Refactor v2 parser (#149)

* Minor changes to serializing IPv4 and IPv6 addresses

* Updated concept names

* Applied clang-format

* Added script to generate IDNA table data for v2

* Refactored punycode function internals

* Refactored host parsing and serialization functions so that they use fmt and ranges libraries (to be later replaced with the std versions)

* Refactored url_parser_context.hpp

* Refactored part of the parser

* These are only minor changes

Author: glynos

CMakeLists.txt

@@ -25,7 +25,7 @@ option(skyr_USE_STATIC_CRT "Use static C Runtime library (/MT or MTd)." ON)
 option(skyr_BUILD_WITH_LLVM_LIBCXX "Instruct Clang to use LLVM's implementation of C++ standard library" OFF)
 option(skyr_ENABLE_FILESYSTEM_FUNCTIONS "Enable functions to convert URL to std::filesystem::path" ON)
 option(skyr_ENABLE_JSON_FUNCTIONS "Enable functions to convert URL components to JSON" ON)
-option(skyr_BUILD_V2 "Build v2, which uses C++20 features" ON)
+option(skyr_BUILD_V2 "Build v2, which uses C++20 features" OFF)
 
 set(CMAKE_VERBOSE_MAKEFILE true)
 if (skyr_BUILD_V2)

include/skyr/v2/concepts/url_concepts.hpp

@@ -1,4 +1,4 @@
-// Copyright (c) Glyn Matthews 2018-19.
+// Copyright (c) Glyn Matthews 2018-20.
 // Distributed under the Boost Software License, Version 1.0.
 // (See accompanying file LICENSE_1_0.txt or copy at
 // http://www.boost.org/LICENSE_1_0.txt)
@@ -25,13 +25,13 @@ template <class T, class charT>
 concept is_char_pointer = std::conjunction_v<std::is_pointer<T>, std::is_same<std::remove_pointer_t<T>, charT>>;
 
 template <class T, class charT>
-concept is_string_convertible =
+concept is_string_container =
     is_basic_string<T, charT> || is_basic_string_view<T, charT> || is_char_array<T, charT> || is_char_pointer<T, charT>;
 
 template <typename T>
-concept is_url_convertible =
-    is_string_convertible<T, char> || is_string_convertible<T, char8_t> || is_string_convertible<T, wchar_t> ||
-    is_string_convertible<T, char16_t> || is_string_convertible<T, char32_t>;
-}  // namespace skyr::v2
+concept is_u8_convertible =
+    is_string_container<T, char> || is_string_container<T, char8_t> || is_string_container<T, wchar_t> ||
+    is_string_container<T, char16_t> || is_string_container<T, char32_t>;
+}  // namespace skyr::inline v2
 
 #endif  // SKYR_V2_CONCEPTS_URL_CONCEPTS_HPP

include/skyr/v2/containers/static_vector.hpp

@@ -15,20 +15,15 @@ namespace skyr::inline v2 {
 ///
 /// \tparam T
 /// \tparam Capacity
-template <
-    class T,
-    std::size_t Capacity
-    >
+template <class T, std::size_t Capacity>
 class static_vector {
  private:
-
   using impl_type = std::array<T, Capacity>;
 
   impl_type impl_;
   std::size_t size_ = 0;
 
  public:
-
   ///
   using value_type = T;
   ///
@@ -105,8 +100,7 @@ class static_vector {
   /// \pre `size() < capacity()`
   /// \post `size() > 0 && size() <= capacity()`
   template <class... Args>
-  constexpr auto emplace_back(Args &&... args)
-    noexcept(std::is_trivially_move_assignable_v<T>) -> reference {
+  constexpr auto emplace_back(Args &&... args) noexcept(std::is_trivially_move_assignable_v<T>) -> reference {
     impl_[size_++] = value_type{std::forward<Args>(args)...};
     return back();
   }
@@ -143,7 +137,13 @@ class static_vector {
   }
 
   ///
-  /// \return `true` if there are elements
+  /// \return
+  [[nodiscard]] constexpr auto capacity() const noexcept -> size_type {
+    return Capacity;
+  }
+
+  ///
+  /// \return `true` if there are no elements
   [[nodiscard]] constexpr auto empty() const noexcept -> bool {
     return size_ == 0;
   }
@@ -195,8 +195,7 @@ class static_vector {
   [[nodiscard]] constexpr auto end() const noexcept -> const_iterator {
     return cend();
   }
-
 };
-}  // namespace skyr::v2
+}  // namespace skyr::inline v2
 
