Decode Base64 data in Java

Question

I have an image that is Base64 encoded. What is the best way to decode that in Java? Hopefully using only the libraries included with Sun Java 6.

Answer 1

Given a test encode/decode example of javax.xml.bind.DatatypeConverter using methods parseBase64Binary() and printBase64Binary() referring to @jeremy-ross and @nightfirecat answer.

@Test
public void EncodeDecode() {
    //ENCODE
    String hello = "Hello World";
    byte[] helloBytes = hello.getBytes(StandardCharsets.UTF_8);
    String encodedHello = DatatypeConverter.printBase64Binary(helloBytes);
    LOGGER.info(hello + " encoded=> " + encodedHello);

    //DECODE
    byte[] encodedHelloBytes = DatatypeConverter.parseBase64Binary(encodedHello);
    String helloAgain = new String(encodedHelloBytes, StandardCharsets.UTF_8) ;
    LOGGER.info(encodedHello + " decoded=> " + helloAgain);

    Assert.assertEquals(hello, helloAgain);
}

Result:

INFO - Hello World encoded=> SGVsbG8gV29ybGQ=
INFO - SGVsbG8gV29ybGQ= decoded=> Hello World

Answer 2

This is a late answer, but Joshua Bloch committed his Base64 class (when he was working for Sun, ahem, Oracle) under the java.util.prefs package. This class existed since JDK 1.4.

E.g.

String currentString = "Hello World";
String base64String = java.util.prefs.Base64.byteArrayToBase64(currentString.getBytes("UTF-8"));

Answer 3

In a code compiled with Java 7 but potentially running in a higher java version, it seems useful to detect presence of java.util.Base64 class and use the approach best for given JVM mentioned in other questions here.

I used this code:

private static final Method JAVA_UTIL_BASE64_GETENCODER;

static {
    Method getEncoderMethod;
    try {
        final Class<?> base64Class = Class.forName("java.util.Base64");
        getEncoderMethod = base64Class.getMethod("getEncoder");
    } catch (ClassNotFoundException | NoSuchMethodException e) {
        getEncoderMethod = null;
    }
    JAVA_UTIL_BASE64_GETENCODER = getEncoderMethod;
}

static String base64EncodeToString(String s) {
    final byte[] bytes = s.getBytes(StandardCharsets.ISO_8859_1);
    if (JAVA_UTIL_BASE64_GETENCODER == null) {
        // Java 7 and older // TODO: remove this branch after switching to Java 8
        return DatatypeConverter.printBase64Binary(bytes);
    } else {
        // Java 8 and newer
        try {
            final Object encoder = JAVA_UTIL_BASE64_GETENCODER.invoke(null);
            final Class<?> encoderClass = encoder.getClass();
            final Method encodeMethod = encoderClass.getMethod("encode", byte[].class);
            final byte[] encodedBytes = (byte[]) encodeMethod.invoke(encoder, bytes);
            return new String(encodedBytes);
        } catch (NoSuchMethodException | IllegalAccessException | InvocationTargetException e) {
            throw new IllegalStateException(e);
        }
    }
}

Answer 4

No need to use commons--Sun ships a base64 encoder with Java. You can import it as such:

import sun.misc.BASE64Decoder;

And then use it like this:

BASE64Decoder decoder = new BASE64Decoder();
byte[] decodedBytes = decoder.decodeBuffer(encodedBytes);

Where encodedBytes is either a java.lang.String or a java.io.InputStream. Just beware that the sun.* classes are not "officially supported" by Sun.

EDIT: Who knew this would be the most controversial answer I'd ever post? I do know that sun.* packages are not supported or guaranteed to continue existing, and I do know about Commons and use it all the time. However, the poster asked for a class that that was "included with Sun Java 6," and that's what I was trying to answer. I agree that Commons is the best way to go in general.

EDIT 2: As amir75 points out below, Java 6+ ships with JAXB, which contains supported code to encode/decode Base64. Please see Jeremy Ross' answer below.

Answer 5

If you are already adding AWS SDK for Java, you can use com.amazonaws.util.Base64.

