VariantSupportpublic class VariantSupport extends Variant Supports reading and writing of variant data.
FIXME (3): Reading and writing should be made more
uniform than it is now. The following items should be resolved:
Reading requires a length parameter that is 4 byte greater than the
actual data, because the variant type field is included. Reading reads from a byte array while writing writes to an byte array
output stream. |
| Fields Summary |
|---|
| private static boolean | logUnsupportedTypes | | protected static List | unsupportedMessage Keeps a list of the variant types an "unsupported" message has already
been issued for. |
| Methods Summary |
|---|
| public static java.lang.String | codepageToEncoding(int codepage)Turns a codepage number into the equivalent character encoding's
name.
if (codepage <= 0)
throw new UnsupportedEncodingException
("Codepage number may not be " + codepage);
switch (codepage)
{
case Constants.CP_UTF16:
return "UTF-16";
case Constants.CP_UTF16_BE:
return "UTF-16BE";
case Constants.CP_UTF8:
return "UTF-8";
case Constants.CP_037:
return "cp037";
case Constants.CP_GBK:
return "GBK";
case Constants.CP_MS949:
return "ms949";
case Constants.CP_WINDOWS_1250:
return "windows-1250";
case Constants.CP_WINDOWS_1251:
return "windows-1251";
case Constants.CP_WINDOWS_1252:
return "windows-1252";
case Constants.CP_WINDOWS_1253:
return "windows-1253";
case Constants.CP_WINDOWS_1254:
return "windows-1254";
case Constants.CP_WINDOWS_1255:
return "windows-1255";
case Constants.CP_WINDOWS_1256:
return "windows-1256";
case Constants.CP_WINDOWS_1257:
return "windows-1257";
case Constants.CP_WINDOWS_1258:
return "windows-1258";
case Constants.CP_JOHAB:
return "johab";
case Constants.CP_MAC_ROMAN:
return "MacRoman";
case Constants.CP_MAC_JAPAN:
return "SJIS";
case Constants.CP_MAC_CHINESE_TRADITIONAL:
return "Big5";
case Constants.CP_MAC_KOREAN:
return "EUC-KR";
case Constants.CP_MAC_ARABIC:
return "MacArabic";
case Constants.CP_MAC_HEBREW:
return "MacHebrew";
case Constants.CP_MAC_GREEK:
return "MacGreek";
case Constants.CP_MAC_CYRILLIC:
return "MacCyrillic";
case Constants.CP_MAC_CHINESE_SIMPLE:
return "EUC_CN";
case Constants.CP_MAC_ROMANIA:
return "MacRomania";
case Constants.CP_MAC_UKRAINE:
return "MacUkraine";
case Constants.CP_MAC_THAI:
return "MacThai";
case Constants.CP_MAC_CENTRAL_EUROPE:
return "MacCentralEurope";
case Constants.CP_MAC_ICELAND:
return "MacIceland";
case Constants.CP_MAC_TURKISH:
return "MacTurkish";
case Constants.CP_MAC_CROATIAN:
return "MacCroatian";
case Constants.CP_US_ACSII:
case Constants.CP_US_ASCII2:
return "US-ASCII";
case Constants.CP_KOI8_R:
return "KOI8-R";
case Constants.CP_ISO_8859_1:
return "ISO-8859-1";
case Constants.CP_ISO_8859_2:
return "ISO-8859-2";
case Constants.CP_ISO_8859_3:
return "ISO-8859-3";
case Constants.CP_ISO_8859_4:
return "ISO-8859-4";
case Constants.CP_ISO_8859_5:
return "ISO-8859-5";
case Constants.CP_ISO_8859_6:
return "ISO-8859-6";
case Constants.CP_ISO_8859_7:
return "ISO-8859-7";
case Constants.CP_ISO_8859_8:
return "ISO-8859-8";
case Constants.CP_ISO_8859_9:
return "ISO-8859-9";
case Constants.CP_ISO_2022_JP1:
case Constants.CP_ISO_2022_JP2:
case Constants.CP_ISO_2022_JP3:
return "ISO-2022-JP";
case Constants.CP_ISO_2022_KR:
return "ISO-2022-KR";
case Constants.CP_EUC_JP:
return "EUC-JP";
case Constants.CP_EUC_KR:
return "EUC-KR";
case Constants.CP_GB2312:
return "GB2312";
case Constants.CP_GB18030:
return "GB18030";
case Constants.CP_SJIS:
return "SJIS";
default:
return "cp" + codepage;
}
| | public static boolean | isLogUnsupportedTypes()Checks whether logging of unsupported variant types warning is turned
on or off.
return logUnsupportedTypes;
| | public static java.lang.Object | read(byte[] src, int offset, int length, long type, int codepage)Reads a variant type from a byte array.
