Fix crash for empty/invalid/not-found filepaths

Add tests to validate empty string path
Add tests to validate non-existent string path
Add tests to validate nullptr buffer parser

6 files changed, 47 insertions, 39 deletions

diff --git a/src/openvic-dataloader/AbstractSyntaxTree.hpp b/src/openvic-dataloader/AbstractSyntaxTree.hpp
index f9f5796..f2d941b 100644
--- a/src/openvic-dataloader/AbstractSyntaxTree.hpp
+++ b/src/openvic-dataloader/AbstractSyntaxTree.hpp
@@ -23,6 +23,7 @@
 
 namespace ovdl {
 	struct AbstractSyntaxTree : SymbolIntern {
+		AbstractSyntaxTree() = default;
 		explicit AbstractSyntaxTree(std::size_t max_elements) : _symbol_interner(max_elements) {}
 
 		symbol_type intern(const char* str, std::size_t length);
@@ -51,6 +52,8 @@ namespace ovdl {
 		using root_node_type = RootNodeT;
 		using node_type = typename file_type::node_type;
 
+		BasicAbstractSyntaxTree() = default;
+
 		explicit BasicAbstractSyntaxTree(file_type&& file)
 			: AbstractSyntaxTree(file.size() * file.visit_buffer([](auto&& buffer) -> size_t { return sizeof(typename std::decay_t<decltype(buffer)>::char_type); })),
 			  _file { std::move(file) } {}
diff --git a/src/openvic-dataloader/File.hpp b/src/openvic-dataloader/File.hpp
index ec25640..cdab377 100644
--- a/src/openvic-dataloader/File.hpp
+++ b/src/openvic-dataloader/File.hpp
@@ -23,6 +23,7 @@ namespace ovdl {
 			lexy::buffer<lexy::utf32_encoding, void>,
 			lexy::buffer<lexy::byte_encoding, void>>;
 
+		File() = default;
 		explicit File(const char* path);
 
 		const char* path() const noexcept;
@@ -74,6 +75,7 @@ namespace ovdl {
 		decltype(auto) visit_buffer(Visitor&& visitor) {
 			switch (_buffer.index()) {
 				SWITCH_LIST
+				case 0: return visitor(lexy::buffer<> {});
 				default: ovdl::detail::unreachable();
 			}
 		}
@@ -82,6 +84,7 @@ namespace ovdl {
 		Return visit_buffer(Visitor&& visitor) {
 			switch (_buffer.index()) {
 				SWITCH_LIST
+				case 0: return visitor(lexy::buffer<> {});
 				default: ovdl::detail::unreachable();
 			}
 		}
@@ -90,6 +93,7 @@ namespace ovdl {
 		decltype(auto) visit_buffer(Visitor&& visitor) const {
 			switch (_buffer.index()) {
 				SWITCH_LIST
+				case 0: return visitor(lexy::buffer<> {});
 				default: ovdl::detail::unreachable();
 			}
 		}
@@ -98,6 +102,7 @@ namespace ovdl {
 		Return visit_buffer(Visitor&& visitor) const {
 			switch (_buffer.index()) {
 				SWITCH_LIST
+				case 0: return visitor(lexy::buffer<> {});
 				default: ovdl::detail::unreachable();
 			}
 		}
@@ -105,7 +110,7 @@ namespace ovdl {
 #undef SWITCH_LIST
 
 	protected:
-		const char* _path;
+		const char* _path = "";
 		std::size_t _buffer_size = 0;
 		detail::type_prepend_t<buffer_ids::variant_type, std::monostate> _buffer;
 	};
@@ -114,6 +119,8 @@ namespace ovdl {
 	struct BasicFile : File {
 		using node_type = NodeT;
 
+		BasicFile() = default;
+
 		template<typename Encoding, typename MemoryResource = void>
 		explicit BasicFile(const char* path, lexy::buffer<Encoding, MemoryResource>&& buffer)
 			: File(path) {
diff --git a/src/openvic-dataloader/ParseState.hpp b/src/openvic-dataloader/ParseState.hpp
index 806829c..259f2a5 100644
--- a/src/openvic-dataloader/ParseState.hpp
+++ b/src/openvic-dataloader/ParseState.hpp
@@ -30,6 +30,8 @@ namespace ovdl {
 		using file_type = typename ast_type::file_type;
 		using diagnostic_logger_type = BasicDiagnosticLogger<ParseState>;
 
+		ParseState() : _ast {}, _logger { this->ast().file() } {}
+
 		ParseState(typename ast_type::file_type&& file, detail::Encoding encoding)
 			: _ast { std::move(file) },
 			  _logger { this->ast().file() },
@@ -69,6 +71,8 @@ namespace ovdl {
 		using file_type = FileT;
 		using diagnostic_logger_type = BasicDiagnosticLogger<FileParseState>;
 
