/* This file is part of TON Blockchain Library. TON Blockchain Library is free software: you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation, either version 2 of the License, or (at your option) any later version. TON Blockchain Library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details. You should have received a copy of the GNU Lesser General Public License along with TON Blockchain Library. If not, see . Copyright 2017-2019 Telegram Systems LLP */ #pragma once #include "td/utils/buffer.h" #include "td/utils/common.h" #include "td/utils/format.h" #include "td/utils/logging.h" #include "td/utils/misc.h" #include "td/utils/Slice.h" #include "td/utils/Status.h" #include "td/utils/UInt.h" #include "td/utils/utf8.h" #include #include #include #include #include namespace td { class TlParser { const unsigned char *data = nullptr; size_t data_len = 0; size_t left_len = 0; size_t error_pos = std::numeric_limits::max(); std::string error; std::unique_ptr data_buf; static constexpr size_t SMALL_DATA_ARRAY_SIZE = 6; std::array small_data_array; alignas(4) static const unsigned char empty_data[sizeof(UInt256)]; public: explicit TlParser(Slice slice) { data_len = left_len = slice.size(); if (is_aligned_pointer<4>(slice.begin())) { data = slice.ubegin(); } else { int32 *buf; if (data_len <= small_data_array.size() * sizeof(int32)) { buf = &small_data_array[0]; } else { LOG(ERROR) << "Unexpected big unaligned data pointer of length " << slice.size() << " at " << slice.begin(); data_buf = std::make_unique(1 + data_len / sizeof(int32)); buf = data_buf.get(); } std::memcpy(buf, slice.begin(), slice.size()); data = reinterpret_cast(buf); } } TlParser(const TlParser &other) = delete; TlParser &operator=(const TlParser &other) = delete; void set_error(const string &error_message); const char *get_error() const { if (error.empty()) { return nullptr; } return error.c_str(); } size_t get_error_pos() const { return error_pos; } Status get_status() const { if (error.empty()) { return Status::OK(); } return Status::Error(PSLICE() << error << " at " << error_pos); } void check_len(const size_t len) { if (unlikely(left_len < len)) { set_error("Not enough data to read"); } else { left_len -= len; } } int32 fetch_int_unsafe() { int32 result; std::memcpy(&result, data, sizeof(int32)); data += sizeof(int32); return result; } int32 fetch_int() { check_len(sizeof(int32)); return fetch_int_unsafe(); } int64 fetch_long_unsafe() { int64 result; std::memcpy(&result, data, sizeof(int64)); data += sizeof(int64); return result; } int64 fetch_long() { check_len(sizeof(int64)); return fetch_long_unsafe(); } double fetch_double_unsafe() { double result; std::memcpy(&result, data, sizeof(double)); data += sizeof(double); return result; } double fetch_double() { check_len(sizeof(double)); return fetch_double_unsafe(); } template T fetch_binary_unsafe() { T result; std::memcpy(&result, data, sizeof(T)); data += sizeof(T); return result; } template T fetch_binary() { static_assert(sizeof(T) <= sizeof(empty_data), "too big fetch_binary"); //static_assert(sizeof(T) % sizeof(int32) == 0, "wrong call to fetch_binary"); check_len(sizeof(T)); return fetch_binary_unsafe(); } template T fetch_string() { check_len(sizeof(int32)); size_t result_len = *data; const char *result_begin; size_t result_aligned_len; if (result_len < 254) { result_begin = reinterpret_cast(data + 1); result_aligned_len = (result_len >> 2) << 2; data += sizeof(int32); } else if (result_len == 254) { result_len = data[1] + (data[2] << 8) + (data[3] << 16); result_begin = reinterpret_cast(data + 4); result_aligned_len = ((result_len + 3) >> 2) << 2; data += sizeof(int32); } else { check_len(sizeof(int32)); result_len = data[1] + (data[2] << 8) + (data[3] << 16) + (data[4] << 24) + (static_cast(data[5]) << 32) + (static_cast(data[6]) << 40) + (static_cast(data[7]) << 48); if (result_len > std::numeric_limits::max() - 3) { set_error("Too big string found"); return T(); } result_begin = reinterpret_cast(data + 8); result_aligned_len = ((result_len + 3) >> 2) << 2; data += sizeof(int64); } check_len(result_aligned_len); if (!error.empty()) { return T(); } data += result_aligned_len; return T(result_begin, result_len); } template T fetch_string_raw(const size_t size) { //CHECK(size % sizeof(int32) == 0); check_len(size); if (!error.empty()) { return T(); } const char *result = reinterpret_cast(data); data += size; return T(result, size); } void fetch_end() { if (left_len) { set_error("Too much data to fetch"); } } size_t get_left_len() const { return left_len; } }; class TlBufferParser : public TlParser { public: explicit TlBufferParser(const BufferSlice *buffer_slice) : TlParser(buffer_slice->as_slice()), parent_(buffer_slice) { } template T fetch_string() { auto result = TlParser::fetch_string(); for (auto &c : result) { if (c == '\0') { c = ' '; } } if (check_utf8(result)) { return result; } CHECK(!result.empty()); LOG(WARNING) << "Wrong UTF-8 string [[" << result << "]] in " << format::as_hex_dump<4>(parent_->as_slice()); // trying to remove last character size_t new_size = result.size() - 1; while (new_size != 0 && !is_utf8_character_first_code_unit(static_cast(result[new_size]))) { new_size--; } result.resize(new_size); if (check_utf8(result)) { return result; } return T(); } template T fetch_string_raw(const size_t size) { return TlParser::fetch_string_raw(size); } private: const BufferSlice *parent_; BufferSlice as_buffer_slice(Slice slice) { if (is_aligned_pointer<4>(slice.data())) { return parent_->from_slice(slice); } return BufferSlice(slice); } }; template <> inline BufferSlice TlBufferParser::fetch_string() { return as_buffer_slice(TlParser::fetch_string()); } template <> inline BufferSlice TlBufferParser::fetch_string_raw(const size_t size) { return as_buffer_slice(TlParser::fetch_string_raw(size)); } } // namespace td