AuroraMiddleware/scuffed-code
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

ICU-4867 Improved mbcs charset detect data.

Browse Source 
X-SVN-Rev: 19147
This commit is contained in:
Andy Heninger 2006-02-13 02:19:06 +00:00
parent 017be20909
commit ab7b0771e8
1 changed files with 23 additions and 31 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

54
icu4j/src/com/ibm/icu/text/CharsetRecog_mbcs.java
View File

@ -191,16 +191,12 @@ abstract class CharsetRecog_mbcs extends CharsetRecognizer {
// TODO: This set of data comes from the character frequency-
// of-occurence analysis tool. The data needs to be moved
// into a resource and loaded from there.
{0x82cc, 0x82c5, 0x82a2, 0x815b, 0x8141, 0x82b5, 0x82c9, 0x82b7, 0x8142, 0x82c4,
0x82c6, 0x82cd, 0x82dc, 0x82f0, 0x82e9, 0x82c8, 0x82aa, 0x8393, 0x82bd, 0x8358,
0x82ea, 0x8343, 0x82a4, 0x82a9, 0x8367, 0x82b1, 0x8145, 0x82e0, 0x838b, 0x834e,
0x82e8, 0x82e7, 0x8140, 0x8362, 0x8389, 0x82c1, 0x838a, 0x82ab, 0x8376, 0x82b3,
0x82f1, 0x82a0, 0x8368, 0x93fa, 0x8175, 0x8176, 0x835e, 0x82e6, 0x8357, 0x82ad,
0x8381, 0x82a6, 0x82b9, 0x82bb, 0x82be, 0x8341, 0x8374, 0x82af, 0x9056, 0x82a8,
0x82c2, 0x8354, 0x8e67, 0x8375, 0x82c7, 0x95f1, 0x8356, 0x967b, 0x92e8, 0x8345,
0x82ce, 0x8385, 0x9770, 0x82df, 0x82dd, 0x836f, 0x8342, 0x8ca7, 0x8352, 0x837d,
0x838d, 0x8346, 0x834f, 0x8380, 0x82ed, 0x8d73, 0x8349, 0x8365, 0x8fee, 0x95b6,
0x8169, 0x816a, 0x836a, 0x8dec, 0x82bf, 0x834c, 0x8366, 0x82e2, 0x838c, 0x945c};
{0x8140, 0x8141, 0x8142, 0x8145, 0x815b, 0x8169, 0x816a, 0x8175, 0x8176, 0x82a0,
0x82a2, 0x82a4, 0x82a9, 0x82aa, 0x82ab, 0x82ad, 0x82af, 0x82b1, 0x82b3, 0x82b5,
0x82b7, 0x82bd, 0x82be, 0x82c1, 0x82c4, 0x82c5, 0x82c6, 0x82c8, 0x82c9, 0x82cc,
0x82cd, 0x82dc, 0x82e0, 0x82e7, 0x82e8, 0x82e9, 0x82ea, 0x82f0, 0x82f1, 0x8341,
0x8343, 0x834e, 0x834f, 0x8358, 0x835e, 0x8362, 0x8367, 0x8375, 0x8376, 0x8389,
0x838a, 0x838b, 0x838d, 0x8393, 0x8e96, 0x93fa, 0x95aa};
boolean nextChar(iteratedChar it, CharsetDetector det) {
it.index = it.nextIndex;
@ -253,11 +249,11 @@ abstract class CharsetRecog_mbcs extends CharsetRecognizer {
0xa558, 0xa569, 0xa5cd, 0xa5e7, 0xa657, 0xa661, 0xa662, 0xa668, 0xa670, 0xa6a8,
0xa6b3, 0xa6b9, 0xa6d3, 0xa6db, 0xa6e6, 0xa6f2, 0xa740, 0xa751, 0xa759, 0xa7da,
0xa8a3, 0xa8a5, 0xa8ad, 0xa8d1, 0xa8d3, 0xa8e4, 0xa8fc, 0xa9c0, 0xa9d2, 0xa9f3,
0xaa6b, 0xaaba, 0xaabe, 0xaacc, 0xaafc, 0xac47, 0xac4f, 0xacb0, 0xacc9, 0xacd2,
0xad59, 0xad6e, 0xaec9, 0xafe0, 0xb0ea, 0xb16f, 0xb2b3, 0xb2c4, 0xb36f, 0xb44c,
0xb44e, 0xb54c, 0xb5a5, 0xb5bd, 0xb5d0, 0xb5d8, 0xb671, 0xb7ed, 0xb867, 0xb944,
0xbad8, 0xbb44, 0xbba1, 0xbc77, 0xbdd1, 0xbe69, 0xc2c4, 0xc3b9, 0xc440, 0xc45f};
0xaa6b, 0xaaba, 0xaabe, 0xaacc, 0xaafc, 0xac47, 0xac4f, 0xacb0, 0xacd2, 0xad59,
0xaec9, 0xafe0, 0xb0ea, 0xb16f, 0xb2b3, 0xb2c4, 0xb36f, 0xb44c, 0xb44e, 0xb54c,
