GROOVY-7089: Base64 URL Safe encoder (closes groovy#450)

Author: jwagenleitner

src/main/org/codehaus/groovy/runtime/EncodingGroovyMethods.java

@@ -27,6 +27,9 @@
 import java.io.UnsupportedEncodingException;
 import java.io.Writer;
 
+import static org.codehaus.groovy.runtime.EncodingGroovyMethodsSupport.TRANSLATE_TABLE;
+import static org.codehaus.groovy.runtime.EncodingGroovyMethodsSupport.TRANSLATE_TABLE_URLSAFE;
+
 /**
  * This class defines all the encoding/decoding groovy methods which enhance
  * the normal JDK classes when inside the Groovy environment.
@@ -36,6 +39,8 @@ public class EncodingGroovyMethods {
 
     private static final char[] T_TABLE = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/=".toCharArray();
 
+    private static final char[] T_TABLE_URLSAFE = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789-_=".toCharArray();
+
     private static final String CHUNK_SEPARATOR = "\r\n";
 
     /**
@@ -76,18 +81,22 @@ public static Writable encodeBase64(Byte[] data) {
      * @since 1.5.7
      */
     public static Writable encodeBase64(final byte[] data, final boolean chunked) {
+        return encodeBase64(data, chunked, false, true);
+    }
+
+    private static Writable encodeBase64(final byte[] data, final boolean chunked, final boolean urlSafe, final boolean pad) {
         return new Writable() {
             public Writer writeTo(final Writer writer) throws IOException {
                 int charCount = 0;
                 final int dLimit = (data.length / 3) * 3;
-
+                final char[] table = urlSafe ? T_TABLE_URLSAFE : T_TABLE;
                 for (int dIndex = 0; dIndex != dLimit; dIndex += 3) {
                     int d = ((data[dIndex] & 0XFF) << 16) | ((data[dIndex + 1] & 0XFF) << 8) | (data[dIndex + 2] & 0XFF);
 
-                    writer.write(T_TABLE[d >> 18]);
-                    writer.write(T_TABLE[(d >> 12) & 0X3F]);
-                    writer.write(T_TABLE[(d >> 6) & 0X3F]);
-                    writer.write(T_TABLE[d & 0X3F]);
+                    writer.write(table[d >> 18]);
+                    writer.write(table[(d >> 12) & 0X3F]);
+                    writer.write(table[(d >> 6) & 0X3F]);
+                    writer.write(table[d & 0X3F]);
 
                     if (chunked && ++charCount == 19) {
                         writer.write(CHUNK_SEPARATOR);
@@ -102,10 +111,16 @@ public Writer writeTo(final Writer writer) throws IOException {
                         d |= (data[dLimit + 1] & 0XFF) << 8;
                     }
 
-                    writer.write(T_TABLE[d >> 18]);
-                    writer.write(T_TABLE[(d >> 12) & 0X3F]);
-                    writer.write((dLimit + 1 < data.length) ? T_TABLE[(d >> 6) & 0X3F] : '=');
-                    writer.write('=');
+                    writer.write(table[d >> 18]);
+                    writer.write(table[(d >> 12) & 0X3F]);
+                    if (pad) {
+                        writer.write((dLimit + 1 < data.length) ? table[(d >> 6) & 0X3F] : '=');
+                        writer.write('=');
+                    } else {
+                        if (dLimit + 1 < data.length) {
+                            writer.write(table[(d >> 6) & 0X3F]);
+                        }
+                    }
                     if (chunked && charCount != 0) {
                         writer.write(CHUNK_SEPARATOR);
                     }
@@ -141,6 +156,74 @@ public static Writable encodeBase64(final byte[] data) {
         return encodeBase64(data, false);
     }
 
