| File: | _build/../src/icu-decoder.cc |
| Warning: | line 108, column 35 This statement is never executed |
Press '?' to see keyboard shortcuts
Keyboard shortcuts:
| 1 | /* |
| 2 | * Copyright © 2019 Christian Persch |
| 3 | * |
| 4 | * This library is free software; you can redistribute it and/or |
| 5 | * modify it under the terms of the GNU Lesser General Public |
| 6 | * License as published by the Free Software Foundation; either |
| 7 | * version 3 of the License, or (at your option) any later version. |
| 8 | * |
| 9 | * This library is distributed in the hope that it will be useful, |
| 10 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 11 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
| 12 | * Lesser General Public License for more details. |
| 13 | * |
| 14 | * You should have received a copy of the GNU General Public License |
| 15 | * along with this program. If not, see <https://www.gnu.org/licenses/>. |
| 16 | */ |
| 17 | |
| 18 | #include "config.h" |
| 19 | |
| 20 | #include <glib.h> |
| 21 | |
| 22 | #include <cassert> |
| 23 | #include <memory> |
| 24 | |
| 25 | #include "icu-decoder.hh" |
| 26 | |
| 27 | namespace bte::base { |
| 28 | |
| 29 | /* |
| 30 | * ICUDecoder::decode: |
| 31 | * @sptr: inout pointer to input data |
| 32 | * @flush: whether to flush |
| 33 | * |
| 34 | * Decodes input, and advances *@sptr for input consumed. At most |
| 35 | * one byte of input is consumed; if flushing, no input is consumed. |
| 36 | * |
| 37 | * Returns: whether there is an output character available |
| 38 | */ |
| 39 | ICUDecoder::Result |
| 40 | ICUDecoder::decode(uint8_t const** sptr, |
| 41 | bool flush) noexcept |
| 42 | { |
| 43 | switch (m_state) { |
| 44 | case State::eOutput: |
| 45 | if (++m_index < m_available) |
| 46 | return Result::eSomething; |
| 47 | |
| 48 | m_state = State::eInput; |
| 49 | [[fallthrough]]; |
| 50 | case State::eInput: { |
| 51 | /* Convert in two stages from charset to UTF-32, pivoting through UTF-16. |
| 52 | * This is similar to ucnv_convertEx(), but that API does not fit our |
| 53 | * requirements completely. |
| 54 | * |
| 55 | * This function is similar to ucnv_getNextUChar, except that it works |
| 56 | * with streaming (and thus may produce no output in some steps), while |
| 57 | * ucnv_getNextUChar does not. |
| 58 | */ |
| 59 | |
| 60 | auto source_ptr = reinterpret_cast<char const**>(sptr); |
| 61 | auto source_start = *source_ptr; |
| 62 | auto source_limit = source_start + (1 - flush); |
| 63 | |
| 64 | auto target_u16_start = u16_buffer(); |
| 65 | auto target_u16_limit = u16_buffer_end(); |
| 66 | auto target_u16 = target_u16_start; |
| 67 | |
| 68 | m_err.reset(); |
| 69 | ucnv_toUnicodeucnv_toUnicode_76(m_charset_converter.get(), |
| 70 | &target_u16, target_u16_limit, |
| 71 | source_ptr, source_limit, |
| 72 | nullptr /* offsets */, |
| 73 | flush, |
| 74 | m_err); |
| 75 | |
| 76 | /* There should be no error here. We use the default callback |
| 77 | * which replaces invalid input with replacment character (either |
| 78 | * U+FFFD or SUB), and we should never hit U_BUFFER_OVERFLOW_ERROR, |
| 79 | * since we process only one input byte at a time and the output |
| 80 | * buffer holds at most 1 UTF-16 character (a high surrogate), and |
| 81 | * there are no charsets where the state is so deep as to suddently |
| 82 | * output 32 characters. |
| 83 | */ |
| 84 | if (m_err.isFailure()) { |
| 85 | m_state = State::eError; |
| 86 | return Result::eError; |
| 87 | } |
| 88 | |
| 89 | /* Now convert from UTF-16 to UTF-32. There will be no overflow here |
| 90 | * either, since the output buffer is empty, and for each UTF-16 code |
| 91 | * point of input, the decoder will output at most one UTF-32 code |
| 92 | * point. |
| 93 | */ |
| 94 | |
| 95 | auto target_u32_start = reinterpret_cast<char*>(u32_buffer()); |
| 96 | auto target_u32_limit = reinterpret_cast<char const*>(u32_buffer_end()); |
| 97 | auto target_u32 = target_u32_start; |
| 98 | auto target_u16_u32_start = const_cast<char16_t const*>(target_u16_start); |
| 99 | auto target_u16_u32_limit = target_u16; |
| 100 | |
| 101 | ucnv_fromUnicodeucnv_fromUnicode_76(m_u32_converter.get(), |
| 102 | &target_u32, target_u32_limit, |
| 103 | &target_u16_u32_start, target_u16_u32_limit, |
| 104 | nullptr /* offsets */, |
| 105 | flush, |
| 106 | m_err); |
| 107 | if (m_err.isFailure()) { |
| 108 | m_state = State::eError; |
This statement is never executed | |
| 109 | return Result::eError; |
| 110 | } |
| 111 | |
| 112 | if (target_u32 == target_u32_start) { |
| 113 | if (*source_ptr == source_start && !flush) { |
| 114 | /* The decoder produced neither output nor consumed input, and |
| 115 | * wan't flushing? That shouldn't happen; go to error state, |
| 116 | * requiring an explicit reset() to proceed further. |
| 117 | */ |
| 118 | m_state = State::eError; |
| 119 | return Result::eError; |
| 120 | } |
| 121 | |
| 122 | return Result::eNothing; |
| 123 | } |
| 124 | |
| 125 | /* We have some output! */ |
| 126 | assert((target_u32 - target_u32_start) % sizeof(m_u32_buffer[0]) == 0)(static_cast <bool> ((target_u32 - target_u32_start) % sizeof (m_u32_buffer[0]) == 0) ? void (0) : __assert_fail ("(target_u32 - target_u32_start) % sizeof(m_u32_buffer[0]) == 0" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 127 | m_available = (target_u32 - target_u32_start) / sizeof(m_u32_buffer[0]); |
| 128 | assert(m_available >= 1)(static_cast <bool> (m_available >= 1) ? void (0) : __assert_fail ("m_available >= 1", __builtin_FILE (), __builtin_LINE () , __extension__ __PRETTY_FUNCTION__)); |
| 129 | |
| 130 | m_index = 0; |
| 131 | return Result::eSomething; |
| 132 | } |
| 133 | |
| 134 | case State::eError: |
| 135 | default: |
| 136 | return Result::eError; |
| 137 | } |
| 138 | } |
| 139 | |
| 140 | void |
| 141 | ICUDecoder::reset() noexcept |
| 142 | { |
| 143 | ucnv_resetToUnicodeucnv_resetToUnicode_76(m_charset_converter.get()); |
| 144 | ucnv_resetFromUnicodeucnv_resetFromUnicode_76(m_u32_converter.get()); |
| 145 | m_err.reset(); |
| 146 | m_state = State::eInput; |
| 147 | m_available = 0; |
| 148 | m_index = 0; |
| 149 | } |
| 150 | |
| 151 | } // namespace bte::base |