cmake option: CUDF_KVIKIO_REMOTE_IO

cpp/CMakeLists.txt

@@ -90,6 +90,12 @@ option(
 mark_as_advanced(CUDF_BUILD_STREAMS_TEST_UTIL)
 option(CUDF_STATIC_LINTERS "Enable static linters during compilation" OFF)
 
+option(
+  CUDF_KVIKIO_REMOTE_IO
+  "Enable remote IO (e.g. AWS S3) support through KvikIO. If disabled, cudf-python will still be able to do remote IO through fsspec."
+  ON
+)
+
 message(VERBOSE "CUDF: Build with NVTX support: ${USE_NVTX}")
 message(VERBOSE "CUDF: Configure CMake to build tests: ${BUILD_TESTS}")
 message(VERBOSE "CUDF: Configure CMake to build (google & nvbench) benchmarks: ${BUILD_BENCHMARKS}")
@@ -109,6 +115,9 @@ message(
   "CUDF: Enable the -lineinfo option for nvcc (useful for cuda-memcheck / profiler): ${CUDA_ENABLE_LINEINFO}"
 )
 message(VERBOSE "CUDF: Statically link the CUDA runtime: ${CUDA_STATIC_RUNTIME}")
+message(VERBOSE
+        "CUDF: Build with remote IO (e.g. AWS S3) support through KvikIO: ${CUDF_KVIKIO_REMOTE_IO}"
+)
 
 # Set a default build type if none was specified
 rapids_cmake_build_type("Release")
@@ -890,6 +899,9 @@ target_compile_definitions(cudf PRIVATE "RMM_LOGGING_LEVEL=LIBCUDF_LOGGING_LEVEL
 # Define spdlog level
 target_compile_definitions(cudf PUBLIC "SPDLOG_ACTIVE_LEVEL=SPDLOG_LEVEL_${LIBCUDF_LOGGING_LEVEL}")
 
+# Enable remote IO through KvikIO
+target_compile_definitions(cudf PRIVATE $<$<BOOL:${CUDF_KVIKIO_REMOTE_IO}>:CUDF_KVIKIO_REMOTE_IO>)
+
 # Compile stringified JIT sources first
 add_dependencies(cudf jitify_preprocess_run)
 

cpp/cmake/thirdparty/get_kvikio.cmake

@@ -22,7 +22,7 @@ function(find_and_configure_kvikio VERSION)
     GIT_REPOSITORY https://github.com/rapidsai/kvikio.git
     GIT_TAG branch-${VERSION}
     GIT_SHALLOW TRUE SOURCE_SUBDIR cpp
-    OPTIONS "KvikIO_BUILD_EXAMPLES OFF"
+    OPTIONS "KvikIO_BUILD_EXAMPLES OFF" "KvikIO_REMOTE_SUPPORT ${CUDF_KVIKIO_REMOTE_IO}"
   )
 
   include("${rapids-cmake-dir}/export/find_package_root.cmake")

cpp/src/io/utilities/datasource.cpp

@@ -26,7 +26,6 @@
 #include <cudf/utilities/span.hpp>
 
 #include <kvikio/file_handle.hpp>
-#include <kvikio/remote_handle.hpp>
 
 #include <rmm/device_buffer.hpp>
 
@@ -37,6 +36,10 @@
 #include <regex>
 #include <vector>
 
+#ifdef CUDF_KVIKIO_REMOTE_IO
+#include <kvikio/remote_handle.hpp>
+#endif
+
 namespace cudf {
 namespace io {
 namespace {
@@ -391,6 +394,7 @@ class user_datasource_wrapper : public datasource {
   datasource* const source;  ///< A non-owning pointer to the user-implemented datasource
 };
 
+#ifdef CUDF_KVIKIO_REMOTE_IO
 /**
  * @brief Remote file source backed by KvikIO, which handles S3 filepaths seamlessly.
  */
@@ -463,14 +467,23 @@ class remote_file_source : public datasource {
   static bool is_supported_remote_url(std::string const& url)
   {
     // Regular expression to match "s3://"
-    std::regex pattern{R"(^s3://)", std::regex_constants::icase};
+    static std::regex pattern{R"(^s3://)", std::regex_constants::icase};
     return std::regex_search(url, pattern);
   }
 
  private:
   kvikio::RemoteHandle _kvikio_file;
 };
-
+#else
+/**
+ * @brief When KvikIO remote IO is disabled, `is_supported_remote_url()` return false always.
+ */
+class remote_file_source : public file_source {
+ public:
+  explicit remote_file_source(char const* filepath) : file_source(filepath) {}
+  static constexpr bool is_supported_remote_url(std::string const&) { return false; }
+};
+#endif
 }  // namespace
 
 std::unique_ptr<datasource> datasource::create(std::string const& filepath,