+		FileParseState() : _file {}, _logger { this->file() } {}
+
 		FileParseState(file_type&& file, detail::Encoding encoding)
 			: _file { std::move(file) },
 			  _logger { this->file() },
diff --git a/src/openvic-dataloader/csv/Parser.cpp b/src/openvic-dataloader/csv/Parser.cpp
index bbd1be4..e4416b2 100644
--- a/src/openvic-dataloader/csv/Parser.cpp
+++ b/src/openvic-dataloader/csv/Parser.cpp
@@ -57,9 +57,6 @@ struct Parser::ParseHandler final : detail::BasicFileParseHandler<CsvParseState>
 	}
 
 	Parser::error_range get_errors() {
-		using iterator = typename decltype(std::declval<const error::Root*>()->children())::iterator;
-		if (!is_valid())
-			return dryad::make_node_range<error::Error>(iterator::from_ptr(nullptr), iterator::from_ptr(nullptr));
 		return parse_state().logger().get_errors();
 	}
 
@@ -129,7 +126,7 @@ constexpr void Parser::_run_load_func(detail::LoadCallback<ParseHandler, Args...
 	if (!error_message.empty()) {
 		_has_error = true;
 		_has_fatal_error = true;
-		_parse_handler->parse_state().logger().template create_log<error::BufferError>(DiagnosticLogger::DiagnosticKind::error, fmt::runtime(error_message));
+		_parse_handler->parse_state().logger().template create_log<error::BufferError>(DiagnosticLogger::DiagnosticKind::error, fmt::runtime(error_message), _file_path);
 	}
 	if (has_error() && &_error_stream.get() != &detail::cnull) {
 		print_errors_to(_error_stream.get());
@@ -235,7 +232,7 @@ void Parser::print_errors_to(std::basic_ostream<char>& stream) const {
 		dryad::visit_tree(
 			error,
 			[&](const error::BufferError* buffer_error) {
-				stream << "buffer error: " << this->error(buffer_error) << '\n';
+				stream << this->error(buffer_error) << '\n';
 			},
 			[&](dryad::child_visitor<error::ErrorKind> visitor, const error::AnnotatedError* annotated_error) {
 				stream << this->error(annotated_error) << '\n';
diff --git a/src/openvic-dataloader/detail/ParseHandler.hpp b/src/openvic-dataloader/detail/ParseHandler.hpp
index 9666a5b..28222ae 100644
--- a/src/openvic-dataloader/detail/ParseHandler.hpp
+++ b/src/openvic-dataloader/detail/ParseHandler.hpp
@@ -26,11 +26,11 @@ namespace ovdl::detail {
 				case buffer_is_null:
 					return "Buffer could not be loaded.";
 				case os_error:
-					return "OS file error.";
+					return "OS file error for '{}'.";
 				case file_not_found:
-					return "File not found.";
+					return "File '{}' not found.";
 				case permission_denied:
-					return "Read Permission denied.";
+					return "Read Permission for '{}' denied.";
 				default:
 					return "";
 			}
@@ -78,19 +78,20 @@ namespace ovdl::detail {
 		virtual const char* path_impl() const = 0;
 
 		template<detail::IsStateType State, detail::IsEncoding BufferEncoding>
-		static constexpr auto generate_state = [](std::optional<State>* state, const char* path, auto&& buffer, Encoding encoding) {
+		static constexpr auto generate_state = [](State* state, const char* path, auto&& buffer, Encoding encoding) {
 			if (path[0] != '\0') {
-				state->emplace(
+				*state = {
 					path,
 					lexy::buffer<BufferEncoding>(std::move(buffer)),
-					encoding);
+					encoding
+				};
 				return;
 			}
-			state->emplace(lexy::buffer<BufferEncoding>(std::move(buffer)), encoding);
+			*state = { lexy::buffer<BufferEncoding>(std::move(buffer)), encoding };
 		};
 
 		template<detail::IsStateType State>
-		static void create_state(std::optional<State>* state, const char* path, lexy::buffer<lexy::default_encoding>&& buffer, std::optional<Encoding> fallback) {
+		static void create_state(State* state, const char* path, lexy::buffer<lexy::default_encoding>&& buffer, std::optional<Encoding> fallback) {
 			if (!_system_fallback_encoding.has_value()) {
 				_detect_system_fallback_encoding();
 			}
@@ -121,15 +122,15 @@ namespace ovdl::detail {
 			}
 
 			if (!is_alone) {
-				(*state)->logger().info("encoding type could not be distinguished");
+				state->logger().info("encoding type could not be distinguished");
 			}
 
 			if (is_bad_fallback) {
-				(*state)->logger().warning("fallback encoding cannot be ascii or utf8");
+				state->logger().warning("fallback encoding cannot be ascii or utf8");
 			}
 
 			if (encoding == ovdl::detail::Encoding::Unknown) {
-				(*state)->logger().warning("could not detect encoding");
+				state->logger().warning("could not detect encoding");
 			}
 		}
 
