Adding support for parse URI for hostnames

src/main/cpp/src/ParseURIJni.cpp

@@ -33,4 +33,18 @@ JNIEXPORT jlong JNICALL Java_com_nvidia_spark_rapids_jni_ParseURI_parseProtocol(
   }
   CATCH_STD(env, 0);
 }
+
+JNIEXPORT jlong JNICALL Java_com_nvidia_spark_rapids_jni_ParseURI_parseHost(JNIEnv* env,
+                                                                            jclass,
+                                                                            jlong input_column)
+{
+  JNI_NULL_CHECK(env, input_column, "input column is null", 0);
+
+  try {
+    cudf::jni::auto_set_device(env);
+    auto const input = reinterpret_cast<cudf::column_view const*>(input_column);
+    return cudf::jni::ptr_as_jlong(spark_rapids_jni::parse_uri_to_host(*input).release());
+  }
+  CATCH_STD(env, 0);
+}
 }

src/main/cpp/src/parse_uri.cu

@@ -292,9 +292,10 @@ bool __device__ validate_domain_name(string_view name)
 {
   // domain name can be alphanum or -.
   // slash can not be the first of last character of the domain name or around a .
-  bool last_was_slash  = false;
-  bool last_was_period = false;
-  bool numeric_start   = false;
+  bool last_was_slash          = false;
+  bool last_was_period         = false;
+  bool numeric_start           = false;
+  int characters_before_period = 0;
   for (auto iter = name.begin(); iter < name.end(); ++iter) {
     auto const c = *iter;
     if (!is_alphanum(c) && c != '-' && c != '.') { return false; }
@@ -311,12 +312,14 @@ bool __device__ validate_domain_name(string_view name)
       last_was_slash  = true;
       last_was_period = false;
     } else if (c == '.') {
-      if (last_was_slash) { return false; }
-      last_was_period = true;
-      last_was_slash  = false;
+      if (last_was_slash || last_was_period || characters_before_period == 0) { return false; }
+      last_was_period          = true;
+      last_was_slash           = false;
+      characters_before_period = 0;
     } else {
       last_was_period = false;
       last_was_slash  = false;
+      characters_before_period++;
     }
   }
 
@@ -847,4 +850,12 @@ std::unique_ptr<column> parse_uri_to_protocol(strings_column_view const& input,
   return detail::parse_uri(input, detail::URI_chunks::PROTOCOL, stream, mr);
 }
 
+std::unique_ptr<column> parse_uri_to_host(strings_column_view const& input,
+                                          rmm::cuda_stream_view stream,
+                                          rmm::mr::device_memory_resource* mr)
+{
+  CUDF_FUNC_RANGE();
+  return detail::parse_uri(input, detail::URI_chunks::HOST, stream, mr);
+}
+
 }  // namespace spark_rapids_jni

src/main/cpp/src/parse_uri.hpp

@@ -39,4 +39,17 @@ std::unique_ptr<cudf::column> parse_uri_to_protocol(
   rmm::cuda_stream_view stream        = cudf::get_default_stream(),
   rmm::mr::device_memory_resource* mr = rmm::mr::get_current_device_resource());
 
+/**
+ * @brief Parse host and copy from the input string column to the output char buffer.
+ *
+ * @param input Input string column of URIs to parse
+ * @param stream Stream on which to operate.
+ * @param mr Memory resource for returned column
+ * @return std::unique_ptr<column> String column of hosts parsed.
+ */
+std::unique_ptr<cudf::column> parse_uri_to_host(
+  cudf::strings_column_view const& input,
+  rmm::cuda_stream_view stream        = rmm::cuda_stream_default,
+  rmm::mr::device_memory_resource* mr = rmm::mr::get_current_device_resource());
+
 }  // namespace spark_rapids_jni