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;
61 boolean fc1InEffect = false;
64 // While still another segment to read...
65 if (bits.available() == 0) {
66 // OK, assume we're done. Really, a TERMINATOR mode should have been recorded here
67 mode = Mode.TERMINATOR;
69 mode = Mode.forBits(bits.readBits(4)); // mode is encoded by 4 bits
71 if (!mode.equals(Mode.TERMINATOR)) {
72 if (mode.equals(Mode.FNC1_FIRST_POSITION) || mode.equals(Mode.FNC1_SECOND_POSITION)) {
73 // We do little with FNC1 except alter the parsed result a bit according to the spec
75 } else if (mode.equals(Mode.ECI)) {
76 // Count doesn't apply to ECI
77 int value = ECI.parseECI(bits);
79 currentCharacterSetECI = CharacterSetECI.getCharacterSetECIByValue(value);
80 } catch (IllegalArgumentException iae) {
81 // unsupported... just continue?
84 // How many characters will follow, encoded in this mode?
85 int count = bits.readBits(mode.getCharacterCountBits(version));
86 if (mode.equals(Mode.NUMERIC)) {
87 decodeNumericSegment(bits, result, count);
88 } else if (mode.equals(Mode.ALPHANUMERIC)) {
89 decodeAlphanumericSegment(bits, result, count, fc1InEffect);
90 } else if (mode.equals(Mode.BYTE)) {
91 decodeByteSegment(bits, result, count, currentCharacterSetECI);
92 } else if (mode.equals(Mode.KANJI)) {
93 decodeKanjiSegment(bits, result, count);
95 throw new ReaderException("Unsupported mode indicator");
99 } while (!mode.equals(Mode.TERMINATOR));
101 // I thought it wasn't allowed to leave extra bytes after the terminator but it happens
103 int bitsLeft = bits.available();
105 if (bitsLeft > 6 || bits.readBits(bitsLeft) != 0) {
106 throw new ReaderException("Excess bits or non-zero bits after terminator mode indicator");
110 return result.toString();
113 private static void decodeKanjiSegment(BitSource bits,
115 int count) throws ReaderException {
116 // Each character will require 2 bytes. Read the characters as 2-byte pairs
117 // and decode as Shift_JIS afterwards
118 byte[] buffer = new byte[2 * count];
121 // Each 13 bits encodes a 2-byte character
122 int twoBytes = bits.readBits(13);
123 int assembledTwoBytes = ((twoBytes / 0x0C0) << 8) | (twoBytes % 0x0C0);
124 if (assembledTwoBytes < 0x01F00) {
125 // In the 0x8140 to 0x9FFC range
126 assembledTwoBytes += 0x08140;
128 // In the 0xE040 to 0xEBBF range
129 assembledTwoBytes += 0x0C140;
131 buffer[offset] = (byte) (assembledTwoBytes >> 8);
132 buffer[offset + 1] = (byte) assembledTwoBytes;
136 // Shift_JIS may not be supported in some environments:
138 result.append(new String(buffer, SHIFT_JIS));
139 } catch (UnsupportedEncodingException uee) {
140 throw new ReaderException(SHIFT_JIS + " encoding is not supported on this device");
144 private static void decodeByteSegment(BitSource bits,
147 CharacterSetECI currentCharacterSetECI) throws ReaderException {
148 byte[] readBytes = new byte[count];
149 if (count << 3 > bits.available()) {
150 throw new ReaderException("Count too large: " + count);
152 for (int i = 0; i < count; i++) {
153 readBytes[i] = (byte) bits.readBits(8);
156 if (currentCharacterSetECI == null) {
157 // The spec isn't clear on this mode; see
158 // section 6.4.5: t does not say which encoding to assuming
159 // upon decoding. I have seen ISO-8859-1 used as well as
160 // Shift_JIS -- without anything like an ECI designator to
162 encoding = guessEncoding(readBytes);
164 encoding = currentCharacterSetECI.getEncodingName();
167 result.append(new String(readBytes, encoding));
168 } catch (UnsupportedEncodingException uce) {
169 throw new ReaderException(uce.toString());
173 private static void decodeAlphanumericSegment(BitSource bits,
176 boolean fc1InEffect) {
177 // Read two characters at a time
178 int start = result.length();
180 int nextTwoCharsBits = bits.readBits(11);
