2 * Copyright 2007 ZXing authors
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 = "SJIS";
44 private static final String EUC_JP = "EUC_JP";
45 private static final boolean ASSUME_SHIFT_JIS;
46 private static final String UTF8 = "UTF8";
47 private static final String ISO88591 = "ISO8859_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();
60 CharacterSetECI currentCharacterSetECI = null;
63 // While still another segment to read...
64 if (bits.available() == 0) {
65 // OK, assume we're done. Really, a TERMINATOR mode should have been recorded here
66 mode = Mode.TERMINATOR;
68 mode = Mode.forBits(bits.readBits(4)); // mode is encoded by 4 bits
70 if (!mode.equals(Mode.TERMINATOR)) {
71 if (mode.equals(Mode.ECI)) {
72 // Count doesn't apply to ECI
73 int value = ECI.parseECI(bits);
75 currentCharacterSetECI = CharacterSetECI.getCharacterSetECIByValue(value);
76 } catch (IllegalArgumentException iae) {
77 // unsupported... just continue?
80 // How many characters will follow, encoded in this mode?
81 int count = bits.readBits(mode.getCharacterCountBits(version));
82 if (mode.equals(Mode.NUMERIC)) {
83 decodeNumericSegment(bits, result, count);
84 } else if (mode.equals(Mode.ALPHANUMERIC)) {
85 decodeAlphanumericSegment(bits, result, count);
86 } else if (mode.equals(Mode.BYTE)) {
87 decodeByteSegment(bits, result, count, currentCharacterSetECI);
88 } else if (mode.equals(Mode.KANJI)) {
89 decodeKanjiSegment(bits, result, count);
91 throw new ReaderException("Unsupported mode indicator");
95 } while (!mode.equals(Mode.TERMINATOR));
97 // I thought it wasn't allowed to leave extra bytes after the terminator but it happens
99 int bitsLeft = bits.available();
101 if (bitsLeft > 6 || bits.readBits(bitsLeft) != 0) {
102 throw new ReaderException("Excess bits or non-zero bits after terminator mode indicator");
106 return result.toString();
109 private static void decodeKanjiSegment(BitSource bits,
111 int count) throws ReaderException {
112 // Each character will require 2 bytes. Read the characters as 2-byte pairs
113 // and decode as Shift_JIS afterwards
114 byte[] buffer = new byte[2 * count];
117 // Each 13 bits encodes a 2-byte character
118 int twoBytes = bits.readBits(13);
119 int assembledTwoBytes = ((twoBytes / 0x0C0) << 8) | (twoBytes % 0x0C0);
120 if (assembledTwoBytes < 0x01F00) {
121 // In the 0x8140 to 0x9FFC range
122 assembledTwoBytes += 0x08140;
124 // In the 0xE040 to 0xEBBF range
125 assembledTwoBytes += 0x0C140;
127 buffer[offset] = (byte) (assembledTwoBytes >> 8);
128 buffer[offset + 1] = (byte) assembledTwoBytes;
132 // Shift_JIS may not be supported in some environments:
134 result.append(new String(buffer, SHIFT_JIS));
135 } catch (UnsupportedEncodingException uee) {
136 throw new ReaderException(SHIFT_JIS + " encoding is not supported on this device");
140 private static void decodeByteSegment(BitSource bits,
143 CharacterSetECI currentCharacterSetECI) throws ReaderException {
144 byte[] readBytes = new byte[count];
145 if (count << 3 > bits.available()) {
146 throw new ReaderException("Count too large: " + count);
148 for (int i = 0; i < count; i++) {
149 readBytes[i] = (byte) bits.readBits(8);
152 if (currentCharacterSetECI == null) {
153 // The spec isn't clear on this mode; see
154 // section 6.4.5: t does not say which encoding to assuming
155 // upon decoding. I have seen ISO-8859-1 used as well as
156 // Shift_JIS -- without anything like an ECI designator to
158 encoding = guessEncoding(readBytes);
160 encoding = currentCharacterSetECI.getEncodingName();
163 result.append(new String(readBytes, encoding));
164 } catch (UnsupportedEncodingException uce) {