+    /**
+     * Produce a Writable object which writes the Base64 URL and Filename Safe encoding of the byte array.
+     * Calling toString() on the result returns the encoding as a String. For more
+     * information on Base64 URL and Filename Safe encoding see <code>RFC 4648 - Section 5
+     * Base 64 Encoding with URL and Filename Safe Alphabet</code>.
+     * <p>
+     * The method omits padding and is equivalent to calling
+     * {@link org.codehaus.groovy.runtime.EncodingGroovyMethods#encodeBase64Url(Byte[], boolean)} with a
+     * value of {@code false}.
+     *
+     * @param data Byte array to be encoded
+     * @return object which will write the Base64 URL and Filename Safe encoding of the byte array
+     * @see org.codehaus.groovy.runtime.EncodingGroovyMethods#encodeBase64Url(Byte[], boolean)
+     * @since 2.5
+     */
+    public static Writable encodeBase64Url(Byte[] data) {
+        return encodeBase64Url(data, false);
+    }
+
+    /**
+     * Produce a Writable object which writes the Base64 URL and Filename Safe encoding of the byte array.
+     * Calling toString() on the result returns the encoding as a String. For more
+     * information on Base64 URL and Filename Safe encoding see <code>RFC 4648 - Section 5
+     * Base 64 Encoding with URL and Filename Safe Alphabet</code>.
+     *
+     * @param data Byte array to be encoded
+     * @param pad whether or not the encoded data should be padded
+     * @return object which will write the Base64 URL and Filename Safe encoding of the byte array
+     * @since 2.5
+     */
+    public static Writable encodeBase64Url(Byte[] data, boolean pad) {
+        return encodeBase64Url(DefaultTypeTransformation.convertToByteArray(data), pad);
+    }
+
+    /**
+     * Produce a Writable object which writes the Base64 URL and Filename Safe encoding of the byte array.
+     * Calling toString() on the result returns the encoding as a String. For more
+     * information on Base64 URL and Filename Safe encoding see <code>RFC 4648 - Section 5
+     * Base 64 Encoding with URL and Filename Safe Alphabet</code>.
+     * <p>
+     * The method omits padding and is equivalent to calling
+     * {@link org.codehaus.groovy.runtime.EncodingGroovyMethods#encodeBase64Url(byte[], boolean)} with a
+     * value of {@code false}.
+     *
+     * @param data Byte array to be encoded
+     * @return object which will write the Base64 URL and Filename Safe encoding of the byte array
+     * @see org.codehaus.groovy.runtime.EncodingGroovyMethods#encodeBase64Url(byte[], boolean)
+     * @since 2.5
+     */
+    public static Writable encodeBase64Url(final byte[] data) {
+        return encodeBase64Url(data, false);
+    }
+
+    /**
+     * Produce a Writable object which writes the Base64 URL and Filename Safe encoding of the byte array.
+     * Calling toString() on the result returns the encoding as a String. For more
+     * information on Base64 URL and Filename Safe encoding see <code>RFC 4648 - Section 5
+     * Base 64 Encoding with URL and Filename Safe Alphabet</code>.
+     *
+     * @param data Byte array to be encoded
+     * @param pad whether or not the encoded data should be padded
+     * @return object which will write the Base64 URL and Filename Safe encoding of the byte array
+     * @since 2.5
+     */
+    public static Writable encodeBase64Url(final byte[] data, final boolean pad) {
+        return encodeBase64(data, false, true, pad);
+    }
+
     /**
      * Decode the String from Base64 into a byte array.
      *
@@ -149,14 +232,29 @@ public static Writable encodeBase64(final byte[] data) {
      * @since 1.0
      */
     public static byte[] decodeBase64(String value) {
+        return decodeBase64(value, false);
+    }
+
+    /**
+     * Decodes a Base64 URL and Filename Safe encoded String into a byte array.
+     *
+     * @param value the string to be decoded
+     * @return the decoded bytes as an array
+     * @since 2.5
+     */
+    public static byte[] decodeBase64Url(String value) {
+        return decodeBase64(value, true);
+    }
+
+    private static byte[] decodeBase64(String value, boolean urlSafe) {
         int byteShift = 4;
         int tmp = 0;
         boolean done = false;
         final StringBuilder buffer = new StringBuilder();
-
+        final byte[] table = urlSafe ? TRANSLATE_TABLE_URLSAFE : TRANSLATE_TABLE;
         for (int i = 0; i != value.length(); i++) {
             final char c = value.charAt(i);
-            final int sixBit = (c < 123) ? EncodingGroovyMethodsSupport.TRANSLATE_TABLE[c] : 66;
+            final int sixBit = (c < 123) ? table[c] : 66;
 
             if (sixBit < 64) {
                 if (done)

