2 * DecodedBitStreamParser.cpp
5 * Created by Christian Brunschen on 20/05/2008.
6 * Copyright 2008 ZXing authors All rights reserved.
8 * Licensed under the Apache License, Version 2.0 (the "License");
9 * you may not use this file except in compliance with the License.
10 * You may obtain a copy of the License at
12 * http://www.apache.org/licenses/LICENSE-2.0
14 * Unless required by applicable law or agreed to in writing, software
15 * distributed under the License is distributed on an "AS IS" BASIS,
16 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
17 * See the License for the specific language governing permissions and
18 * limitations under the License.
21 #include <zxing/qrcode/decoder/DecodedBitStreamParser.h>
25 // Required for compatibility. TODO: test on Symbian
26 #ifdef ZXING_ICONV_CONST
28 #define ICONV_CONST const
32 #define ICONV_CONST /**/
35 using namespace zxing;
42 const char DecodedBitStreamParser::ALPHANUMERIC_CHARS[] = { '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', 'A', 'B',
43 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X',
44 'Y', 'Z', ' ', '$', '%', '*', '+', '-', '.', '/', ':'
47 const char *DecodedBitStreamParser::ASCII = "ASCII";
48 const char *DecodedBitStreamParser::ISO88591 = "ISO-8859-1";
49 const char *DecodedBitStreamParser::UTF8 = "UTF-8";
50 const char *DecodedBitStreamParser::SHIFT_JIS = "SHIFT_JIS";
51 const char *DecodedBitStreamParser::EUC_JP = "EUC-JP";
53 void DecodedBitStreamParser::append(ostream &ost, const unsigned char *bufIn, size_t nIn, const char *src) {
58 iconv_t cd = iconv_open(UTF8, src);
59 const int maxOut = 4 * nIn + 1;
60 unsigned char* bufOut = new unsigned char[maxOut];
62 ICONV_CONST char *fromPtr = (ICONV_CONST char *)bufIn;
64 char *toPtr = (char *)bufOut;
68 size_t oneway = iconv(cd, &fromPtr, &nFrom, &toPtr, &nTo);
69 if (oneway == (size_t)(-1)) {
72 throw ReaderException("error converting characters");
77 int nResult = maxOut - nTo;
78 bufOut[nResult] = '\0';
84 void DecodedBitStreamParser::decodeKanjiSegment(Ref<BitSource> bits, ostringstream &result, int count) {
85 // Each character will require 2 bytes. Read the characters as 2-byte pairs
86 // and decode as Shift_JIS afterwards
87 size_t nBytes = 2 * count;
88 unsigned char* buffer = new unsigned char[nBytes];
91 // Each 13 bits encodes a 2-byte character
93 int twoBytes = bits->readBits(13);
94 int assembledTwoBytes = ((twoBytes / 0x0C0) << 8) | (twoBytes % 0x0C0);
95 if (assembledTwoBytes < 0x01F00) {
96 // In the 0x8140 to 0x9FFC range
97 assembledTwoBytes += 0x08140;
99 // In the 0xE040 to 0xEBBF range
100 assembledTwoBytes += 0x0C140;
102 buffer[offset] = (unsigned char)(assembledTwoBytes >> 8);
103 buffer[offset + 1] = (unsigned char)assembledTwoBytes;
108 append(result, buffer, nBytes, SHIFT_JIS);
112 void DecodedBitStreamParser::decodeByteSegment(Ref<BitSource> bits, ostringstream &result, int count) {
114 unsigned char* readBytes = new unsigned char[nBytes];
115 if (count << 3 > bits->available()) {
117 s << "Count too large: " << count;
119 throw ReaderException(s.str().c_str());
121 for (int i = 0; i < count; i++) {
122 readBytes[i] = (unsigned char)bits->readBits(8);
124 // The spec isn't clear on this mode; see
125 // section 6.4.5: t does not say which encoding to assuming
126 // upon decoding. I have seen ISO-8859-1 used as well as
127 // Shift_JIS -- without anything like an ECI designator to
129 const char *encoding = guessEncoding(readBytes, nBytes);
130 append(result, readBytes, nBytes, encoding);
134 void DecodedBitStreamParser::decodeNumericSegment(Ref<BitSource> bits, ostringstream &result, int count) {
136 unsigned char* bytes = new unsigned char[nBytes];
138 // Read three digits at a time
140 // Each 10 bits encodes three digits
141 int threeDigitsBits = bits->readBits(10);
142 if (threeDigitsBits >= 1000) {
144 s << "Illegal value for 3-digit unit: " << threeDigitsBits;
146 throw ReaderException(s.str().c_str());
148 bytes[i++] = ALPHANUMERIC_CHARS[threeDigitsBits / 100];
149 bytes[i++] = ALPHANUMERIC_CHARS[(threeDigitsBits / 10) % 10];
150 bytes[i++] = ALPHANUMERIC_CHARS[threeDigitsBits % 10];
154 // Two digits left over to read, encoded in 7 bits
