| 1 | /**************************************************************************** |
| 2 | ** |
| 3 | ** Copyright (C) 2016 The Qt Company Ltd. |
| 4 | ** Contact: https://www.qt.io/licensing/ |
| 5 | ** |
| 6 | ** This file is part of the QtNetwork module of the Qt Toolkit. |
| 7 | ** |
| 8 | ** $QT_BEGIN_LICENSE:LGPL$ |
| 9 | ** Commercial License Usage |
| 10 | ** Licensees holding valid commercial Qt licenses may use this file in |
| 11 | ** accordance with the commercial license agreement provided with the |
| 12 | ** Software or, alternatively, in accordance with the terms contained in |
| 13 | ** a written agreement between you and The Qt Company. For licensing terms |
| 14 | ** and conditions see https://www.qt.io/terms-conditions. For further |
| 15 | ** information use the contact form at https://www.qt.io/contact-us. |
| 16 | ** |
| 17 | ** GNU Lesser General Public License Usage |
| 18 | ** Alternatively, this file may be used under the terms of the GNU Lesser |
| 19 | ** General Public License version 3 as published by the Free Software |
| 20 | ** Foundation and appearing in the file LICENSE.LGPL3 included in the |
| 21 | ** packaging of this file. Please review the following information to |
| 22 | ** ensure the GNU Lesser General Public License version 3 requirements |
| 23 | ** will be met: https://www.gnu.org/licenses/lgpl-3.0.html. |
| 24 | ** |
| 25 | ** GNU General Public License Usage |
| 26 | ** Alternatively, this file may be used under the terms of the GNU |
| 27 | ** General Public License version 2.0 or (at your option) the GNU General |
| 28 | ** Public license version 3 or any later version approved by the KDE Free |
| 29 | ** Qt Foundation. The licenses are as published by the Free Software |
| 30 | ** Foundation and appearing in the file LICENSE.GPL2 and LICENSE.GPL3 |
| 31 | ** included in the packaging of this file. Please review the following |
| 32 | ** information to ensure the GNU General Public License requirements will |
| 33 | ** be met: https://www.gnu.org/licenses/gpl-2.0.html and |
| 34 | ** https://www.gnu.org/licenses/gpl-3.0.html. |
| 35 | ** |
| 36 | ** $QT_END_LICENSE$ |
| 37 | ** |
| 38 | ****************************************************************************/ |
| 39 | |
| 40 | #include "bitstreams_p.h" |
| 41 | #include "huffman_p.h" |
| 42 | |
| 43 | #include <QtCore/qbytearray.h> |
| 44 | |
| 45 | #include <limits> |
| 46 | |
| 47 | QT_BEGIN_NAMESPACE |
| 48 | |
| 49 | static_assert(std::numeric_limits<uchar>::digits == 8, "octets expected" ); |
| 50 | |
| 51 | namespace HPack |
| 52 | { |
| 53 | |
| 54 | BitOStream::BitOStream(std::vector<uchar> &b) |
| 55 | : buffer(b), |
| 56 | // All data 'packed' before: |
| 57 | bitsSet(8 * quint64(b.size())) |
| 58 | { |
| 59 | } |
| 60 | |
| 61 | void BitOStream::writeBits(uchar bits, quint8 bitLength) |
| 62 | { |
| 63 | Q_ASSERT(bitLength <= 8); |
| 64 | |
| 65 | quint8 count = bitsSet % 8; // bits used in buffer.back(), but 0 means 8 |
| 66 | bits <<= 8 - bitLength; // at top of byte, lower bits clear |
| 67 | if (count) { // we have a part-used byte; fill it some more: |
| 68 | buffer.back() |= bits >> count; |
| 69 | count = 8 - count; |
| 70 | } // count bits have been consumed (and 0 now means 0) |
| 71 | if (bitLength > count) |
| 72 | buffer.push_back(bits << count); |
| 73 | |
| 74 | bitsSet += bitLength; |
| 75 | } |
| 76 | |
| 77 | void BitOStream::write(quint32 src) |
| 78 | { |
| 79 | const quint8 prefixLen = 8 - bitsSet % 8; |
| 80 | const quint32 fullPrefix = (1 << prefixLen) - 1; |
| 81 | |
| 82 | // https://http2.github.io/http2-spec/compression.html#low-level.representation, |
| 83 | // 5.1 |
| 84 | if (src < fullPrefix) { |
| 85 | writeBits(uchar(src), prefixLen); |
| 86 | } else { |
| 87 | writeBits(uchar(fullPrefix), prefixLen); |
| 88 | // We're on the byte boundary now, |
| 89 | // so we can just 'push_back'. |
| 90 | Q_ASSERT(!(bitsSet % 8)); |
| 91 | src -= fullPrefix; |