Object value;
int o1 = offset;
int l1 = length - LittleEndian.INT_SIZE;
long lType = type;
/* Instead of trying to read 8-bit characters from a Unicode string,
* read 16-bit characters. */
if (codepage == Constants.CP_UNICODE && type == Variant.VT_LPSTR)
lType = Variant.VT_LPWSTR;
switch ((int) lType)
{
case Variant.VT_EMPTY:
{
value = null;
break;
}
case Variant.VT_I2:
{
/*
* Read a short. In Java it is represented as an
* Integer object.
*/
value = new Integer(LittleEndian.getShort(src, o1));
break;
}
case Variant.VT_I4:
{
/*
* Read a word. In Java it is represented as an
* Integer object.
*/
value = new Integer(LittleEndian.getInt(src, o1));
break;
}
case Variant.VT_I8:
{
/*
* Read a double word. In Java it is represented as a
* Long object.
*/
value = new Long(LittleEndian.getLong(src, o1));
break;
}
case Variant.VT_R8:
{
/*
* Read an eight-byte double value. In Java it is represented as
* a Double object.
*/
value = new Double(LittleEndian.getDouble(src, o1));
break;
}
case Variant.VT_FILETIME:
{
/*
* Read a FILETIME object. In Java it is represented
* as a Date object.
*/
final long low = LittleEndian.getUInt(src, o1);
o1 += LittleEndian.INT_SIZE;
final long high = LittleEndian.getUInt(src, o1);
value = Util.filetimeToDate((int) high, (int) low);
break;
}
case Variant.VT_LPSTR:
{
/*
* Read a byte string. In Java it is represented as a
* String object. The 0x00 bytes at the end must be
* stripped.
*/
final int first = o1 + LittleEndian.INT_SIZE;
long last = first + LittleEndian.getUInt(src, o1) - 1;
o1 += LittleEndian.INT_SIZE;
while (src[(int) last] == 0 && first <= last)
last--;
final int l = (int) (last - first + 1);
value = codepage != -1 ?
new String(src, first, l,
codepageToEncoding(codepage)) :
new String(src, first, l);
break;
}
case Variant.VT_LPWSTR:
{
/*
* Read a Unicode string. In Java it is represented as
* a String object. The 0x00 bytes at the end must be
* stripped.
*/
final int first = o1 + LittleEndian.INT_SIZE;
long last = first + LittleEndian.getUInt(src, o1) - 1;
long l = last - first;
o1 += LittleEndian.INT_SIZE;
StringBuffer b = new StringBuffer((int) (last - first));
for (int i = 0; i <= l; i++)
{
final int i1 = o1 + (i * 2);
final int i2 = i1 + 1;
final int high = src[i2] << 8;
final int low = src[i1] & 0x00ff;
final char c = (char) (high | low);
b.append(c);
}
/* Strip 0x00 characters from the end of the string: */
while (b.length() > 0 && b.charAt(b.length() - 1) == 0x00)
b.setLength(b.length() - 1);
value = b.toString();
break;
}
case Variant.VT_CF:
{
final byte[] v = new byte[l1];
for (int i = 0; i < l1; i++)
v[i] = src[(o1 + i)];
value = v;
break;
}
case Variant.VT_BOOL:
{
/*
* The first four bytes in src, from src[offset] to
* src[offset + 3] contain the DWord for VT_BOOL, so
* skip it, we don't need it.
*/
// final int first = offset + LittleEndian.INT_SIZE;
long bool = LittleEndian.getUInt(src, o1);
if (bool != 0)
value = Boolean.TRUE;
else
value = Boolean.FALSE;
break;
}
default:
{
final byte[] v = new byte[l1];
for (int i = 0; i < l1; i++)
v[i] = src[(o1 + i)];
throw new ReadingNotSupportedException(type, v);
}
}
return value;
| | public static void | setLogUnsupportedTypes(boolean logUnsupportedTypes)Specifies whether warnings about unsupported variant types are to be
written to System.err or not.
