summaryrefslogtreecommitdiff
path: root/enc/iso_8859_16.c
diff options
context:
space:
mode:
authorduerst <duerst@b2dd03c8-39d4-4d8f-98ff-823fe69b080e>2016-07-06 09:34:23 +0000
committerduerst <duerst@b2dd03c8-39d4-4d8f-98ff-823fe69b080e>2016-07-06 09:34:23 +0000
commit79b1758b65563815abb5d626a21b597e39fcfea4 (patch)
tree0e31c28ad588ef35f1407e288bc03ead405ac00b /enc/iso_8859_16.c
parent64fea27496bd10a5ccdb78abcd541f9c1a9174c8 (diff)
* enc/iso_8859_16.c, test/ruby/enc/test_case_comprehensive.rb:
Implement non-ASCII case conversion for ISO-8859-16, by Satoshi Kayama. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55590 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Diffstat (limited to 'enc/iso_8859_16.c')
-rw-r--r--enc/iso_8859_16.c58
1 files changed, 57 insertions, 1 deletions
diff --git a/enc/iso_8859_16.c b/enc/iso_8859_16.c
index 7e04403..2ed2f3c 100644
--- a/enc/iso_8859_16.c
+++ b/enc/iso_8859_16.c
@@ -222,6 +222,62 @@ get_case_fold_codes_by_str(OnigCaseFoldType flag,
flag, p, end, items);
}
+#ifdef ONIG_CASE_MAPPING
+static int
+case_map (OnigCaseFoldType* flagP, const OnigUChar** pp,
+ const OnigUChar* end, OnigUChar* to, OnigUChar* to_end,
+ const struct OnigEncodingTypeST* enc)
+{
+ OnigCodePoint code;
+ OnigUChar *to_start = to;
+ OnigCaseFoldType flags = *flagP;
+
+ while (*pp<end && to<to_end) {
+ code = *(*pp)++;
+ if (code==SHARP_s) {
+ if (flags&ONIGENC_CASE_UPCASE) {
+ flags |= ONIGENC_CASE_MODIFIED;
+ *to++ = 'S';
+ code = (flags&ONIGENC_CASE_TITLECASE) ? 's' : 'S';
+ }
+ else if (flags&ONIGENC_CASE_FOLD) {
+ flags |= ONIGENC_CASE_MODIFIED;
+ *to++ = 's';
+ code = 's';
+ }
+ }
+ //else if (code==0xAA || code==0xBA) ;
+ else if ((EncISO_8859_16_CtypeTable[code] & BIT_CTYPE_UPPER)
+ && (flags & (ONIGENC_CASE_DOWNCASE|ONIGENC_CASE_FOLD))) {
+ flags |= ONIGENC_CASE_MODIFIED;
+ if(code==0xA1 || code==0xBC) code++;
+ else if(code==0xA3 || code==0xAA || code==0xAF) code += 0x10;
+ else if(code==0xA6 || code==0xAC) code += 0x02;
+ else if(code==0xB2) code += 0x07;
+ else if(code==0xB4) code += 0x04;
+ else if(code==0xBE) code += 0x41;
+ else code += 0x20;
+ }
+ else if ((EncISO_8859_16_CtypeTable[code]&BIT_CTYPE_LOWER)
+ && (flags&ONIGENC_CASE_UPCASE)) {
+ flags |= ONIGENC_CASE_MODIFIED;
+ if(code==0xA2 || code==0xBD) code--;
+ else if(code==0xB3 || code==0xBA || code==0xBF) code -= 0x10;
+ else if(code==0xA8 || code==0xAE) code -= 0x02;
+ else if(code==0xB9) code -= 0x07;
+ else if(code==0xB8) code -= 0x04;
+ else if(code==0xFF) code -= 0x41;
+ else code -= 0x20;
+ }
+ *to++ = code;
+ if (flags&ONIGENC_CASE_TITLECASE) /* switch from titlecase to lowercase for capitalize */
+ flags ^= (ONIGENC_CASE_UPCASE|ONIGENC_CASE_DOWNCASE|ONIGENC_CASE_TITLECASE);
+ }
+ *flagP = flags;
+ return (int)(to-to_start);
+}
+#endif /* ONIG_CASE_MAPPING */
+
OnigEncodingDefine(iso_8859_16, ISO_8859_16) = {
onigenc_single_byte_mbc_enc_len,
"ISO-8859-16", /* name */
@@ -242,7 +298,7 @@ OnigEncodingDefine(iso_8859_16, ISO_8859_16) = {
0,
ONIGENC_FLAG_NONE,
#ifdef ONIG_CASE_MAPPING
- onigenc_single_byte_ascii_only_case_map,
+ case_map,
#endif /* ONIG_CASE_MAPPING */
};
ENC_ALIAS("ISO8859-16", "ISO-8859-16")