Answer 6

I used android.util.base64 that works pretty good without any dependances:

Usage:

byte[] decodedKey = Base64.decode(encodedPublicKey, Base64.DEFAULT);

---

package com.test;

import java.io.UnsupportedEncodingException;

/**
 * Utilities for encoding and decoding the Base64 representation of
 * binary data.  See RFCs <a
 * href="http://www.ietf.org/rfc/rfc2045.txt">2045</a> and <a
 * href="http://www.ietf.org/rfc/rfc3548.txt">3548</a>.
 */
public class Base64 {

    public static final int DEFAULT = 0;


    public static final int NO_PADDING = 1;


    public static final int NO_WRAP = 2;


    public static final int CRLF = 4;


    public static final int URL_SAFE = 8;


    public static final int NO_CLOSE = 16;

    //  --------------------------------------------------------
    //  shared code
    //  --------------------------------------------------------

    /* package */ static abstract class Coder {
        public byte[] output;
        public int op;

        public abstract boolean process(byte[] input, int offset, int len, boolean finish);


        public abstract int maxOutputSize(int len);
    }

    //  --------------------------------------------------------
    //  decoding
    //  --------------------------------------------------------


    public static byte[] decode(String str, int flags) {
        return decode(str.getBytes(), flags);
    }

    public static byte[] decode(byte[] input, int flags) {
        return decode(input, 0, input.length, flags);
    }


    public static byte[] decode(byte[] input, int offset, int len, int flags) {
        // Allocate space for the most data the input could represent.
        // (It could contain less if it contains whitespace, etc.)
        Decoder decoder = new Decoder(flags, new byte[len*3/4]);

        if (!decoder.process(input, offset, len, true)) {
            throw new IllegalArgumentException("bad base-64");
        }

        // Maybe we got lucky and allocated exactly enough output space.
        if (decoder.op == decoder.output.length) {
            return decoder.output;
        }

        // Need to shorten the array, so allocate a new one of the
        // right size and copy.
        byte[] temp = new byte[decoder.op];
        System.arraycopy(decoder.output, 0, temp, 0, decoder.op);
        return temp;
    }

   static class Decoder extends Coder {       
        private static final int DECODE[] = {
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 62, -1, -1, -1, 63,
            52, 53, 54, 55, 56, 57, 58, 59, 60, 61, -1, -1, -1, -2, -1, -1,
            -1,  0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14,
            15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, -1, -1, -1, -1, -1,
            -1, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40,
            41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, -1, -1, -1, -1, -1,
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
        };

        /**
         * Decode lookup table for the "web safe" variant (RFC 3548
         * sec. 4) where - and _ replace + and /.
         */
        private static final int DECODE_WEBSAFE[] = {
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 62, -1, -1,
            52, 53, 54, 55, 56, 57, 58, 59, 60, 61, -1, -1, -1, -2, -1, -1,
            -1,  0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14,
            15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, -1, -1, -1, -1, 63,
            -1, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40,
            41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, -1, -1, -1, -1, -1,
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
            -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
        };

        /** Non-data values in the DECODE arrays. */
        private static final int SKIP = -1;
        private static final int EQUALS = -2;


        private int state;   // state number (0 to 6)
        private int value;

        final private int[] alphabet;

        public Decoder(int flags, byte[] output) {
            this.output = output;

            alphabet = ((flags & URL_SAFE) == 0) ? DECODE : DECODE_WEBSAFE;
            state = 0;
            value = 0;
        }


        public int maxOutputSize(int len) {
            return len * 3/4 + 10;
        }

        /**
         * Decode another block of input data.
         *
         * @return true if the state machine is still healthy.  false if
         *         bad base-64 data has been detected in the input stream.
         */
        public boolean process(byte[] input, int offset, int len, boolean finish) {
            if (this.state == 6) return false;

            int p = offset;
            len += offset;

            int state = this.state;
            int value = this.value;
            int op = 0;
            final byte[] output = this.output;
            final int[] alphabet = this.alphabet;

            while (p < len) {                
                if (state == 0) {
                    while (p+4 <= len &&
                           (value = ((alphabet[input[p] & 0xff] << 18) |
                                     (alphabet[input[p+1] & 0xff] << 12) |
                                     (alphabet[input[p+2] & 0xff] << 6) |
                                     (alphabet[input[p+3] & 0xff]))) >= 0) {
                        output[op+2] = (byte) value;
                        output[op+1] = (byte) (value >> 8);
                        output[op] = (byte) (value >> 16);
                        op += 3;
                        p += 4;
                    }
                    if (p >= len) break;
                }

                int d = alphabet[input[p++] & 0xff];

                switch (state) {
                case 0:
                    if (d >= 0) {
                        value = d;
                        ++state;
                    } else if (d != SKIP) {
                        this.state = 6;
                        return false;
                    }
                    break;

                case 1:
                    if (d >= 0) {
                        value = (value << 6) | d;
                        ++state;
                    } else if (d != SKIP) {
                        this.state = 6;
                        return false;
                    }
                    break;

                case 2:
                    if (d >= 0) {
                        value = (value << 6) | d;
                        ++state;
                    } else if (d == EQUALS) {
                        // Emit the last (partial) output tuple;
                        // expect exactly one more padding character.
                        output[op++] = (byte) (value >> 4);
                        state = 4;
                    } else if (d != SKIP) {
                        this.state = 6;
                        return false;
                    }
                    break;

