (using custom Status/StatusOr
classes) and functions that are part of the interface and return
Status/StatusOr are named with a NoExceptions suffix.

Author: michaelbautin

CMakeLists.txt

@@ -7,34 +7,38 @@ include(GNUInstallDirs)
 include(CheckCXXCompilerFlag)
 include(FetchContent)
 
+# These example/test targets catch exceptions, so exceptions should always be
+# enabled building these files even if they are disabled in other targets.
+set(HNSWLIB_TARGETS_REQUIRING_EXCEPTIONS
+    example_mt_filter
+    example_mt_replace_deleted
+    example_mt_search
+    multiThread_replace_test
+    test_updates)
+
+
 function(add_example_or_test TARGET_NAME ...)
     add_executable(${ARGV})
     target_link_libraries("${TARGET_NAME}" ${common_link_libraries})
-    target_include_directories("${TARGET_NAME}" PUBLIC "${abseil-cpp_SOURCE_DIR}")
-    if(HNSWLIB_USE_ABSEIL)
-        target_compile_definitions("${TARGET_NAME}" PUBLIC HNSWLIB_USE_ABSEIL)
+    list(FIND HNSWLIB_TARGETS_REQUIRING_EXCEPTIONS "${TARGET_NAME}" _index)
+    if(_index GREATER -1)
+        if(NOT HNSWLIB_ENABLE_EXCEPTIONS)
+            message("Enabling exceptions for target ${TARGET_NAME} as a special case")
+        endif()
+        set(should_enable_exceptions ON)
+    else()
+        set(should_enable_exceptions "${HNSWLIB_ENABLE_EXCEPTIONS}")
+    endif()
+    if(should_enable_exceptions)
+        target_compile_options("${TARGET_NAME}" PUBLIC -fexceptions)
+    else()
+        target_compile_options("${TARGET_NAME}" PUBLIC -fno-exceptions)
     endif()
 endfunction()
 
-option(HNSWLIB_USE_ABSEIL "Whether to use Abseil in hnswlib" OFF)
+option(HNSWLIB_ENABLE_EXCEPTIONS "Whether to enable exceptions in hnswlib" ON)
 
-if(HNSWLIB_USE_ABSEIL)
-    FetchContent_Declare(
-        abseil-cpp
-        GIT_REPOSITORY https://github.com/abseil/abseil-cpp
-        GIT_TAG 20250127.1
-    )
-
-    FetchContent_MakeAvailable(abseil-cpp)
-
-    if("${abseil-cpp_SOURCE_DIR}" STREQUAL "")
-        message(FATAL "Failed to find Abseil source directory")
-    endif()
-    # Abseil requires C++14 or later.
-    set(CMAKE_CXX_STANDARD 14)
-else()
-    set(CMAKE_CXX_STANDARD 11)
-endif()
+set(CMAKE_CXX_STANDARD 11)
 
 add_library(hnswlib INTERFACE)
 add_library(hnswlib::hnswlib ALIAS hnswlib)
@@ -83,18 +87,16 @@ if(HNSWLIB_EXAMPLES)
 
     set(common_link_libraries
         hnswlib)
-    if(HNSWLIB_USE_ABSEIL)
-      list(APPEND common_link_libraries absl::status absl::statusor)
-    endif()
 
     set(EXAMPLE_NAMES
-        example_search
         example_epsilon_search
-        example_multivector_search
         example_filter
-        example_replace_deleted
+        example_mt_filter
+        example_mt_replace_deleted
         example_mt_search
-        example_mt_replace_deleted)
+        example_multivector_search
+        example_replace_deleted
+        example_search)
 
     foreach(example_name IN LISTS EXAMPLE_NAMES)
       add_example_or_test("${example_name}" "examples/cpp/${example_name}.cpp")

examples/cpp/example_epsilon_search.cpp

@@ -53,7 +53,7 @@ int main() {
         }
         std::cout << "Query #" << i << "\n";
         hnswlib::EpsilonSearchStopCondition<dist_t> stop_condition(epsilon2, min_num_candidates, max_elements);
-        HNSWLIB_ASSIGN_VALUE_OR_THROW_IN_TEST(result, alg_hnsw->searchStopConditionClosest(query_data, stop_condition));
+        auto result = alg_hnsw->searchStopConditionClosest(query_data, stop_condition);
         size_t num_vectors = result.size();
         std::cout << "Found " << num_vectors << " vectors\n";
         delete[] query_data;

examples/cpp/example_filter.cpp

@@ -45,8 +45,7 @@ int main() {
     // Query the elements for themselves with filter and check returned labels
     int k = 10;
     for (int i = 0; i < max_elements; i++) {
-        HNSWLIB_ASSIGN_VALUE_OR_THROW_IN_TEST(
-            result, alg_hnsw->searchKnnCloserFirst(data + i * dim, k, &pickIdsDivisibleByTwo));
+        auto result = alg_hnsw->searchKnnCloserFirst(data + i * dim, k, &pickIdsDivisibleByTwo);
         for (auto item: result) {
             if (item.second % 2 == 1) std::cout << "Error: found odd label\n";
         }

examples/cpp/example_mt_filter.cpp

@@ -106,9 +106,7 @@ int main() {
     int k = 10;
     std::vector<hnswlib::labeltype> neighbors(max_elements * k);
     ParallelFor(0, max_elements, num_threads, [&](size_t row, size_t threadId) {
-        HNSWLIB_ASSIGN_VALUE_OR_THROW_IN_TEST(
-            result,
-            alg_hnsw->searchKnn(data + dim * row, k, &pickIdsDivisibleByTwo));
+        auto result = alg_hnsw->searchKnn(data + dim * row, k, &pickIdsDivisibleByTwo);
         for (int i = 0; i < k; i++) {
             hnswlib::labeltype label = result.top().second;
             result.pop();

