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;
21 import java.io.UnsupportedEncodingException;
24 * <p>QR Codes can encode text as bits in one of several modes, and can use multiple modes
25 * in one QR Code. This class decodes the bits back into text.</p>
27 * <p>See ISO 18004:2006, 6.4.3 - 6.4.7</p>
29 * @author srowen@google.com (Sean Owen)
31 final class DecodedBitStreamParser {
34 * See ISO 18004:2006, 6.4.4 Table 5
36 private static final char[] ALPHANUMERIC_CHARS = new char[]{
37 '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', 'A', 'B',
38 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J', 'K', 'L', 'M', 'N',
39 'O', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z',
40 ' ', '$', '%', '*', '+', '-', '.', '/', ':'
42 private static final String SHIFT_JIS = "Shift_JIS";
43 private static final boolean ASSUME_SHIFT_JIS;
44 private static final String UTF8 = "UTF-8";
45 private static final String ISO88591 = "ISO-8859-1";
48 String platformDefault = System.getProperty("file.encoding");
49 ASSUME_SHIFT_JIS = SHIFT_JIS.equalsIgnoreCase(platformDefault) ||
50 "EUC-JP".equalsIgnoreCase(platformDefault);
53 private DecodedBitStreamParser() {
56 static String decode(byte[] bytes, Version version) throws ReaderException {
57 BitSource bits = new BitSource(bytes);
58 StringBuffer result = new StringBuffer();
61 // While still another segment to read...
62 mode = Mode.forBits(bits.readBits(4)); // mode is encoded by 4 bits
63 if (!mode.equals(Mode.TERMINATOR)) {
64 // How many characters will follow, encoded in this mode?
65 int count = bits.readBits(mode.getCharacterCountBits(version));
66 if (mode.equals(Mode.NUMERIC)) {
67 decodeNumericSegment(bits, result, count);
68 } else if (mode.equals(Mode.ALPHANUMERIC)) {
69 decodeAlphanumericSegment(bits, result, count);
70 } else if (mode.equals(Mode.BYTE)) {
71 decodeByteSegment(bits, result, count);
72 } else if (mode.equals(Mode.KANJI)) {
73 decodeKanjiSegment(bits, result, count);
75 throw new ReaderException("Unsupported mode indicator");
78 } while (!mode.equals(Mode.TERMINATOR));
80 // I thought it wasn't allowed to leave extra bytes after the terminator but it happens
82 int bitsLeft = bits.available();
84 if (bitsLeft > 6 || bits.readBits(bitsLeft) != 0) {
85 throw new ReaderException("Excess bits or non-zero bits after terminator mode indicator");
89 return result.toString();
92 private static void decodeKanjiSegment(BitSource bits,
94 int count) throws ReaderException {
95 // Each character will require 2 bytes. Read the characters as 2-byte pairs
96 // and decode as Shift_JIS afterwards
97 byte[] buffer = new byte[2 * count];
100 // Each 13 bits encodes a 2-byte character
101 int twoBytes = bits.readBits(13);
102 int assembledTwoBytes = ((twoBytes / 0x0C0) << 8) | (twoBytes % 0x0C0);
103 if (assembledTwoBytes < 0x01F00) {
104 // In the 0x8140 to 0x9FFC range
105 assembledTwoBytes += 0x08140;
107 // In the 0xE040 to 0xEBBF range
108 assembledTwoBytes += 0x0C140;
110 buffer[offset] = (byte) (assembledTwoBytes >> 8);
111 buffer[offset + 1] = (byte) assembledTwoBytes;
115 // Shift_JIS may not be supported in some environments:
117 result.append(new String(buffer, SHIFT_JIS));
118 } catch (UnsupportedEncodingException uee) {
119 throw new ReaderException(SHIFT_JIS + " encoding is not supported on this device");
123 private static void decodeByteSegment(BitSource bits,
125 int count) throws ReaderException {
126 byte[] readBytes = new byte[count];
127 if (count << 3 > bits.available()) {
128 throw new ReaderException("Count too large: " + count);
130 for (int i = 0; i < count; i++) {
131 readBytes[i] = (byte) bits.readBits(8);
133 // The spec isn't clear on this mode; see
134 // section 6.4.5: t does not say which encoding to assuming
135 // upon decoding. I have seen ISO-8859-1 used as well as
136 // Shift_JIS -- without anything like an ECI designator to
138 String encoding = guessEncoding(readBytes);
140 result.append(new String(readBytes, encoding));
141 } catch (UnsupportedEncodingException uce) {
142 throw new ReaderException(uce.toString());
146 private static void decodeAlphanumericSegment(BitSource bits,
149 // Read two characters at a time
151 int nextTwoCharsBits = bits.readBits(11);
152 result.append(ALPHANUMERIC_CHARS[nextTwoCharsBits / 45]);
153 result.append(ALPHANUMERIC_CHARS[nextTwoCharsBits % 45]);
157 // special case: one character left
158 result.append(ALPHANUMERIC_CHARS[bits.readBits(6)]);
162 private static void decodeNumericSegment(BitSource bits,
164 int count) throws ReaderException {
165 // Read three digits at a time
167 // Each 10 bits encodes three digits
168 int threeDigitsBits = bits.readBits(10);
169 if (threeDigitsBits >= 1000) {
170 throw new ReaderException("Illegal value for 3-digit unit: " + threeDigitsBits);
172 result.append(ALPHANUMERIC_CHARS[threeDigitsBits / 100]);
173 result.append(ALPHANUMERIC_CHARS[(threeDigitsBits / 10) % 10]);
174 result.append(ALPHANUMERIC_CHARS[threeDigitsBits % 10]);
178 // Two digits left over to read, encoded in 7 bits
179 int twoDigitsBits = bits.readBits(7);
180 if (twoDigitsBits >= 100) {
181 throw new ReaderException("Illegal value for 2-digit unit: " + twoDigitsBits);
183 result.append(ALPHANUMERIC_CHARS[twoDigitsBits / 10]);
184 result.append(ALPHANUMERIC_CHARS[twoDigitsBits % 10]);
185 } else if (count == 1) {
186 // One digit left over to read
187 int digitBits = bits.readBits(4);
188 if (digitBits >= 10) {
189 throw new ReaderException("Illegal value for digit unit: " + digitBits);
191 result.append(ALPHANUMERIC_CHARS[digitBits]);
195 private static String guessEncoding(byte[] bytes) {
196 if (ASSUME_SHIFT_JIS) {
199 // Does it start with the UTF-8 byte order mark? then guess it's UTF-8
200 if (bytes.length > 3 && bytes[0] == (byte) 0xEF && bytes[1] == (byte) 0xBB && bytes[2] == (byte) 0xBF) {
203 // For now, merely tries to distinguish ISO-8859-1, UTF-8 and Shift_JIS,
204 // which should be by far the most common encodings. ISO-8859-1
205 // should not have bytes in the 0x80 - 0x9F range, while Shift_JIS
206 // uses this as a first byte of a two-byte character. If we see this
207 // followed by a valid second byte in Shift_JIS, assume it is Shift_JIS.
208 // If we see something else in that second byte, we'll make the risky guess
210 int length = bytes.length;
211 boolean canBeISO88591 = true;
212 boolean 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;