                case 3:
                    if (d >= 0) {
                        // Emit the output triple and return to state 0.
                        value = (value << 6) | d;
                        output[op+2] = (byte) value;
                        output[op+1] = (byte) (value >> 8);
                        output[op] = (byte) (value >> 16);
                        op += 3;
                        state = 0;
                    } else if (d == EQUALS) {
                        // Emit the last (partial) output tuple;
                        // expect no further data or padding characters.
                        output[op+1] = (byte) (value >> 2);
                        output[op] = (byte) (value >> 10);
                        op += 2;
                        state = 5;
                    } else if (d != SKIP) {
                        this.state = 6;
                        return false;
                    }
                    break;

                case 4:
                    if (d == EQUALS) {
                        ++state;
                    } else if (d != SKIP) {
                        this.state = 6;
                        return false;
                    }
                    break;

                case 5:
                    if (d != SKIP) {
                        this.state = 6;
                        return false;
                    }
                    break;
                }
            }

            if (!finish) {
                // We're out of input, but a future call could provide
                // more.
                this.state = state;
                this.value = value;
                this.op = op;
                return true;
            }


            switch (state) {
            case 0:
                break;
            case 1:
                this.state = 6;
                return false;
            case 2:
                output[op++] = (byte) (value >> 4);
                break;
            case 3:
               output[op++] = (byte) (value >> 10);
                output[op++] = (byte) (value >> 2);
                break;
            case 4:
                this.state = 6;
                return false;
            case 5:
                break;
            }

            this.state = state;
            this.op = op;
            return true;
        }
    }

    //  --------------------------------------------------------
    //  encoding
    //  --------------------------------------------------------    
    public static String encodeToString(byte[] input, int flags) {
        try {
            return new String(encode(input, flags), "US-ASCII");
        } catch (UnsupportedEncodingException e) {
            // US-ASCII is guaranteed to be available.
            throw new AssertionError(e);
        }
    }


    public static String encodeToString(byte[] input, int offset, int len, int flags) {
        try {
            return new String(encode(input, offset, len, flags), "US-ASCII");
        } catch (UnsupportedEncodingException e) {
            // US-ASCII is guaranteed to be available.
            throw new AssertionError(e);
        }
    }


    public static byte[] encode(byte[] input, int flags) {
        return encode(input, 0, input.length, flags);
    }


    public static byte[] encode(byte[] input, int offset, int len, int flags) {
        Encoder encoder = new Encoder(flags, null);

        // Compute the exact length of the array we will produce.
        int output_len = len / 3 * 4;

        // Account for the tail of the data and the padding bytes, if any.
        if (encoder.do_padding) {
            if (len % 3 > 0) {
                output_len += 4;
            }
        } else {
            switch (len % 3) {
                case 0: break;
                case 1: output_len += 2; break;
                case 2: output_len += 3; break;
            }
        }

        // Account for the newlines, if any.
        if (encoder.do_newline && len > 0) {
            output_len += (((len-1) / (3 * Encoder.LINE_GROUPS)) + 1) *
                (encoder.do_cr ? 2 : 1);
        }

        encoder.output = new byte[output_len];
        encoder.process(input, offset, len, true);

        assert encoder.op == output_len;

        return encoder.output;
    }

    /* package */ static class Encoder extends Coder {
        /**
         * Emit a new line every this many output tuples.  Corresponds to
         * a 76-character line length (the maximum allowable according to
         * <a href="http://www.ietf.org/rfc/rfc2045.txt">RFC 2045</a>).
         */
        public static final int LINE_GROUPS = 19;

        /**
         * Lookup table for turning Base64 alphabet positions (6 bits)
         * into output bytes.
         */
        private static final byte ENCODE[] = {
            'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O', 'P',
            'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z', 'a', 'b', 'c', 'd', 'e', 'f',
            'g', 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', 'p', 'q', 'r', 's', 't', 'u', 'v',
            'w', 'x', 'y', 'z', '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', '+', '/',
        };

        /**
         * Lookup table for turning Base64 alphabet positions (6 bits)
         * into output bytes.
         */
        private static final byte ENCODE_WEBSAFE[] = {
            'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O', 'P',
            'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z', 'a', 'b', 'c', 'd', 'e', 'f',
            'g', 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', 'p', 'q', 'r', 's', 't', 'u', 'v',
            'w', 'x', 'y', 'z', '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', '-', '_',
        };

        final private byte[] tail;
        /* package */ int tailLen;
        private int count;

        final public boolean do_padding;
        final public boolean do_newline;
        final public boolean do_cr;
        final private byte[] alphabet;

        public Encoder(int flags, byte[] output) {
            this.output = output;