165 throw new ReaderException(uce.toString());
169 private static void decodeAlphanumericSegment(BitSource bits,
172 // Read two characters at a time
174 int nextTwoCharsBits = bits.readBits(11);
175 result.append(ALPHANUMERIC_CHARS[nextTwoCharsBits / 45]);
176 result.append(ALPHANUMERIC_CHARS[nextTwoCharsBits % 45]);
180 // special case: one character left
181 result.append(ALPHANUMERIC_CHARS[bits.readBits(6)]);
185 private static void decodeNumericSegment(BitSource bits,
187 int count) throws ReaderException {
188 // Read three digits at a time
190 // Each 10 bits encodes three digits
191 int threeDigitsBits = bits.readBits(10);
192 if (threeDigitsBits >= 1000) {
193 throw new ReaderException("Illegal value for 3-digit unit: " + threeDigitsBits);
195 result.append(ALPHANUMERIC_CHARS[threeDigitsBits / 100]);
196 result.append(ALPHANUMERIC_CHARS[(threeDigitsBits / 10) % 10]);
197 result.append(ALPHANUMERIC_CHARS[threeDigitsBits % 10]);
201 // Two digits left over to read, encoded in 7 bits
202 int twoDigitsBits = bits.readBits(7);
203 if (twoDigitsBits >= 100) {
204 throw new ReaderException("Illegal value for 2-digit unit: " + twoDigitsBits);
206 result.append(ALPHANUMERIC_CHARS[twoDigitsBits / 10]);
207 result.append(ALPHANUMERIC_CHARS[twoDigitsBits % 10]);
208 } else if (count == 1) {
209 // One digit left over to read
210 int digitBits = bits.readBits(4);
211 if (digitBits >= 10) {
212 throw new ReaderException("Illegal value for digit unit: " + digitBits);
214 result.append(ALPHANUMERIC_CHARS[digitBits]);
218 private static String guessEncoding(byte[] bytes) {
219 if (ASSUME_SHIFT_JIS) {
222 // Does it start with the UTF-8 byte order mark? then guess it's UTF-8
223 if (bytes.length > 3 && bytes[0] == (byte) 0xEF && bytes[1] == (byte) 0xBB && bytes[2] == (byte) 0xBF) {
226 // For now, merely tries to distinguish ISO-8859-1, UTF-8 and Shift_JIS,
227 // which should be by far the most common encodings. ISO-8859-1
228 // should not have bytes in the 0x80 - 0x9F range, while Shift_JIS
229 // uses this as a first byte of a two-byte character. If we see this
230 // followed by a valid second byte in Shift_JIS, assume it is Shift_JIS.
231 // If we see something else in that second byte, we'll make the risky guess
233 int length = bytes.length;
234 boolean canBeISO88591 = true;
235 boolean lastWasPossibleDoubleByteStart = false;
236 for (int i = 0; i < length; i++) {
237 int value = bytes[i] & 0xFF;
238 if (value >= 0x80 && value <= 0x9F && i < length - 1) {
239 canBeISO88591 = false;
240 // ISO-8859-1 shouldn't use this, but before we decide it is Shift_JIS,
241 // just double check that it is followed by a byte that's valid in
242 // the Shift_JIS encoding
243 if (lastWasPossibleDoubleByteStart) {
244 // If we just checked this and the last byte for being a valid double-byte
245 // char, don't check starting on this byte. If the this and the last byte
246 // formed a valid pair, then this shouldn't be checked to see if it starts
247 // a double byte pair of course.
248 lastWasPossibleDoubleByteStart = false;
250 // ... otherwise do check to see if this plus the next byte form a valid
251 // double byte pair encoding a character.
252 lastWasPossibleDoubleByteStart = true;
253 int nextValue = bytes[i + 1] & 0xFF;
254 if ((value & 0x1) == 0) {
255 // if even, next value should be in [0x9F,0xFC]
256 // if not, we'll guess UTF-8
257 if (nextValue < 0x9F || nextValue > 0xFC) {
261 // if odd, next value should be in [0x40,0x9E]
262 // if not, we'll guess UTF-8
263 if (nextValue < 0x40 || nextValue > 0x9E) {
270 return canBeISO88591 ? ISO88591 : SHIFT_JIS;