155 int twoDigitsBits = bits->readBits(7);
156 if (twoDigitsBits >= 100) {
158 s << "Illegal value for 2-digit unit: " << twoDigitsBits;
160 throw ReaderException(s.str().c_str());
162 bytes[i++] = ALPHANUMERIC_CHARS[twoDigitsBits / 10];
163 bytes[i++] = ALPHANUMERIC_CHARS[twoDigitsBits % 10];
164 } else if (count == 1) {
165 // One digit left over to read
166 int digitBits = bits->readBits(4);
167 if (digitBits >= 10) {
169 s << "Illegal value for digit unit: " << digitBits;
171 throw ReaderException(s.str().c_str());
173 bytes[i++] = ALPHANUMERIC_CHARS[digitBits];
175 append(result, bytes, nBytes, ASCII);
179 void DecodedBitStreamParser::decodeAlphanumericSegment(Ref<BitSource> bits, ostringstream &result, int count) {
181 unsigned char* bytes = new unsigned char[nBytes];
183 // Read two characters at a time
185 int nextTwoCharsBits = bits->readBits(11);
186 bytes[i++] = ALPHANUMERIC_CHARS[nextTwoCharsBits / 45];
187 bytes[i++] = ALPHANUMERIC_CHARS[nextTwoCharsBits % 45];
191 bytes[i++] = ALPHANUMERIC_CHARS[bits->readBits(6)];
193 append(result, bytes, nBytes, ASCII);
198 DecodedBitStreamParser::guessEncoding(unsigned char *bytes, int length) {
199 // Does it start with the UTF-8 byte order mark? then guess it's UTF-8
200 if (length > 3 && bytes[0] == (unsigned char)0xEF && bytes[1] == (unsigned char)0xBB && bytes[2]
201 == (unsigned char)0xBF) {
204 // For now, merely tries to distinguish ISO-8859-1, UTF-8 and Shift_JIS,
205 // which should be by far the most common encodings. ISO-8859-1
206 // should not have bytes in the 0x80 - 0x9F range, while Shift_JIS
207 // uses this as a first byte of a two-byte character. If we see this
208 // followed by a valid second byte in Shift_JIS, assume it is Shift_JIS.
209 // If we see something else in that second byte, we'll make the risky guess
211 bool canBeISO88591 = true;
212 bool lastWasPossibleDoubleByteStart = false;
213 for (int i = 0; i < length; i++) {
214 int value = bytes[i] & 0xFF;
215 if (value >= 0x80 && value <= 0x9F && i < length - 1) {
216 canBeISO88591 = false;
217 // ISO-8859-1 shouldn't use this, but before we decide it is Shift_JIS,
218 // just double check that it is followed by a byte that's valid in
219 // the Shift_JIS encoding
220 if (lastWasPossibleDoubleByteStart) {
221 // If we just checked this and the last byte for being a valid double-byte
222 // char, don't check starting on this byte. If the this and the last byte
223 // formed a valid pair, then this shouldn't be checked to see if it starts
224 // a double byte pair of course.
225 lastWasPossibleDoubleByteStart = false;
227 // ... otherwise do check to see if this plus the next byte form a valid
228 // double byte pair encoding a character.
229 lastWasPossibleDoubleByteStart = true;
230 int nextValue = bytes[i + 1] & 0xFF;
231 if ((value & 0x1) == 0) {
232 // if even, next value should be in [0x9F,0xFC]
233 // if not, we'll guess UTF-8
234 if (nextValue < 0x9F || nextValue > 0xFC) {
238 // if odd, next value should be in [0x40,0x9E]
239 // if not, we'll guess UTF-8
240 if (nextValue < 0x40 || nextValue > 0x9E) {
247 return canBeISO88591 ? ISO88591 : SHIFT_JIS;
250 string DecodedBitStreamParser::decode(ArrayRef<unsigned char> bytes, Version *version) {
251 ostringstream result;
252 Ref<BitSource> bits(new BitSource(bytes));
253 Mode *mode = &Mode::TERMINATOR;
255 // While still another segment to read...
256 if (bits->available() < 4) {
257 // OK, assume we're done. Really, a TERMINATOR mode should have been recorded here
258 mode = &Mode::TERMINATOR;
260 mode = &Mode::forBits(bits->readBits(4)); // mode is encoded by 4 bits
262 if (mode != &Mode::TERMINATOR) {
263 // How many characters will follow, encoded in this mode?
264 int count = bits->readBits(mode->getCharacterCountBits(version));
265 if (mode == &Mode::NUMERIC) {
266 decodeNumericSegment(bits, result, count);
267 } else if (mode == &Mode::ALPHANUMERIC) {
268 decodeAlphanumericSegment(bits, result, count);
269 } else if (mode == &Mode::BYTE) {
270 decodeByteSegment(bits, result, count);
271 } else if (mode == &Mode::KANJI) {
272 decodeKanjiSegment(bits, result, count);
274 throw ReaderException("Unsupported mode indicator");
277 } while (mode != &Mode::TERMINATOR);