2 * Copyright 2007 Google Inc.
4 * Licensed under the Apache License, Version 2.0 (the "License");
5 * you may not use this file except in compliance with the License.
6 * You may obtain a copy of the License at
8 * http://www.apache.org/licenses/LICENSE-2.0
10 * Unless required by applicable law or agreed to in writing, software
11 * distributed under the License is distributed on an "AS IS" BASIS,
12 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13 * See the License for the specific language governing permissions and
14 * limitations under the License.
17 package com.google.zxing.qrcode.decoder;
19 import com.google.zxing.ReaderException;
20 import com.google.zxing.common.BitSource;
22 import java.io.UnsupportedEncodingException;
25 * <p>QR Codes can encode text as bits in one of several modes, and can use multiple modes
26 * in one QR Code. This class decodes the bits back into text.</p>
28 * <p>See ISO 18004:2006, 6.4.3 - 6.4.7</p>
30 * @author srowen@google.com (Sean Owen)
32 final class DecodedBitStreamParser {
35 * See ISO 18004:2006, 6.4.4 Table 5
37 private static final char[] ALPHANUMERIC_CHARS = {
38 '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', 'A', 'B',
39 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J', 'K', 'L', 'M', 'N',
40 'O', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z',
41 ' ', '$', '%', '*', '+', '-', '.', '/', ':'
43 private static final String SHIFT_JIS = "Shift_JIS";
44 private static final String EUC_JP = "EUC-JP";
45 private static final boolean ASSUME_SHIFT_JIS;
46 private static final String UTF8 = "UTF-8";
47 private static final String ISO88591 = "ISO-8859-1";
50 String platformDefault = System.getProperty("file.encoding");
51 ASSUME_SHIFT_JIS = SHIFT_JIS.equalsIgnoreCase(platformDefault) || EUC_JP.equalsIgnoreCase(platformDefault);
54 private DecodedBitStreamParser() {
57 static String decode(byte[] bytes, Version version) throws ReaderException {
58 BitSource bits = new BitSource(bytes);
59 StringBuffer result = new StringBuffer();
62 // While still another segment to read...
63 if (bits.available() == 0) {
64 // OK, assume we're done. Really, a TERMINATOR mode should have been recorded here
65 mode = Mode.TERMINATOR;
67 mode = Mode.forBits(bits.readBits(4)); // mode is encoded by 4 bits
69 if (!mode.equals(Mode.TERMINATOR)) {
70 if (mode.equals(Mode.ECI)) {
71 // Count doesn't apply to ECI
73 // We don't currently do anything with ECI, since there seems to be no reference
74 // defining what each value means. AIM's "Extended Channel Interpretations" does
75 // not define it. I have never observed a QR Code using it. So for now, we at least
76 // parse it but don't know how to take action on it.
78 // How many characters will follow, encoded in this mode?
79 int count = bits.readBits(mode.getCharacterCountBits(version));
80 if (mode.equals(Mode.NUMERIC)) {
81 decodeNumericSegment(bits, result, count);
82 } else if (mode.equals(Mode.ALPHANUMERIC)) {
83 decodeAlphanumericSegment(bits, result, count);
84 } else if (mode.equals(Mode.BYTE)) {
85 decodeByteSegment(bits, result, count);
86 } else if (mode.equals(Mode.KANJI)) {
87 decodeKanjiSegment(bits, result, count);
89 throw new ReaderException("Unsupported mode indicator");
93 } while (!mode.equals(Mode.TERMINATOR));
95 // I thought it wasn't allowed to leave extra bytes after the terminator but it happens
97 int bitsLeft = bits.available();
99 if (bitsLeft > 6 || bits.readBits(bitsLeft) != 0) {
100 throw new ReaderException("Excess bits or non-zero bits after terminator mode indicator");
104 return result.toString();
107 private static int parseECI(BitSource bits) {
108 int firstByte = bits.readBits(8);
109 if (firstByte & 0x80 == 0) {
111 return firstByte & 0x7F;
112 } else if (firstByte & 0xC0 == 0x80) {
114 int secondByte = bits.readBits(8);
115 return ((firstByte & 0x3F) << 8) | secondByte;
116 } else if (firstByte & 0xE0 == 0xC0) {
118 int secondByte = bits.readBits(8);
119 int thirdByte = bits.readBits(8);
120 return ((firstByte & 0x1F) << 16) | (secondByte << 8) | thirdByte;
124 private static void decodeKanjiSegment(BitSource bits,
126 int count) throws ReaderException {
127 // Each character will require 2 bytes. Read the characters as 2-byte pairs
128 // and decode as Shift_JIS afterwards
129 byte[] buffer = new byte[2 * count];
132 // Each 13 bits encodes a 2-byte character
133 int twoBytes = bits.readBits(13);
134 int assembledTwoBytes = ((twoBytes / 0x0C0) << 8) | (twoBytes % 0x0C0);
135 if (assembledTwoBytes < 0x01F00) {
136 // In the 0x8140 to 0x9FFC range
137 assembledTwoBytes += 0x08140;
139 // In the 0xE040 to 0xEBBF range
140 assembledTwoBytes += 0x0C140;
