mirrors/cpython
1
0
Fork 0
mirror of https://github.com/python/cpython.git synced 2025-11-25 21:11:09 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 10

Update big5hkscs codec to conform to the HKSCS:2004 revision.

Browse source
This commit is contained in:
Hye-Shik Chang 2008-02-08 17:10:20 +00:00
parent 27d339446a
commit 01612e7dec
6 changed files with 1868 additions and 1777 deletions
Download patch file Download diff file Expand all files Collapse all files

56
Modules/cjkcodecs/_codecs_hk.c
View file

@ -26,6 +26,16 @@ CODEC_INIT(big5hkscs)
return 0;
}
/*
* There are four possible pair unicode -> big5hkscs maps as in HKSCS 2004:
* U+00CA U+0304 -> 8862 (U+00CA alone is mapped to 8866)
* U+00CA U+030C -> 8864
* U+00EA U+0304 -> 88a3 (U+00EA alone is mapped to 88a7)
* U+00EA U+030C -> 88a5
* These are handled by not mapping tables but a hand-written code.
*/
static const DBCHAR big5hkscs_pairenc_table[4] = {0x8862, 0x8864, 0x88a3, 0x88a5};
ENCODER(big5hkscs)
{
while (inleft > 0) {
@ -46,7 +56,27 @@ ENCODER(big5hkscs)
REQUIRE_OUTBUF(2)
if (c < 0x10000) {
TRYMAP_ENC(big5hkscs_bmp, code, c);
TRYMAP_ENC(big5hkscs_bmp, code, c) {
if (code == MULTIC) {
if (inleft >= 2 &&
((c & 0xffdf) == 0x00ca) &&
(((*inbuf)[1] & 0xfff7) == 0x0304)) {
code = big5hkscs_pairenc_table[
((c >> 4) |
((*inbuf)[1] >> 3)) & 3];
insize = 2;
}
else if (inleft < 2 &&
!(flags & MBENC_FLUSH))
return MBERR_TOOFEW;
else {
if (c == 0xca)
code = 0x8866;
else /* c == 0xea */
code = 0x88a7;
}
}
}
else TRYMAP_ENC(big5, code, c);
else return 1;
}
@ -67,7 +97,7 @@ ENCODER(big5hkscs)
return 0;
}
#define BH2S(c1, c2) (((c1) - 0x88) * (0xfe - 0x40 + 1) + ((c2) - 0x40))
#define BH2S(c1, c2) (((c1) - 0x87) * (0xfe - 0x40 + 1) + ((c2) - 0x40))
DECODER(big5hkscs)
{
@ -96,19 +126,19 @@ hkscsdec: TRYMAP_DEC(big5hkscs, decoded, c, IN2) {
int s = BH2S(c, IN2);
const unsigned char *hintbase;
assert(0x88 <= c && c <= 0xfe);
assert(0x87 <= c && c <= 0xfe);
assert(0x40 <= IN2 && IN2 <= 0xfe);
if (BH2S(0x88, 0x40) <= s && s <= BH2S(0xa0, 0xfe)) {
if (BH2S(0x87, 0x40) <= s && s <= BH2S(0xa0, 0xfe)) {
hintbase = big5hkscs_phint_0;
s -= BH2S(0x88, 0x40);
s -= BH2S(0x87, 0x40);
}
else if (BH2S(0xc6,0xa1) <= s && s <= BH2S(0xc8,0xfe)){
hintbase = big5hkscs_phint_11939;
hintbase = big5hkscs_phint_12130;
s -= BH2S(0xc6, 0xa1);
}
else if (BH2S(0xf9,0xd6) <= s && s <= BH2S(0xfe,0xfe)){
hintbase = big5hkscs_phint_21733;
hintbase = big5hkscs_phint_21924;
s -= BH2S(0xf9, 0xd6);
}
else
@ -123,7 +153,17 @@ hkscsdec: TRYMAP_DEC(big5hkscs, decoded, c, IN2) {
NEXT(2, 1)
}
}
else return 2;
else {
switch ((c << 8) | IN2) {
case 0x8862: WRITE2(0x00ca, 0x0304); break;
case 0x8864: WRITE2(0x00ca, 0x030c); break;
case 0x88a3: WRITE2(0x00ea, 0x0304); break;
case 0x88a5: WRITE2(0x00ea, 0x030c); break;
default: return 2;
}
NEXT(2, 2) /* all decoded codepoints are pairs, above. */
}
}
return 0;