examples/cpp/example_mt_search.cpp

@@ -89,7 +89,7 @@ int main() {
     // Query the elements for themselves and measure recall
     std::vector<hnswlib::labeltype> neighbors(max_elements);
     ParallelFor(0, max_elements, num_threads, [&](size_t row, size_t threadId) {
-        HNSWLIB_ASSIGN_VALUE_OR_THROW_IN_TEST(result, alg_hnsw->searchKnn(data + dim * row, 1));
+        auto result = alg_hnsw->searchKnn(data + dim * row, 1);
         hnswlib::labeltype label = result.top().second;
         neighbors[row] = label;
     });

examples/cpp/example_multivector_search.cpp

@@ -63,9 +63,7 @@ int main() {
         }
         std::cout << "Query #" << i << "\n";
         hnswlib::MultiVectorSearchStopCondition<docidtype, dist_t> stop_condition(space, num_docs, ef_collection);
-        HNSWLIB_ASSIGN_VALUE_OR_THROW_IN_TEST(
-            result,
-            alg_hnsw->searchStopConditionClosest(query_data, stop_condition));
+        auto result = alg_hnsw->searchStopConditionClosest(query_data, stop_condition);
         size_t num_vectors = result.size();
 
         std::unordered_map<docidtype, size_t> doc_counter;

examples/cpp/example_search.cpp

@@ -29,7 +29,7 @@ int main() {
     // Query the elements for themselves and measure recall
     float correct = 0;
     for (int i = 0; i < max_elements; i++) {
-        HNSWLIB_ASSIGN_VALUE_OR_THROW_IN_TEST(result, alg_hnsw->searchKnn(data + i * dim, 1));
+        auto result = alg_hnsw->searchKnn(data + i * dim, 1);
         hnswlib::labeltype label = result.top().second;
         if (label == i) correct++;
     }
@@ -45,7 +45,7 @@ int main() {
     alg_hnsw = new hnswlib::HierarchicalNSW<float>(&space, hnsw_path);
     correct = 0;
     for (int i = 0; i < max_elements; i++) {
-        HNSWLIB_ASSIGN_VALUE_OR_THROW_IN_TEST(result, alg_hnsw->searchKnn(data + i * dim, 1));
+        auto result = alg_hnsw->searchKnn(data + i * dim, 1);
         hnswlib::labeltype label = result.top().second;
         if (label == i) correct++;
     }

hnswlib/bruteforce.h

@@ -54,7 +54,7 @@ class BruteforceSearch : public AlgorithmInterface<dist_t> {
         size_per_element_ = data_size_ + sizeof(labeltype);
         data_ = (char *) malloc(maxElements * size_per_element_);
         if (data_ == nullptr)
-            throw std::runtime_error("Not enough memory: BruteforceSearch failed to allocate data");
+            HNSWLIB_THROW_RUNTIME_ERROR("Not enough memory: BruteforceSearch failed to allocate data");
         cur_element_count = 0;
     }
 
@@ -64,7 +64,7 @@ class BruteforceSearch : public AlgorithmInterface<dist_t> {
     }
 
 
-    HNSWLIB_STATUS_TYPE addPoint(const void *datapoint, labeltype label, bool replace_deleted = false) override {
+    Status addPointNoExceptions(const void *datapoint, labeltype label, bool replace_deleted = false) override {
         int idx;
         {
             std::unique_lock<std::mutex> lock(index_lock);
@@ -74,7 +74,7 @@ class BruteforceSearch : public AlgorithmInterface<dist_t> {
                 idx = search->second;
             } else {
                 if (cur_element_count >= maxelements_) {
-                    throw std::runtime_error("The number of elements exceeds the specified limit\n");
+                    HNSWLIB_THROW_RUNTIME_ERROR("The number of elements exceeds the specified limit\n");
                 }
                 idx = cur_element_count;
                 dict_external_to_internal[label] = idx;
@@ -83,7 +83,7 @@ class BruteforceSearch : public AlgorithmInterface<dist_t> {
         }
         memcpy(data_ + size_per_element_ * idx + data_size_, &label, sizeof(labeltype));
         memcpy(data_ + size_per_element_ * idx, datapoint, data_size_);
-        HNSWLIB_RETURN_OK_STATUS;
+        return OkStatus();
     }
 
 
@@ -108,8 +108,8 @@ class BruteforceSearch : public AlgorithmInterface<dist_t> {
 
 
     using DistanceLabelPriorityQueue = typename AlgorithmInterface<dist_t>::DistanceLabelPriorityQueue;
-    HNSWLIB_STATUS_OR_TYPE(DistanceLabelPriorityQueue)
-    searchKnn(const void *query_data, size_t k, BaseFilterFunctor* isIdAllowed = nullptr) const override {
+    StatusOr<DistanceLabelPriorityQueue>
+    searchKnnNoExceptions(const void *query_data, size_t k, BaseFilterFunctor* isIdAllowed = nullptr) const override {
         assert(k <= cur_element_count);
         std::priority_queue<std::pair<dist_t, labeltype >> topResults;
         if (cur_element_count == 0) return topResults;
@@ -168,7 +168,7 @@ class BruteforceSearch : public AlgorithmInterface<dist_t> {
         size_per_element_ = data_size_ + sizeof(labeltype);
         data_ = (char *) malloc(maxelements_ * size_per_element_);
         if (data_ == nullptr)
-            throw std::runtime_error("Not enough memory: loadIndex failed to allocate data");
+            HNSWLIB_THROW_RUNTIME_ERROR("Not enough memory: loadIndex failed to allocate data");
 
         input.read(data_, maxelements_ * size_per_element_);