0xb5a5, 0xb5bd, 0xb5d0, 0xb5d8, 0xb671, 0xb7ed, 0xb867, 0xb944, 0xbad8, 0xbb44,
0xbba1, 0xbdd1, 0xc2c4, 0xc3b9, 0xc440, 0xc45f};
boolean nextChar(iteratedChar it, CharsetDetector det) {
it.index = it.nextIndex;
it.error = false;
@ -279,7 +275,7 @@ abstract class CharsetRecog_mbcs extends CharsetRecognizer {
it.charValue = (it.charValue << 8) | secondByte;
if (secondByte < 0x40 ||
secondByte >=0x7f && secondByte <= 0xa0 ||
secondByte ==0x7f ||
secondByte == 0xff) {
it.error = true;
}
@ -496,21 +492,17 @@ abstract class CharsetRecog_mbcs extends CharsetRecognizer {
// TODO: This set of data comes from the character frequency-
// of-occurence analysis tool. The data needs to be moved
// into a resource and loaded from there.
{0xa3ac, 0xb5c4, 0xa1a1, 0xa1a4, 0xa1a3, 0xcac7, 0xd2bb, 0xb4f3, 0xd4da, 0xd6d0,
0xcafd, 0xd3d0, 0xa1f3, 0xb2bb, 0xa3ba, 0xbbfa, 0xc8cb, 0xa1a2, 0xd3c3, 0xd1a7,
0xc8d5, 0xbedd, 0xb8f6, 0xd0c2, 0xcdf8, 0xd2aa, 0xb9fa, 0xc1cb, 0xc9cf, 0xa1b0,
0xa1b1, 0xced2, 0xbcfe, 0xcec4, 0xd2d4, 0xc4dc, 0xc0b4, 0xd4c2, 0xcab1, 0xd0d0,
0xbdcc, 0xbfc9, 0xb6d4, 0xbcdb, 0xb1be, 0xb3f6, 0xb8b4, 0xc9fa, 0xb1b8, 0xbcbc,
0xcfc2, 0xbacd, 0xbecd, 0xb3c9, 0xd5e2, 0xb8df, 0xb7d6, 0xc5cc, 0xbfc6, 0xbbe1,
0xceaa, 0xc8e7, 0xcfb5, 0xa1f1, 0xc4ea, 0xb1a8, 0xb6af, 0xc0ed, 0xd3fd, 0xb7a2,
0xc8ab, 0xb7bd, 0xcee5, 0xc2db, 0xbba7, 0xd0d4, 0xb9c9, 0xc3c7, 0xb9fd, 0xcad0,
0xb5e3, 0xbbd6, 0xcfd6, 0xcab5, 0xd2b2, 0xbfb4, 0xb6e0, 0xccec, 0xc7f8, 0xd0c5,
0xcad6, 0xb9d8, 0xb5bd, 0xb7dd, 0xc6f7, 0xcaf5, 0xa3a1, 0xb7a8, 0xb9ab, 0xd2b5,
0xcbf9, 0xcdbc, 0xc6e4, 0xd3da, 0xd0a1, 0xd1a1, 0xd3ce, 0xbfaa, 0xb4e6, 0xc4bf,
0xd7f7, 0xb5e7, 0xcdb3, 0xc7e9, 0xd7ee, 0xc6c0, 0xcfdf, 0xb5d8, 0xb5c0, 0xbead,
0xb4c5, 0xc6b7, 0xc4da, 0xd0c4, 0xb9a4, 0xd4aa, 0xc2bc, 0xc3c0, 0xbaf3, 0xcabd,
0xbcd2, 0xcef1, 0xbdab, 0xa3ad, 0xa3bf, 0xb3a4, 0xb9fb, 0xd6ae, 0xc1bf, 0xbbd8,
0xb8f1, 0xb6f8, 0xb6a8, 0xcde2, 0xbac3, 0xb3cc, 0xccd8, 0xd7d4, 0xcbb5};
{0xa1a1, 0xa1a2, 0xa1a3, 0xa1a4, 0xa1b0, 0xa1b1, 0xa1f1, 0xa1f3, 0xa3a1, 0xa3ac,
0xa3ba, 0xb1a8, 0xb1b8, 0xb1be, 0xb2bb, 0xb3c9, 0xb3f6, 0xb4f3, 0xb5bd, 0xb5c4,
0xb5e3, 0xb6af, 0xb6d4, 0xb6e0, 0xb7a2, 0xb7a8, 0xb7bd, 0xb7d6, 0xb7dd, 0xb8b4,
0xb8df, 0xb8f6, 0xb9ab, 0xb9c9, 0xb9d8, 0xb9fa, 0xb9fd, 0xbacd, 0xbba7, 0xbbd6,
0xbbe1, 0xbbfa, 0xbcbc, 0xbcdb, 0xbcfe, 0xbdcc, 0xbecd, 0xbedd, 0xbfb4, 0xbfc6,
0xbfc9, 0xc0b4, 0xc0ed, 0xc1cb, 0xc2db, 0xc3c7, 0xc4dc, 0xc4ea, 0xc5cc, 0xc6f7,
0xc7f8, 0xc8ab, 0xc8cb, 0xc8d5, 0xc8e7, 0xc9cf, 0xc9fa, 0xcab1, 0xcab5, 0xcac7,
0xcad0, 0xcad6, 0xcaf5, 0xcafd, 0xccec, 0xcdf8, 0xceaa, 0xcec4, 0xced2, 0xcee5,
0xcfb5, 0xcfc2, 0xcfd6, 0xd0c2, 0xd0c5, 0xd0d0, 0xd0d4, 0xd1a7, 0xd2aa, 0xd2b2,
0xd2b5, 0xd2bb, 0xd2d4, 0xd3c3, 0xd3d0, 0xd3fd, 0xd4c2, 0xd4da, 0xd5e2, 0xd6d0};
String getName() {
return "GB18030";