181 result.append(ALPHANUMERIC_CHARS[nextTwoCharsBits / 45]);
182 result.append(ALPHANUMERIC_CHARS[nextTwoCharsBits % 45]);
186 // special case: one character left
187 result.append(ALPHANUMERIC_CHARS[bits.readBits(6)]);
189 // See section 6.4.8.1, 6.4.8.2
191 // We need to massage the result a bit if in an FNC1 mode:
192 for (int i = start; i < result.length(); i++) {
193 if (result.charAt(i) == '%') {
194 if (i < result.length() - 1 && result.charAt(i + 1) == '%') {
195 // %% is rendered as %
196 result.deleteCharAt(i + 1);
198 // In alpha mode, % should be converted to FNC1 separator 0x1D
199 result.setCharAt(i, (char) 0x1D);
206 private static void decodeNumericSegment(BitSource bits,
208 int count) throws ReaderException {
209 // Read three digits at a time
211 // Each 10 bits encodes three digits
212 int threeDigitsBits = bits.readBits(10);
213 if (threeDigitsBits >= 1000) {
214 throw new ReaderException("Illegal value for 3-digit unit: " + threeDigitsBits);
216 result.append(ALPHANUMERIC_CHARS[threeDigitsBits / 100]);
217 result.append(ALPHANUMERIC_CHARS[(threeDigitsBits / 10) % 10]);
218 result.append(ALPHANUMERIC_CHARS[threeDigitsBits % 10]);
222 // Two digits left over to read, encoded in 7 bits
223 int twoDigitsBits = bits.readBits(7);
224 if (twoDigitsBits >= 100) {
225 throw new ReaderException("Illegal value for 2-digit unit: " + twoDigitsBits);
227 result.append(ALPHANUMERIC_CHARS[twoDigitsBits / 10]);
228 result.append(ALPHANUMERIC_CHARS[twoDigitsBits % 10]);
229 } else if (count == 1) {
230 // One digit left over to read
231 int digitBits = bits.readBits(4);
232 if (digitBits >= 10) {
233 throw new ReaderException("Illegal value for digit unit: " + digitBits);
235 result.append(ALPHANUMERIC_CHARS[digitBits]);
239 private static String guessEncoding(byte[] bytes) {
240 if (ASSUME_SHIFT_JIS) {
243 // Does it start with the UTF-8 byte order mark? then guess it's UTF-8
244 if (bytes.length > 3 && bytes[0] == (byte) 0xEF && bytes[1] == (byte) 0xBB && bytes[2] == (byte) 0xBF) {
247 // For now, merely tries to distinguish ISO-8859-1, UTF-8 and Shift_JIS,
248 // which should be by far the most common encodings. ISO-8859-1
249 // should not have bytes in the 0x80 - 0x9F range, while Shift_JIS
250 // uses this as a first byte of a two-byte character. If we see this
251 // followed by a valid second byte in Shift_JIS, assume it is Shift_JIS.
252 // If we see something else in that second byte, we'll make the risky guess
254 int length = bytes.length;
255 boolean canBeISO88591 = true;
256 boolean lastWasPossibleDoubleByteStart = false;
257 for (int i = 0; i < length; i++) {
258 int value = bytes[i] & 0xFF;
259 if (value >= 0x80 && value <= 0x9F && i < length - 1) {
260 canBeISO88591 = false;
261 // ISO-8859-1 shouldn't use this, but before we decide it is Shift_JIS,
262 // just double check that it is followed by a byte that's valid in
263 // the Shift_JIS encoding
264 if (lastWasPossibleDoubleByteStart) {
265 // If we just checked this and the last byte for being a valid double-byte
266 // char, don't check starting on this byte. If the this and the last byte
267 // formed a valid pair, then this shouldn't be checked to see if it starts
268 // a double byte pair of course.
269 lastWasPossibleDoubleByteStart = false;
271 // ... otherwise do check to see if this plus the next byte form a valid
272 // double byte pair encoding a character.
273 lastWasPossibleDoubleByteStart = true;
274 int nextValue = bytes[i + 1] & 0xFF;
275 if ((value & 0x1) == 0) {
276 // if even, next value should be in [0x9F,0xFC]
277 // if not, we'll guess UTF-8
278 if (nextValue < 0x9F || nextValue > 0xFC) {
282 // if odd, next value should be in [0x40,0x9E]
283 // if not, we'll guess UTF-8
284 if (nextValue < 0x40 || nextValue > 0x9E) {
291 return canBeISO88591 ? ISO88591 : SHIFT_JIS;