src/main/org/codehaus/groovy/runtime/EncodingGroovyMethodsSupport.java

@@ -54,4 +54,37 @@ public class EncodingGroovyMethodsSupport {
                     + "\u0029\u002A\u002B\u002C\u002D\u002E\u002F\u0030"
                     //        x     y     z
                     + "\u0031\u0032\u0033").getBytes();
+
+    static final byte[] TRANSLATE_TABLE_URLSAFE = (
+            "\u0042\u0042\u0042\u0042\u0042\u0042\u0042\u0042"
+                    //                    \t    \n                \r
+                    + "\u0042\u0042\u0041\u0041\u0042\u0042\u0041\u0042"
+                    //
+                    + "\u0042\u0042\u0042\u0042\u0042\u0042\u0042\u0042"
+                    //
+                    + "\u0042\u0042\u0042\u0042\u0042\u0042\u0042\u0042"
+                    //        sp    !     "     #     $     %     &     '
+                    + "\u0041\u0042\u0042\u0042\u0042\u0042\u0042\u0042"
+                    //         (    )     *     +     ,     -     .     /
+                    + "\u0042\u0042\u0042\u0042\u0042\u003E\u0042\u0042"
+                    //         0    1     2     3     4     5     6     7
+                    + "\u0034\u0035\u0036\u0037\u0038\u0039\u003A\u003B"
+                    //         8    9     :     ;     <     =     >     ?
+                    + "\u003C\u003D\u0042\u0042\u0042\u0040\u0042\u0042"
+                    //         @    A     B     C     D     E     F     G
+                    + "\u0042\u0000\u0001\u0002\u0003\u0004\u0005\u0006"
+                    //         H    I   J K   L     M   N   O
+                    + "\u0007\u0008\t\n\u000B\u000C\r\u000E"
+                    //         P    Q     R     S     T     U     V    W
+                    + "\u000F\u0010\u0011\u0012\u0013\u0014\u0015\u0016"
+                    //         X    Y     Z     [     \     ]     ^    _
+                    + "\u0017\u0018\u0019\u0042\u0042\u0042\u0042\u003F"
+                    //         '    a     b     c     d     e     f     g
+                    + "\u0042\u001A\u001B\u001C\u001D\u001E\u001F\u0020"
+                    //        h   i   j     k     l     m     n     o    p
+                    + "\u0021\"\u0023\u0024\u0025\u0026\u0027\u0028"
+                    //        p     q     r     s     t     u     v     w
+                    + "\u0029\u002A\u002B\u002C\u002D\u002E\u002F\u0030"
+                    //        x     y     z
+                    + "\u0031\u0032\u0033").getBytes();
 }

src/test/groovy/Base64Test.groovy

@@ -18,10 +18,15 @@
  */
 package groovy
 
+import java.nio.charset.StandardCharsets
+
 class Base64Test extends GroovyTestCase {
     String testString = '\u00A71234567890-=\u00B1!@\u00A3\$%^&*()_+qwertyuiop[]QWERTYUIOP{}asdfghjkl;\'\\ASDFGHJKL:"|`zxcvbnm,./~ZXCVBNM<>?\u0003\u00ff\u00f0\u000f'
     byte[] testBytes = testString.getBytes("ISO-8859-1")
 
