diff options
Diffstat (limited to 'include/lib/modernjson/detail/input/input_adapters.hpp')
| -rw-r--r-- | include/lib/modernjson/detail/input/input_adapters.hpp | 169 |
1 files changed, 95 insertions, 74 deletions
diff --git a/include/lib/modernjson/detail/input/input_adapters.hpp b/include/lib/modernjson/detail/input/input_adapters.hpp index 5abaee3..dfb8caf 100644 --- a/include/lib/modernjson/detail/input/input_adapters.hpp +++ b/include/lib/modernjson/detail/input/input_adapters.hpp @@ -18,7 +18,7 @@ namespace nlohmann namespace detail { /// the supported input formats -enum class input_format_t { json, cbor, msgpack, ubjson }; +enum class input_format_t { json, cbor, msgpack, ubjson, bson }; //////////////////// // input adapters // @@ -60,8 +60,8 @@ class input_stream_adapter : public input_adapter_protocol ~input_stream_adapter() override { // clear stream flags; we use underlying streambuf I/O, do not - // maintain ifstream flags - is.clear(); + // maintain ifstream flags, except eof + is.clear(is.rdstate() & std::ios::eofbit); } explicit input_stream_adapter(std::istream& i) @@ -71,13 +71,21 @@ class input_stream_adapter : public input_adapter_protocol // delete because of pointer members input_stream_adapter(const input_stream_adapter&) = delete; input_stream_adapter& operator=(input_stream_adapter&) = delete; + input_stream_adapter(input_stream_adapter&&) = delete; + input_stream_adapter& operator=(input_stream_adapter&&) = delete; // std::istream/std::streambuf use std::char_traits<char>::to_int_type, to // ensure that std::char_traits<char>::eof() and the character 0xFF do not // end up as the same value, eg. 0xFFFFFFFF. std::char_traits<char>::int_type get_character() override { - return sb.sbumpc(); + auto res = sb.sbumpc(); + // set eof manually, as we don't use the istream interface. + if (res == EOF) + { + is.clear(is.rdstate() | std::ios::eofbit); + } + return res; } private: @@ -90,13 +98,16 @@ class input_stream_adapter : public input_adapter_protocol class input_buffer_adapter : public input_adapter_protocol { public: - input_buffer_adapter(const char* b, const std::size_t l) + input_buffer_adapter(const char* b, const std::size_t l) noexcept : cursor(b), limit(b + l) {} // delete because of pointer members input_buffer_adapter(const input_buffer_adapter&) = delete; input_buffer_adapter& operator=(input_buffer_adapter&) = delete; + input_buffer_adapter(input_buffer_adapter&&) = delete; + input_buffer_adapter& operator=(input_buffer_adapter&&) = delete; + ~input_buffer_adapter() override = default; std::char_traits<char>::int_type get_character() noexcept override { @@ -115,38 +126,66 @@ class input_buffer_adapter : public input_adapter_protocol const char* const limit; }; -template<typename WideStringType> -class wide_string_input_adapter : public input_adapter_protocol +template<typename WideStringType, size_t T> +struct wide_string_input_helper { - public: - explicit wide_string_input_adapter(const WideStringType& w) : str(w) {} - - std::char_traits<char>::int_type get_character() noexcept override + // UTF-32 + static void fill_buffer(const WideStringType& str, size_t& current_wchar, std::array<std::char_traits<char>::int_type, 4>& utf8_bytes, size_t& utf8_bytes_index, size_t& utf8_bytes_filled) { - // check if buffer needs to be filled - if (utf8_bytes_index == utf8_bytes_filled) + utf8_bytes_index = 0; + + if (current_wchar == str.size()) + { + utf8_bytes[0] = std::char_traits<char>::eof(); + utf8_bytes_filled = 1; + } + else { - if (sizeof(typename WideStringType::value_type) == 2) + // get the current character + const auto wc = static_cast<int>(str[current_wchar++]); + + // UTF-32 to UTF-8 encoding + if (wc < 0x80) + { + utf8_bytes[0] = wc; + utf8_bytes_filled = 1; + } + else if (wc <= 0x7FF) + { + utf8_bytes[0] = 0xC0 | ((wc >> 6) & 0x1F); + utf8_bytes[1] = 0x80 | (wc & 0x3F); + utf8_bytes_filled = 2; + } + else if (wc <= 0xFFFF) { - fill_buffer_utf16(); + utf8_bytes[0] = 0xE0 | ((wc >> 12) & 0x0F); + utf8_bytes[1] = 0x80 | ((wc >> 6) & 0x3F); + utf8_bytes[2] = 0x80 | (wc & 0x3F); + utf8_bytes_filled = 3; + } + else if (wc <= 0x10FFFF) + { + utf8_bytes[0] = 0xF0 | ((wc >> 18) & 0x07); + utf8_bytes[1] = 0x80 | ((wc >> 12) & 0x3F); + utf8_bytes[2] = 0x80 | ((wc >> 6) & 0x3F); + utf8_bytes[3] = 0x80 | (wc & 0x3F); + utf8_bytes_filled = 4; } else { - fill_buffer_utf32(); + // unknown character + utf8_bytes[0] = wc; + utf8_bytes_filled = 1; } - - assert(utf8_bytes_filled > 0); - assert(utf8_bytes_index == 0); } - - // use buffer - assert(utf8_bytes_filled > 0); - assert(utf8_bytes_index < utf8_bytes_filled); - return utf8_bytes[utf8_bytes_index++]; } +}; - private: - void fill_buffer_utf16() +template<typename WideStringType> +struct wide_string_input_helper<WideStringType, 2> +{ + // UTF-16 + static void fill_buffer(const WideStringType& str, size_t& current_wchar, std::array<std::char_traits<char>::int_type, 4>& utf8_bytes, size_t& utf8_bytes_index, size_t& utf8_bytes_filled) { utf8_bytes_index = 0; @@ -158,7 +197,7 @@ class wide_string_input_adapter : public input_adapter_protocol else { // get the current character - const int wc = static_cast<int>(str[current_wchar++]); + const auto wc = static_cast<int>(str[current_wchar++]); // UTF-16 to UTF-8 encoding if (wc < 0x80) @@ -183,7 +222,7 @@ class wide_string_input_adapter : public input_adapter_protocol { if (current_wchar < str.size()) { - const int wc2 = static_cast<int>(str[current_wchar++]); + const auto wc2 = static_cast<int>(str[current_wchar++]); const int charcode = 0x10000 + (((wc & 0x3FF) << 10) | (wc2 & 0x3FF)); utf8_bytes[0] = 0xf0 | (charcode >> 18); utf8_bytes[1] = 0x80 | ((charcode >> 12) & 0x3F); @@ -201,58 +240,40 @@ class wide_string_input_adapter : public input_adapter_protocol } } } +}; - void fill_buffer_utf32() - { - utf8_bytes_index = 0; +template<typename WideStringType> +class wide_string_input_adapter : public input_adapter_protocol +{ + public: + explicit wide_string_input_adapter(const WideStringType& w) noexcept + : str(w) + {} - if (current_wchar == str.size()) - { - utf8_bytes[0] = std::char_traits<char>::eof(); - utf8_bytes_filled = 1; - } - else + std::char_traits<char>::int_type get_character() noexcept override + { + // check if buffer needs to be filled + if (utf8_bytes_index == utf8_bytes_filled) { - // get the current character - const int wc = static_cast<int>(str[current_wchar++]); + fill_buffer<sizeof(typename WideStringType::value_type)>(); - // UTF-32 to UTF-8 encoding - if (wc < 0x80) - { - utf8_bytes[0] = wc; - utf8_bytes_filled = 1; - } - else if (wc <= 0x7FF) - { - utf8_bytes[0] = 0xC0 | ((wc >> 6) & 0x1F); - utf8_bytes[1] = 0x80 | (wc & 0x3F); - utf8_bytes_filled = 2; - } - else if (wc <= 0xFFFF) - { - utf8_bytes[0] = 0xE0 | ((wc >> 12) & 0x0F); - utf8_bytes[1] = 0x80 | ((wc >> 6) & 0x3F); - utf8_bytes[2] = 0x80 | (wc & 0x3F); - utf8_bytes_filled = 3; - } - else if (wc <= 0x10FFFF) - { - utf8_bytes[0] = 0xF0 | ((wc >> 18 ) & 0x07); - utf8_bytes[1] = 0x80 | ((wc >> 12) & 0x3F); - utf8_bytes[2] = 0x80 | ((wc >> 6) & 0x3F); - utf8_bytes[3] = 0x80 | (wc & 0x3F); - utf8_bytes_filled = 4; - } - else - { - // unknown character - utf8_bytes[0] = wc; - utf8_bytes_filled = 1; - } + assert(utf8_bytes_filled > 0); + assert(utf8_bytes_index == 0); } + + // use buffer + assert(utf8_bytes_filled > 0); + assert(utf8_bytes_index < utf8_bytes_filled); + return utf8_bytes[utf8_bytes_index++]; } private: + template<size_t T> + void fill_buffer() + { + wide_string_input_helper<WideStringType, T>::fill_buffer(str, current_wchar, utf8_bytes, utf8_bytes_index, utf8_bytes_filled); + } + /// the wstring to process const WideStringType& str; @@ -373,5 +394,5 @@ class input_adapter /// the actual adapter input_adapter_t ia = nullptr; }; -} -} +} // namespace detail +} // namespace nlohmann |