 #endif  // SKYR_V2_CONTAINERS_STATIC_VECTOR_HPP

include/skyr/v2/core/check_input.hpp

@@ -12,7 +12,7 @@
 #include <iterator>
 
 namespace skyr::inline v2 {
-constexpr static auto is_c0_control_or_space = [] (auto byte) {
+constexpr static auto is_c0_control_or_space = [](auto byte) {
   return std::iscntrl(byte, std::locale::classic()) || std::isspace(byte, std::locale::classic());
 };
 
@@ -31,16 +31,6 @@ constexpr inline auto remove_trailing_c0_control_or_space(std::string_view input
   input.remove_suffix(std::distance(first, it));
   return input;
 }
-
-inline auto remove_tabs_and_newlines(std::string &input, bool *validation_error) {
-  constexpr auto is_tab_or_newline = [] (auto byte) {
-    return (byte == '\t') || (byte == '\r') || (byte == '\n');
-  };
-
-  auto it = std::remove_if(std::begin(input), std::end(input), is_tab_or_newline);
-  *validation_error |= (it != std::cend(input));
-  input.erase(it, std::cend(input));
-}
-}  // namespace skyr::v2
+}  // namespace skyr::inline v2
 
 #endif  // SKYR_V2_CORE_CHECK_INPUT_HPP

include/skyr/v2/core/errors.hpp

@@ -43,24 +43,36 @@ enum class url_parse_errc {
 namespace details {
 class url_parse_error_category : public std::error_category {
  public:
-  [[nodiscard]] auto name() const noexcept -> const char * override {
+  [[nodiscard]] auto name() const noexcept -> const char* override {
     return "url parse";
   }
 
   [[nodiscard]] auto message(int error) const noexcept -> std::string override {
     switch (static_cast<url_parse_errc>(error)) {
-      case url_parse_errc::invalid_scheme_character:return "Invalid URL scheme";
-      case url_parse_errc::not_an_absolute_url_with_fragment:return "Not an absolute URL with fragment";
-      case url_parse_errc::empty_hostname:return "Empty hostname";
-      case url_parse_errc::invalid_ipv4_address:return "Invalid IPv4 address";
-      case url_parse_errc::invalid_ipv6_address:return "Invalid IPv6 address";
-      case url_parse_errc::forbidden_host_point:return "Forbidden host point";
-      case url_parse_errc::cannot_decode_host_point:return "Cannot decode host point";
-      case url_parse_errc::domain_error:return "Domain error";
-      case url_parse_errc::cannot_be_a_base_url:return "Cannot be a base URL";
-      case url_parse_errc::cannot_have_a_username_password_or_port:return "Cannot have a username, password or port";
-      case url_parse_errc::invalid_port:return "Invalid port";
-      default:return "(Unknown error)";
+      case url_parse_errc::invalid_scheme_character:
+        return "Invalid URL scheme";
+      case url_parse_errc::not_an_absolute_url_with_fragment:
+        return "Not an absolute URL with fragment";
+      case url_parse_errc::empty_hostname:
+        return "Empty hostname";
+      case url_parse_errc::invalid_ipv4_address:
+        return "Invalid IPv4 address";
+      case url_parse_errc::invalid_ipv6_address:
+        return "Invalid IPv6 address";
+      case url_parse_errc::forbidden_host_point:
+        return "Forbidden host point";
+      case url_parse_errc::cannot_decode_host_point:
+        return "Cannot decode host point";
+      case url_parse_errc::domain_error:
+        return "Domain error";
+      case url_parse_errc::cannot_be_a_base_url:
+        return "Cannot be a base URL";
+      case url_parse_errc::cannot_have_a_username_password_or_port:
+        return "Cannot have a username, password or port";
+      case url_parse_errc::invalid_port:
+        return "Invalid port";
+      default:
+        return "(Unknown error)";
     }
   }
 };
@@ -73,11 +85,11 @@ inline auto make_error_code(url_parse_errc error) noexcept -> std::error_code {
   static const details::url_parse_error_category category{};
   return std::error_code(static_cast<int>(error), category);
 }
-}  // namespace skyr::v2
+}  // namespace skyr::inline v2
 
 namespace std {
 template <>
 struct is_error_code_enum<skyr::v2::url_parse_errc> : true_type {};
 }  // namespace std
 
-#endif // SKYR_V2_CORE_ERRORS_HPP
+#endif  // SKYR_V2_CORE_ERRORS_HPP