@@ -143,36 +144,37 @@ namespace ovdl::detail {
 		using parse_state_type = ParseState;
 
 		virtual constexpr bool is_valid_impl() const {
-			if (!_parse_state) return false;
-			return _parse_state.value().file().is_valid();
+			return _parse_state.file().is_valid();
 		}
 
 		constexpr virtual buffer_error load_buffer_impl(lexy::buffer<lexy::default_encoding>&& buffer, const char* path, std::optional<Encoding> fallback) {
-			if (buffer.data() == nullptr) return buffer_error::buffer_is_null;
+			if (buffer.data() == nullptr) {
+				_parse_state = {};
+				return buffer_error::buffer_is_null;
+			}
 			create_state(&_parse_state, path, std::move(buffer), fallback);
 			return is_valid_impl() ? buffer_error::success : buffer_error::buffer_is_null;
 		}
 
 		virtual const char* path_impl() const {
-			if (!_parse_state) return "";
-			return _parse_state.value().file().path();
+			return _parse_state.file().path();
 		}
 
 		parse_state_type& parse_state() {
-			return _parse_state.value();
+			return _parse_state;
 		}
 
 		const parse_state_type& parse_state() const {
-			return _parse_state.value();
+			return _parse_state;
 		}
 
 		template<typename Encoding>
 		constexpr const auto& buffer() const {
-			return _parse_state.value().file().template get_buffer_as<Encoding>();
+			return _parse_state.file().template get_buffer_as<Encoding>();
 		}
 
 	protected:
-		std::optional<parse_state_type> _parse_state;
+		parse_state_type _parse_state;
 	};
 
 	template<detail::IsParseState ParseState>
@@ -180,8 +182,7 @@ namespace ovdl::detail {
 		using parse_state_type = ParseState;
 
 		virtual constexpr bool is_valid_impl() const {
-			if (!_parse_state) return false;
-			return _parse_state.value().ast().file().is_valid();
+			return _parse_state.ast().file().is_valid();
 		}
 
 		constexpr virtual buffer_error load_buffer_impl(lexy::buffer<lexy::default_encoding>&& buffer, const char* path, std::optional<Encoding> fallback) {
@@ -191,24 +192,23 @@ namespace ovdl::detail {
 		}
 
 		virtual const char* path_impl() const {
-			if (!_parse_state) return "";
-			return _parse_state.value().ast().file().path();
+			return _parse_state.ast().file().path();
 		}
 
 		parse_state_type& parse_state() {
-			return _parse_state.value();
+			return _parse_state;
 		}
 
 		const parse_state_type& parse_state() const {
-			return _parse_state.value();
+			return _parse_state;
 		}
 
 		template<typename Encoding>
 		constexpr const auto& buffer() const {
-			return _parse_state.value().ast().file().template get_buffer_as<Encoding>();
+			return _parse_state.ast().file().template get_buffer_as<Encoding>();
 		}
 
 	protected:
-		std::optional<parse_state_type> _parse_state;
+		parse_state_type _parse_state;
 	};
 }
\ No newline at end of file
diff --git a/src/openvic-dataloader/v2script/Parser.cpp b/src/openvic-dataloader/v2script/Parser.cpp
index e5234d6..9b36a95 100644
--- a/src/openvic-dataloader/v2script/Parser.cpp
+++ b/src/openvic-dataloader/v2script/Parser.cpp
@@ -77,9 +77,6 @@ struct Parser::ParseHandler final : detail::BasicStateParseHandler<v2script::ast
 	}
 
 	Parser::error_range get_errors() {
-		using iterator = typename decltype(std::declval<const error::Root*>()->children())::iterator;
-		if (!is_valid())
-			return dryad::make_node_range<error::Error>(iterator::from_ptr(nullptr), iterator::from_ptr(nullptr));
 		return parse_state().logger().get_errors();
 	}
 };
@@ -146,7 +143,7 @@ constexpr void Parser::_run_load_func(detail::LoadCallback<Parser::ParseHandler*
 	if (!error_message.empty()) {
 		_has_error = true;
 		_has_fatal_error = true;
-		_parse_handler->parse_state().logger().template create_log<error::BufferError>(DiagnosticLogger::DiagnosticKind::error, fmt::runtime(error_message));
+		_parse_handler->parse_state().logger().template create_log<error::BufferError>(DiagnosticLogger::DiagnosticKind::error, fmt::runtime(error_message), _file_path);
 	}
 	if (has_error() && &_error_stream.get() != &detail::cnull) {
 		print_errors_to(_error_stream.get());
@@ -356,7 +353,7 @@ void Parser::print_errors_to(std::basic_ostream<char>& stream) const {
 		dryad::visit_tree(
 			error,
 			[&](const error::BufferError* buffer_error) {
-				stream << "buffer error: " << this->error(buffer_error) << '\n';
+				stream << this->error(buffer_error) << '\n';
 			},
 			[&](dryad::child_visitor<error::ErrorKind> visitor, const error::AnnotatedError* annotated_error) {
 				stream << this->error(annotated_error) << '\n';