package genson;
import java.io.*;
/**
* This is an internal class that might evolve in the future into a JsonReader Factory and be moved
* to the stream package.
*/
public final class EncodingAwareReaderFactory
{
static enum UTFEncoding
{
UTF_32BE
(4
), UTF_32LE
(4
), UTF_16BE
(2
), UTF_16LE
(2
), UTF_8
(1
),
5+0%2Fdocs%2Fapi+UNKNOWN">UNKNOWN(-1
);
final int bytes
;
private UTFEncoding
(int bytes
) {
this.
bytes = bytes
;
}
public 1.5.0/docs/api/java/lang/String.html">String encoding
() {
return name
().
replace('_',
'-');
}
}
/**
* Creates java.io.Reader instances with detected encoding from the input stream
* using BOM if present or JSON spec.
*
* Some links:
* http://www.herongyang.com/Unicode/
* http://www.ietf.org/rfc/rfc4627.txt
*
* @throws java.io.IOException
* @throws java.io.UnsupportedEncodingException
*/
public 1.5.0/docs/api/java/io/Reader.html">Reader createReader
(5+0%2Fdocs%2Fapi+InputStream">InputStream is
) throws 1.5.0/docs/api/java/io/IOException.html">IOException {
byte[] bytes =
new byte[4
];
int len = fetchBytes
(bytes, is
);
if (len
< 1
) return new 1.5.0/docs/api/java/io/InputStreamReader.html">InputStreamReader(is
);
// read first 4 bytes if available
int bits_32 =
(bytes
[0
] & 0xFF
) << 24
|
(bytes
[1
] & 0xFF
) << 16
|
(bytes
[2
] & 0xFF
) << 8
|
(bytes
[3
] & 0xFF
);
UTFEncoding encoding = UTFEncoding.
5+0%2Fdocs%2Fapi+UNKNOWN">UNKNOWN;
boolean hasBOM =
false;
// try to detect the encoding from those 4 bytes if BOM is used
if (len == 4
) encoding = detectEncodingFromBOM
(bits_32
);
// no BOM then fall back to JSON spec
if (encoding == UTFEncoding.
5+0%2Fdocs%2Fapi+UNKNOWN">UNKNOWN) {
encoding = detectEncodingUsingJSONSpec
(bits_32
);
} else hasBOM =
true;
// should not happen as we default to UTF-8
if (encoding == UTFEncoding.
5+0%2Fdocs%2Fapi+UNKNOWN">UNKNOWN) {
throw new 1.5.0/docs/api/java/io/UnsupportedEncodingException.html">UnsupportedEncodingException("The encoding could not be detected from the stream.");
}
int usedBOMBytes = hasBOM
? len -
(4 - encoding.
bytes) :
0;
int bytesToUnread = len - usedBOMBytes
;
// small optimization to avoid encapsulation when there is nothing to unread
if (bytesToUnread == 0
) {
return new 1.5.0/docs/api/java/io/InputStreamReader.html">InputStreamReader(is, encoding.
encoding());
} else {
1.5.0/docs/api/java/io/PushbackInputStream.html">PushbackInputStream pis =
new 1.5.0/docs/api/java/io/PushbackInputStream.html">PushbackInputStream(is, bytesToUnread
);
pis.
unread(bytes, usedBOMBytes, bytesToUnread
);
return new 1.5.0/docs/api/java/io/InputStreamReader.html">InputStreamReader(pis, encoding.
encoding());
}
}
private UTFEncoding detectEncodingFromBOM
(int bits_32
) {
int bits_16 = bits_32
>>> 16;
if (bits_32 == 0x0000FEFF
) return UTFEncoding.
UTF_32BE;
else if (bits_32 == 0xFFFE0000
) return UTFEncoding.
UTF_32LE;
else if (bits_16 == 0xFEFF
) return UTFEncoding.
UTF_16BE;
else if (bits_16 == 0xFFFE
) return UTFEncoding.
UTF_16LE;
else if (bits_32
>>> 8 == 0xEFBBBF
) return UTFEncoding.
UTF_8;
else return UTFEncoding.
5+0%2Fdocs%2Fapi+UNKNOWN">UNKNOWN;
}
private UTFEncoding detectEncodingUsingJSONSpec
(int bits_32
) {
int bits_16 = bits_32
>>> 16;
if (bits_32
>>> 8 == 0
) return UTFEncoding.
UTF_32BE;
else if ((bits_32
& 0x00FFFFFF
) == 0
) return UTFEncoding.
UTF_32LE;
else if ((bits_16
& 0xFF00
) == 0
) return UTFEncoding.
UTF_16BE;
else if ((bits_16
& 0x00FF
) == 0
) return UTFEncoding.
UTF_16LE;
else return UTFEncoding.
UTF_8;
}
private int fetchBytes
(byte[] bytes,
5+0%2Fdocs%2Fapi+InputStream">InputStream is
) throws 1.5.0/docs/api/java/io/IOException.html">IOException {
int start =
0;
int bytesRead
;
while(start
< bytes.
length-1
&& (bytesRead = is.
read(bytes, start, bytes.
length-start
)) > -1
) {
start += bytesRead
;
}
return start
;
}
}