+    // Test bytes that have both the 62nd and 63rd base64 alphabet in the encoded string
+    static final byte[] testBytesChar62And63 = new BigInteger('4bf7ce5201fe239ab42ebead5acd8fa3', 16).toByteArray()
+
     void testCodec() {
         // turn the bytes back into a string for later comparison
         def savedString = new String(testBytes, "ISO-8859-1")
@@ -62,4 +67,146 @@ class Base64Test extends GroovyTestCase {
         def encodedBytes = testBytes.encodeBase64().toString()
         assert encodedBytes == 'pzEyMzQ1Njc4OTAtPbEhQKMkJV4mKigpXytxd2VydHl1aW9wW11RV0VSVFlVSU9Qe31hc2RmZ2hqa2w7J1xBU0RGR0hKS0w6InxgenhjdmJubSwuL35aWENWQk5NPD4/A//wDw=='
     }
+
+    void testRfc4648Section10Encoding() {
+        assert b64('') == ''
+        assert b64('f') == 'Zg=='
+        assert b64('fo') == 'Zm8='
+        assert b64('foo') == 'Zm9v'
+        assert b64('foob') == 'Zm9vYg=='
+        assert b64('fooba') == 'Zm9vYmE='
+        assert b64('foobar') == 'Zm9vYmFy'
+    }
+
+    void testRfc4648Section10Decoding() {
+        assert decodeB64('') == ''
+
+        assert decodeB64('Zg') == 'f'
+        assert decodeB64('Zg==') == 'f'
+
+        assert decodeB64('Zm8') == 'fo'
+        assert decodeB64('Zm8=') == 'fo'
+
+        assert decodeB64('Zm9v') == 'foo'
+
+        assert decodeB64('Zm9vYg') == 'foob'
+        assert decodeB64('Zm9vYg==') == 'foob'
+
+        assert decodeB64('Zm9vYmE') == 'fooba'
+        assert decodeB64('Zm9vYmE=') == 'fooba'
+
+        assert decodeB64('Zm9vYmFy') == 'foobar'
+    }
+
+    void testRfc4648Section10EncodingUrlSafe() {
+        assert b64url('') == ''
+        assert b64url('f') == 'Zg'
+        assert b64url('fo') == 'Zm8'
+        assert b64url('foo') == 'Zm9v'
+        assert b64url('foob') == 'Zm9vYg'
+        assert b64url('fooba') == 'Zm9vYmE'
+        assert b64url('foobar') == 'Zm9vYmFy'
+    }
+
+    void testRfc4648Section10EncodingUrlSafeWithPadding() {
+        assert b64url('', true) == ''
+        assert b64url('f', true) == 'Zg=='
+        assert b64url('fo', true) == 'Zm8='
+        assert b64url('foo', true) == 'Zm9v'
+        assert b64url('foob', true) == 'Zm9vYg=='
+        assert b64url('fooba', true) == 'Zm9vYmE='
+        assert b64url('foobar', true) == 'Zm9vYmFy'
+    }
+
+    void testRfc4648Section10DecodingUrlSafe() {
+        assert decodeB64url('') == ''
+
+        assert decodeB64url('Zg') == 'f'
+        assert decodeB64url('Zg==') == 'f'
+
+        assert decodeB64url('Zm8') == 'fo'
+        assert decodeB64url('Zm8=') == 'fo'
+
+        assert decodeB64url('Zm9v') == 'foo'
+
+        assert decodeB64url('Zm9vYg') == 'foob'
+        assert decodeB64url('Zm9vYg==') == 'foob'
+
+        assert decodeB64url('Zm9vYmE') == 'fooba'
+        assert decodeB64url('Zm9vYmE=') == 'fooba'
+
+        assert decodeB64url('Zm9vYmFy') == 'foobar'
+    }
+
+    void testEncodingWithChar62And63() {
+        assert testBytesChar62And63.encodeBase64().toString() == 'S/fOUgH+I5q0Lr6tWs2Pow=='
+    }
+
+    void testUrlSafeEncodingWithChar62And63() {
+        assert testBytesChar62And63.encodeBase64Url().toString() == 'S_fOUgH-I5q0Lr6tWs2Pow'
+        assert testBytesChar62And63.encodeBase64Url(true).toString() == 'S_fOUgH-I5q0Lr6tWs2Pow=='
+    }
+
+    void testDecodingWithChar62And63() {
+        assert 'S/fOUgH+I5q0Lr6tWs2Pow=='.decodeBase64() == testBytesChar62And63
+        assert 'S/fOUgH+I5q0Lr6tWs2Pow'.decodeBase64() == testBytesChar62And63
+    }
+
+    void testUrlSafeDecodingWithChar62And63() {
+        assert 'S_fOUgH-I5q0Lr6tWs2Pow=='.decodeBase64Url() == testBytesChar62And63
+        assert 'S_fOUgH-I5q0Lr6tWs2Pow'.decodeBase64Url() == testBytesChar62And63
+    }
+
+    void testUrlSafeEncodingByDefaultOmitsPadding() {
+        assert testBytes.encodeBase64Url().toString() ==
+                'pzEyMzQ1Njc4OTAtPbEhQKMkJV4mKigpXytxd2VydHl1aW9wW11RV0VSVFlVSU9Qe31h' +
+                'c2RmZ2hqa2w7J1xBU0RGR0hKS0w6InxgenhjdmJubSwuL35aWENWQk5NPD4_A__wDw'
+    }
+
+    void testUrlSafeEncodingWithPadding() {
+        assert testBytes.encodeBase64Url(true).toString() ==
+                'pzEyMzQ1Njc4OTAtPbEhQKMkJV4mKigpXytxd2VydHl1aW9wW11RV0VSVFlVSU9Qe31h' +
+                'c2RmZ2hqa2w7J1xBU0RGR0hKS0w6InxgenhjdmJubSwuL35aWENWQk5NPD4_A__wDw=='
+    }
+
+    void testDecodingNonBase64Alphabet() {
+        shouldFail {
+            decodeB64('S_fOUgH-I5q0Lr6tWs2Pow==')
+        }
+    }
+
+    void testUrlSafeDecodingNonUrlSafeAlphabet() {
+        shouldFail {
+            decodeB64url('S/fOUgH+I5q0Lr6tWs2Pow==')
+        }
+    }
+
+    void testDecodingWithInnerPad() {
+        shouldFail {
+            decodeB64('Zm9v=YmE=')
+        }
+    }
+
+    void testUrlSafeDecodingWithInnerPad() {
+        shouldFail {
+            decodeB64url('Zm9v=YmE=')
+        }
+    }
+
+    // Test helper methods
+    private static String b64(String s) {
+        s.getBytes(StandardCharsets.UTF_8).encodeBase64().toString()
+    }
+
+    private static String b64url(String s, boolean pad=false) {
+        s.getBytes(StandardCharsets.UTF_8).encodeBase64Url(pad).toString()
+    }
+
+    private static String decodeB64(String s) {
+        new String(s.decodeBase64(), StandardCharsets.UTF_8)
+    }
+
+    private static String decodeB64url(String s) {
+        new String(s.decodeBase64Url(), StandardCharsets.UTF_8)
+    }
 }