| 92 | while (src >= 128) { |
| 93 | buffer.push_back(uchar(src % 128 + 128)); |
| 94 | src /= 128; |
| 95 | bitsSet += 8; |
| 96 | } |
| 97 | buffer.push_back(src); |
| 98 | bitsSet += 8; |
| 99 | } |
| 100 | } |
| 101 | |
| 102 | void BitOStream::write(const QByteArray &src, bool compressed) |
| 103 | { |
| 104 | quint32 byteLen = src.size(); |
| 105 | if (compressed && byteLen) { |
| 106 | const auto bitLen = huffman_encoded_bit_length(src); |
| 107 | Q_ASSERT(bitLen && std::numeric_limits<quint32>::max() >= (bitLen + 7) / 8); |
| 108 | byteLen = (bitLen + 7) / 8; |
| 109 | writeBits(uchar(1), 1); // bit set - compressed |
| 110 | } else { |
| 111 | writeBits(uchar(0), 1); // no compression. |
| 112 | } |
| 113 | |
| 114 | write(byteLen); |
| 115 | |
| 116 | if (compressed) { |
| 117 | huffman_encode_string(src, *this); |
| 118 | } else { |
| 119 | bitsSet += quint64(src.size()) * 8; |
| 120 | buffer.insert(buffer.end(), src.begin(), src.end()); |
| 121 | } |
| 122 | } |
| 123 | |
| 124 | quint64 BitOStream::bitLength() const |
| 125 | { |
| 126 | return bitsSet; |
| 127 | } |
| 128 | |
| 129 | quint64 BitOStream::byteLength() const |
| 130 | { |
| 131 | return buffer.size(); |
| 132 | } |
| 133 | |
| 134 | const uchar *BitOStream::begin() const |
| 135 | { |
| 136 | return &buffer[0]; |
| 137 | } |
| 138 | |
| 139 | const uchar *BitOStream::end() const |
| 140 | { |
| 141 | return &buffer[0] + buffer.size(); |
| 142 | } |
| 143 | |
| 144 | void BitOStream::clear() |
| 145 | { |
| 146 | buffer.clear(); |
| 147 | bitsSet = 0; |
| 148 | } |
| 149 | |
| 150 | BitIStream::BitIStream() |
| 151 | : first(), |
| 152 | last(), |
| 153 | offset(), |
| 154 | streamError(Error::NoError) |
| 155 | { |
| 156 | } |
| 157 | |
| 158 | BitIStream::BitIStream(const uchar *begin, const uchar *end) |
| 159 | : first(begin), |
| 160 | last(end), |
| 161 | offset(), |
| 162 | streamError(Error::NoError) |
| 163 | { |
| 164 | } |
| 165 | |
| 166 | quint64 BitIStream::bitLength() const |
| 167 | { |
| 168 | return quint64(last - first) * 8; |
| 169 | } |
| 170 | |
| 171 | bool BitIStream::hasMoreBits() const |
| 172 | { |
| 173 | return offset < bitLength(); |
| 174 | } |
| 175 | |
| 176 | bool BitIStream::skipBits(quint64 nBits) |
| 177 | { |
| 178 | if (nBits > bitLength() || bitLength() - nBits < offset) |
| 179 | return false; |
| 180 | |
| 181 | offset += nBits; |
| 182 | return true; |
| 183 | } |
| 184 | |
| 185 | bool BitIStream::rewindOffset(quint64 nBits) |
| 186 | { |
| 187 | if (nBits > offset) |
| 188 | return false; |
| 189 | |
| 190 | offset -= nBits; |
| 191 | return true; |
| 192 | } |
| 193 | |
| 194 | bool BitIStream::read(quint32 *dstPtr) |
| 195 | { |
| 196 | Q_ASSERT(dstPtr); |
| 197 | quint32 &dst = *dstPtr; |
| 198 | |
| 199 | // 5.1 Integer Representation |
| 200 | // |
| 201 | // Integers are used to represent name indexes, header field indexes, or string lengths. |
| 202 | // An integer representation can start anywhere within an octet. |
| 203 | // To allow for optimized processing, an integer representation always finishes at the end of an octet. |
| 204 | // An integer is represented in two parts: a prefix that fills the current octet and an optional |
| 205 | // list of octets that are used if the integer value does not fit within the prefix. |
| 206 | // The number of bits of the prefix (called N) is a parameter of the integer representation. |
| 207 | // If the integer value is small enough, i.e., strictly less than 2N-1, it is compressed within the N-bit prefix. |
| 208 | // ... |
| 209 | // The prefix size, N, is always between 1 and 8 bits. An integer |
| 210 | // starting at an octet boundary will have an 8-bit prefix. |
| 211 | |
| 212 | // Technically, such integers can be of any size, but as we do not have arbitrary-long integers, |
| 213 | // everything that does not fit into 'dst' we consider as an error (after all, try to allocate a string |