VariantSupport.logUnsupportedTypes = logUnsupportedTypes;
| | public static int | write(java.io.OutputStream out, long type, java.lang.Object value, int codepage)Writes a variant value to an output stream. This method ensures that
always a multiple of 4 bytes is written.
If the codepage is UTF-16, which is encouraged, strings
must always be written as {@link Variant#VT_LPWSTR}
strings, not as {@link Variant#VT_LPSTR} strings. This method ensure this
by converting strings appropriately, if needed.
int length = 0;
switch ((int) type)
{
case Variant.VT_BOOL:
{
int trueOrFalse;
if (((Boolean) value).booleanValue())
trueOrFalse = 1;
else
trueOrFalse = 0;
length = TypeWriter.writeUIntToStream(out, trueOrFalse);
break;
}
case Variant.VT_LPSTR:
{
final byte[] bytes =
(codepage == -1 ?
((String) value).getBytes() :
((String) value).getBytes(codepageToEncoding(codepage)));
length = TypeWriter.writeUIntToStream(out, bytes.length + 1);
final byte[] b = new byte[bytes.length + 1];
System.arraycopy(bytes, 0, b, 0, bytes.length);
b[b.length - 1] = 0x00;
out.write(b);
length += b.length;
break;
}
case Variant.VT_LPWSTR:
{
final int nrOfChars = ((String) value).length() + 1;
length += TypeWriter.writeUIntToStream(out, nrOfChars);
char[] s = Util.pad4((String) value);
for (int i = 0; i < s.length; i++)
{
final int high = ((s[i] & 0x0000ff00) >> 8);
final int low = (s[i] & 0x000000ff);
final byte highb = (byte) high;
final byte lowb = (byte) low;
out.write(lowb);
out.write(highb);
length += 2;
}
out.write(0x00);
out.write(0x00);
length += 2;
break;
}
case Variant.VT_CF:
{
final byte[] b = (byte[]) value;
out.write(b);
length = b.length;
break;
}
case Variant.VT_EMPTY:
{
TypeWriter.writeUIntToStream(out, Variant.VT_EMPTY);
length = LittleEndianConsts.INT_SIZE;
break;
}
case Variant.VT_I2:
{
TypeWriter.writeToStream(out, ((Integer) value).shortValue());
length = LittleEndianConsts.SHORT_SIZE;
break;
}
case Variant.VT_I4:
{
if (!(value instanceof Integer))
{
throw new ClassCastException("Could not cast an object to "
+ Integer.class.toString() + ": "
+ value.getClass().toString() + ", "
+ value.toString());
}
length += TypeWriter.writeToStream(out,
((Integer) value).intValue());
break;
}
case Variant.VT_I8:
{
TypeWriter.writeToStream(out, ((Long) value).longValue());
length = LittleEndianConsts.LONG_SIZE;
break;
}
case Variant.VT_R8:
{
length += TypeWriter.writeToStream(out,
((Double) value).doubleValue());
break;
}
case Variant.VT_FILETIME:
{
long filetime = Util.dateToFileTime((Date) value);
int high = (int) ((filetime >> 32) & 0x00000000FFFFFFFFL);
int low = (int) (filetime & 0x00000000FFFFFFFFL);
length += TypeWriter.writeUIntToStream
(out, 0x0000000FFFFFFFFL & low);
length += TypeWriter.writeUIntToStream
(out, 0x0000000FFFFFFFFL & high);
break;
}
default:
{
/* The variant type is not supported yet. However, if the value
* is a byte array we can write it nevertheless. */
if (value instanceof byte[])
{
final byte[] b = (byte[]) value;
out.write(b);
length = b.length;
writeUnsupportedTypeMessage
(new WritingNotSupportedException(type, value));
}
else
throw new WritingNotSupportedException(type, value);
break;
}
}
return length;
| | protected static void | writeUnsupportedTypeMessage(org.apache.poi.hpsf.UnsupportedVariantTypeException ex)Writes a warning to System.err that a variant type is
unsupported by HPSF. Such a warning is written only once for each variant
type. Log messages can be turned on or off by
if (isLogUnsupportedTypes())
{
if (unsupportedMessage == null)
unsupportedMessage = new LinkedList();
Long vt = new Long(ex.getVariantType());
if (!unsupportedMessage.contains(vt))
{
System.err.println(ex.getMessage());
unsupportedMessage.add(vt);
}
}
|
|