142 buffer[offset] = (byte) (assembledTwoBytes >> 8);
143 buffer[offset + 1] = (byte) assembledTwoBytes;
147 // Shift_JIS may not be supported in some environments:
149 result.append(new String(buffer, SHIFT_JIS));
150 } catch (UnsupportedEncodingException uee) {
151 throw new ReaderException(SHIFT_JIS + " encoding is not supported on this device");
155 private static void decodeByteSegment(BitSource bits,
157 int count) throws ReaderException {
158 byte[] readBytes = new byte[count];
159 if (count << 3 > bits.available()) {
160 throw new ReaderException("Count too large: " + count);
162 for (int i = 0; i < count; i++) {
163 readBytes[i] = (byte) bits.readBits(8);
165 // The spec isn't clear on this mode; see
166 // section 6.4.5: t does not say which encoding to assuming
167 // upon decoding. I have seen ISO-8859-1 used as well as
168 // Shift_JIS -- without anything like an ECI designator to
170 String encoding = guessEncoding(readBytes);
172 result.append(new String(readBytes, encoding));
173 } catch (UnsupportedEncodingException uce) {
174 throw new ReaderException(uce.toString());
178 private static void decodeAlphanumericSegment(BitSource bits,
181 // Read two characters at a time
183 int nextTwoCharsBits = bits.readBits(11);
184 result.append(ALPHANUMERIC_CHARS[nextTwoCharsBits / 45]);
185 result.append(ALPHANUMERIC_CHARS[nextTwoCharsBits % 45]);
189 // special case: one character left
190 result.append(ALPHANUMERIC_CHARS[bits.readBits(6)]);
194 private static void decodeNumericSegment(BitSource bits,
196 int count) throws ReaderException {
197 // Read three digits at a time
199 // Each 10 bits encodes three digits
200 int threeDigitsBits = bits.readBits(10);
201 if (threeDigitsBits >= 1000) {
202 throw new ReaderException("Illegal value for 3-digit unit: " + threeDigitsBits);
204 result.append(ALPHANUMERIC_CHARS[threeDigitsBits / 100]);
205 result.append(ALPHANUMERIC_CHARS[(threeDigitsBits / 10) % 10]);
206 result.append(ALPHANUMERIC_CHARS[threeDigitsBits % 10]);
210 // Two digits left over to read, encoded in 7 bits
211 int twoDigitsBits = bits.readBits(7);
212 if (twoDigitsBits >= 100) {
213 throw new ReaderException("Illegal value for 2-digit unit: " + twoDigitsBits);
215 result.append(ALPHANUMERIC_CHARS[twoDigitsBits / 10]);
216 result.append(ALPHANUMERIC_CHARS[twoDigitsBits % 10]);
217 } else if (count == 1) {
218 // One digit left over to read
219 int digitBits = bits.readBits(4);
220 if (digitBits >= 10) {
221 throw new ReaderException("Illegal value for digit unit: " + digitBits);
223 result.append(ALPHANUMERIC_CHARS[digitBits]);
227 private static String guessEncoding(byte[] bytes) {
228 if (ASSUME_SHIFT_JIS) {
231 // Does it start with the UTF-8 byte order mark? then guess it's UTF-8
232 if (bytes.length > 3 && bytes[0] == (byte) 0xEF && bytes[1] == (byte) 0xBB && bytes[2] == (byte) 0xBF) {
235 // For now, merely tries to distinguish ISO-8859-1, UTF-8 and Shift_JIS,
236 // which should be by far the most common encodings. ISO-8859-1
237 // should not have bytes in the 0x80 - 0x9F range, while Shift_JIS
238 // uses this as a first byte of a two-byte character. If we see this
239 // followed by a valid second byte in Shift_JIS, assume it is Shift_JIS.
240 // If we see something else in that second byte, we'll make the risky guess
242 int length = bytes.length;
243 boolean canBeISO88591 = true;
244 boolean lastWasPossibleDoubleByteStart = false;
245 for (int i = 0; i < length; i++) {
246 int value = bytes[i] & 0xFF;
247 if (value >= 0x80 && value <= 0x9F && i < length - 1) {
248 canBeISO88591 = false;
249 // ISO-8859-1 shouldn't use this, but before we decide it is Shift_JIS,
250 // just double check that it is followed by a byte that's valid in
251 // the Shift_JIS encoding
252 if (lastWasPossibleDoubleByteStart) {
253 // If we just checked this and the last byte for being a valid double-byte
254 // char, don't check starting on this byte. If the this and the last byte
255 // formed a valid pair, then this shouldn't be checked to see if it starts
256 // a double byte pair of course.
257 lastWasPossibleDoubleByteStart = false;
259 // ... otherwise do check to see if this plus the next byte form a valid
260 // double byte pair encoding a character.
261 lastWasPossibleDoubleByteStart = true;
262 int nextValue = bytes[i + 1] & 0xFF;
263 if ((value & 0x1) == 0) {
264 // if even, next value should be in [0x9F,0xFC]
265 // if not, we'll guess UTF-8
266 if (nextValue < 0x9F || nextValue > 0xFC) {
270 // if odd, next value should be in [0x40,0x9E]
271 // if not, we'll guess UTF-8
272 if (nextValue < 0x40 || nextValue > 0x9E) {
279 return canBeISO88591 ? ISO88591 : SHIFT_JIS;