            do_padding = (flags & NO_PADDING) == 0;
            do_newline = (flags & NO_WRAP) == 0;
            do_cr = (flags & CRLF) != 0;
            alphabet = ((flags & URL_SAFE) == 0) ? ENCODE : ENCODE_WEBSAFE;

            tail = new byte[2];
            tailLen = 0;

            count = do_newline ? LINE_GROUPS : -1;
        }

        /**
         * @return an overestimate for the number of bytes {@code
         * len} bytes could encode to.
         */
        public int maxOutputSize(int len) {
            return len * 8/5 + 10;
        }

        public boolean process(byte[] input, int offset, int len, boolean finish) {
            // Using local variables makes the encoder about 9% faster.
            final byte[] alphabet = this.alphabet;
            final byte[] output = this.output;
            int op = 0;
            int count = this.count;

            int p = offset;
            len += offset;
            int v = -1;

            // First we need to concatenate the tail of the previous call
            // with any input bytes available now and see if we can empty
            // the tail.

            switch (tailLen) {
                case 0:
                    // There was no tail.
                    break;

                case 1:
                    if (p+2 <= len) {
                        // A 1-byte tail with at least 2 bytes of
                        // input available now.
                        v = ((tail[0] & 0xff) << 16) |
                            ((input[p++] & 0xff) << 8) |
                            (input[p++] & 0xff);
                        tailLen = 0;
                    };
                    break;

                case 2:
                    if (p+1 <= len) {
                        // A 2-byte tail with at least 1 byte of input.
                        v = ((tail[0] & 0xff) << 16) |
                            ((tail[1] & 0xff) << 8) |
                            (input[p++] & 0xff);
                        tailLen = 0;
                    }
                    break;
            }

            if (v != -1) {
                output[op++] = alphabet[(v >> 18) & 0x3f];
                output[op++] = alphabet[(v >> 12) & 0x3f];
                output[op++] = alphabet[(v >> 6) & 0x3f];
                output[op++] = alphabet[v & 0x3f];
                if (--count == 0) {
                    if (do_cr) output[op++] = '\r';
                    output[op++] = '\n';
                    count = LINE_GROUPS;
                }
            }

            // At this point either there is no tail, or there are fewer
            // than 3 bytes of input available.

            // The main loop, turning 3 input bytes into 4 output bytes on
            // each iteration.
            while (p+3 <= len) {
                v = ((input[p] & 0xff) << 16) |
                    ((input[p+1] & 0xff) << 8) |
                    (input[p+2] & 0xff);
                output[op] = alphabet[(v >> 18) & 0x3f];
                output[op+1] = alphabet[(v >> 12) & 0x3f];
                output[op+2] = alphabet[(v >> 6) & 0x3f];
                output[op+3] = alphabet[v & 0x3f];
                p += 3;
                op += 4;
                if (--count == 0) {
                    if (do_cr) output[op++] = '\r';
                    output[op++] = '\n';
                    count = LINE_GROUPS;
                }
            }

            if (finish) {

                if (p-tailLen == len-1) {
                    int t = 0;
                    v = ((tailLen > 0 ? tail[t++] : input[p++]) & 0xff) << 4;
                    tailLen -= t;
                    output[op++] = alphabet[(v >> 6) & 0x3f];
                    output[op++] = alphabet[v & 0x3f];
                    if (do_padding) {
                        output[op++] = '=';
                        output[op++] = '=';
                    }
                    if (do_newline) {
                        if (do_cr) output[op++] = '\r';
                        output[op++] = '\n';
                    }
                } else if (p-tailLen == len-2) {
                    int t = 0;
                    v = (((tailLen > 1 ? tail[t++] : input[p++]) & 0xff) << 10) |
                        (((tailLen > 0 ? tail[t++] : input[p++]) & 0xff) << 2);
                    tailLen -= t;
                    output[op++] = alphabet[(v >> 12) & 0x3f];
                    output[op++] = alphabet[(v >> 6) & 0x3f];
                    output[op++] = alphabet[v & 0x3f];
                    if (do_padding) {
                        output[op++] = '=';
                    }
                    if (do_newline) {
                        if (do_cr) output[op++] = '\r';
                        output[op++] = '\n';
                    }
                } else if (do_newline && op > 0 && count != LINE_GROUPS) {
                    if (do_cr) output[op++] = '\r';
                    output[op++] = '\n';
                }

                assert tailLen == 0;
                assert p == len;
            } else {
                // Save the leftovers in tail to be consumed on the next
                // call to encodeInternal.

                if (p == len-1) {
                    tail[tailLen++] = input[p];
                } else if (p == len-2) {
                    tail[tailLen++] = input[p];
                    tail[tailLen++] = input[p+1];
                }
            }

            this.op = op;
            this.count = count;

            return true;
        }
    }

    private Base64() { }   // don't instantiate
}