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 mode = Mode.forBits(bits.readBits(4)); // mode is encoded by 4 bits
64 if (!mode.equals(Mode.TERMINATOR)) {
65 // How many characters will follow, encoded in this mode?
66 int count = bits.readBits(mode.getCharacterCountBits(version));
67 if (mode.equals(Mode.NUMERIC)) {
68 decodeNumericSegment(bits, result, count);
69 } else if (mode.equals(Mode.ALPHANUMERIC)) {
70 decodeAlphanumericSegment(bits, result, count);
71 } else if (mode.equals(Mode.BYTE)) {
72 decodeByteSegment(bits, result, count);
73 } else if (mode.equals(Mode.KANJI)) {
74 decodeKanjiSegment(bits, result, count);
76 throw new ReaderException("Unsupported mode indicator");
79 } while (!mode.equals(Mode.TERMINATOR));
81 // I thought it wasn't allowed to leave extra bytes after the terminator but it happens
83 int bitsLeft = bits.available();
85 if (bitsLeft > 6 || bits.readBits(bitsLeft) != 0) {
86 throw new ReaderException("Excess bits or non-zero bits after terminator mode indicator");
90 return result.toString();
93 private static void decodeKanjiSegment(BitSource bits,
95 int count) throws ReaderException {
96 // Each character will require 2 bytes. Read the characters as 2-byte pairs
97 // and decode as Shift_JIS afterwards
98 byte[] buffer = new byte[2 * count];
101 // Each 13 bits encodes a 2-byte character
102 int twoBytes = bits.readBits(13);
103 int assembledTwoBytes = ((twoBytes / 0x0C0) << 8) | (twoBytes % 0x0C0);
104 if (assembledTwoBytes < 0x01F00) {
105 // In the 0x8140 to 0x9FFC range
106 assembledTwoBytes += 0x08140;
108 // In the 0xE040 to 0xEBBF range
109 assembledTwoBytes += 0x0C140;
111 buffer[offset] = (byte) (assembledTwoBytes >> 8);
112 buffer[offset + 1] = (byte) assembledTwoBytes;
116 // Shift_JIS may not be supported in some environments:
118 result.append(new String(buffer, SHIFT_JIS));
119 } catch (UnsupportedEncodingException uee) {
120 throw new ReaderException(SHIFT_JIS + " encoding is not supported on this device");
124 private static void decodeByteSegment(BitSource bits,
126 int count) throws ReaderException {
127 byte[] readBytes = new byte[count];
128 if (count << 3 > bits.available()) {
129 throw new ReaderException("Count too large: " + count);
131 for (int i = 0; i < count; i++) {
132 readBytes[i] = (byte) bits.readBits(8);
134 // The spec isn't clear on this mode; see
135 // section 6.4.5: t does not say which encoding to assuming
136 // upon decoding. I have seen ISO-8859-1 used as well as
137 // Shift_JIS -- without anything like an ECI designator to
139 String encoding = guessEncoding(readBytes);
141 result.append(new String(readBytes, encoding));
142 } catch (UnsupportedEncodingException uce) {
143 throw new ReaderException(uce.toString());
147 private static void decodeAlphanumericSegment(BitSource bits,
150 // Read two characters at a time
152 int nextTwoCharsBits = bits.readBits(11);
153 result.append(ALPHANUMERIC_CHARS[nextTwoCharsBits / 45]);
154 result.append(ALPHANUMERIC_CHARS[nextTwoCharsBits % 45]);
158 // special case: one character left
159 result.append(ALPHANUMERIC_CHARS[bits.readBits(6)]);
163 private static void decodeNumericSegment(BitSource bits,
165 int count) throws ReaderException {
166 // Read three digits at a time
168 // Each 10 bits encodes three digits
169 int threeDigitsBits = bits.readBits(10);
170 if (threeDigitsBits >= 1000) {
171 throw new ReaderException("Illegal value for 3-digit unit: " + threeDigitsBits);
173 result.append(ALPHANUMERIC_CHARS[threeDigitsBits / 100]);
174 result.append(ALPHANUMERIC_CHARS[(threeDigitsBits / 10) % 10]);
175 result.append(ALPHANUMERIC_CHARS[threeDigitsBits % 10]);
179 // Two digits left over to read, encoded in 7 bits
180 int twoDigitsBits = bits.readBits(7);
181 if (twoDigitsBits >= 100) {
182 throw new ReaderException("Illegal value for 2-digit unit: " + twoDigitsBits);
184 result.append(ALPHANUMERIC_CHARS[twoDigitsBits / 10]);
185 result.append(ALPHANUMERIC_CHARS[twoDigitsBits % 10]);
186 } else if (count == 1) {
187 // One digit left over to read
188 int digitBits = bits.readBits(4);
189 if (digitBits >= 10) {
190 throw new ReaderException("Illegal value for digit unit: " + digitBits);
192 result.append(ALPHANUMERIC_CHARS[digitBits]);
196 private static String guessEncoding(byte[] bytes) {
197 if (ASSUME_SHIFT_JIS) {
200 // Does it start with the UTF-8 byte order mark? then guess it's UTF-8
201 if (bytes.length > 3 && bytes[0] == (byte) 0xEF && bytes[1] == (byte) 0xBB && bytes[2] == (byte) 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 int length = bytes.length;
212 boolean canBeISO88591 = true;
213 boolean lastWasPossibleDoubleByteStart = false;
214 for (int i = 0; i < length; i++) {
215 int value = bytes[i] & 0xFF;
216 if (value >= 0x80 && value <= 0x9F && i < length - 1) {
217 canBeISO88591 = false;
218 // ISO-8859-1 shouldn't use this, but before we decide it is Shift_JIS,
219 // just double check that it is followed by a byte that's valid in
220 // the Shift_JIS encoding
221 if (lastWasPossibleDoubleByteStart) {
222 // If we just checked this and the last byte for being a valid double-byte
223 // char, don't check starting on this byte. If the this and the last byte
224 // formed a valid pair, then this shouldn't be checked to see if it starts
225 // a double byte pair of course.
226 lastWasPossibleDoubleByteStart = false;
228 // ... otherwise do check to see if this plus the next byte form a valid
229 // double byte pair encoding a character.
230 lastWasPossibleDoubleByteStart = true;
231 int nextValue = bytes[i + 1] & 0xFF;
232 if ((value & 0x1) == 0) {
233 // if even, next value should be in [0x9F,0xFC]
234 // if not, we'll guess UTF-8
235 if (nextValue < 0x9F || nextValue > 0xFC) {
239 // if odd, next value should be in [0x40,0x9E]
240 // if not, we'll guess UTF-8
241 if (nextValue < 0x40 || nextValue > 0x9E) {
248 return canBeISO88591 ? ISO88591 : SHIFT_JIS;