| 214 | // of such size and ... hehehe - send it as a part of a header! |
| 215 | |
| 216 | // This function updates the offset _only_ if the read was successful. |
| 217 | if (offset >= bitLength()) { |
| 218 | setError(Error::NotEnoughData); |
| 219 | return false; |
| 220 | } |
| 221 | |
| 222 | setError(Error::NoError); |
| 223 | |
| 224 | const quint32 prefixLen = 8 - offset % 8; |
| 225 | const quint32 fullPrefix = (1 << prefixLen) - 1; |
| 226 | |
| 227 | const uchar prefix = uchar(first[offset / 8] & fullPrefix); |
| 228 | if (prefix < fullPrefix) { |
| 229 | // The number fitted into the prefix bits. |
| 230 | dst = prefix; |
| 231 | offset += prefixLen; |
| 232 | return true; |
| 233 | } |
| 234 | |
| 235 | quint32 newOffset = offset + prefixLen; |
| 236 | // We have a list of bytes representing an integer ... |
| 237 | quint64 val = prefix; |
| 238 | quint32 octetPower = 0; |
| 239 | |
| 240 | while (true) { |
| 241 | if (newOffset >= bitLength()) { |
| 242 | setError(Error::NotEnoughData); |
| 243 | return false; |
| 244 | } |
| 245 | |
| 246 | const uchar octet = first[newOffset / 8]; |
| 247 | |
| 248 | if (octetPower == 28 && octet > 15) { |
| 249 | qCritical("integer is too big" ); |
| 250 | setError(Error::InvalidInteger); |
| 251 | return false; |
| 252 | } |
| 253 | |
| 254 | val += quint32(octet & 0x7f) << octetPower; |
| 255 | newOffset += 8; |
| 256 | |
| 257 | if (!(octet & 0x80)) { |
| 258 | // The most significant bit of each octet is used |
| 259 | // as a continuation flag: its value is set to 1 |
| 260 | // except for the last octet in the list. |
| 261 | break; |
| 262 | } |
| 263 | |
| 264 | octetPower += 7; |
| 265 | } |
| 266 | |
| 267 | dst = val; |
| 268 | offset = newOffset; |
| 269 | Q_ASSERT(!(offset % 8)); |
| 270 | |
| 271 | return true; |
| 272 | } |
| 273 | |
| 274 | bool BitIStream::read(QByteArray *dstPtr) |
| 275 | { |
| 276 | Q_ASSERT(dstPtr); |
| 277 | QByteArray &dst = *dstPtr; |
| 278 | //5.2 String Literal Representation |
| 279 | // |
| 280 | // Header field names and header field values can be represented as string literals. |
| 281 | // A string literal is compressed as a sequence of octets, either by directly encoding |
| 282 | // the string literal's octets or by using a Huffman code. |
| 283 | |
| 284 | // We update the offset _only_ if the read was successful. |
| 285 | |
| 286 | const quint64 oldOffset = offset; |
| 287 | uchar compressed = 0; |
| 288 | if (peekBits(offset, 1, &compressed) != 1 || !skipBits(1)) { |
| 289 | setError(Error::NotEnoughData); |
| 290 | return false; |
| 291 | } |
| 292 | |
| 293 | setError(Error::NoError); |
| 294 | |
| 295 | quint32 len = 0; |
| 296 | if (read(&len)) { |
| 297 | Q_ASSERT(!(offset % 8)); |
| 298 | if (len <= (bitLength() - offset) / 8) { // We have enough data to read a string ... |
| 299 | if (!compressed) { |
| 300 | // Now good news, integer always ends on a byte boundary. |
| 301 | // We can read 'len' bytes without any bit magic. |
| 302 | const char *src = reinterpret_cast<const char *>(first + offset / 8); |
| 303 | dst = QByteArray(src, len); |
| 304 | offset += quint64(len) * 8; |
| 305 | return true; |
| 306 | } |
| 307 | |
| 308 | BitIStream slice(first + offset / 8, first + offset / 8 + len); |
| 309 | if (huffman_decode_string(slice, &dst)) { |
| 310 | offset += quint64(len) * 8; |
| 311 | return true; |
| 312 | } |
| 313 | |
| 314 | setError(Error::CompressionError); |
| 315 | } else { |
| 316 | setError(Error::NotEnoughData); |
| 317 | } |
| 318 | } // else the exact reason was set by read(quint32). |
| 319 | |
| 320 | offset = oldOffset; |
| 321 | return false; |
| 322 | } |
| 323 | |
| 324 | BitIStream::Error BitIStream::error() const |
| 325 | { |
| 326 | return streamError; |
| 327 | } |
| 328 | |
| 329 | void BitIStream::setError(Error newState) |
| 330 | { |
| 331 | streamError = newState; |
| 332 | } |
| 333 | |
| 334 | } // namespace HPack |
| 335 | |
| 336 | QT_END_NAMESPACE |
| 337 | |