diff options
| author | duerst <duerst@b2dd03c8-39d4-4d8f-98ff-823fe69b080e> | 2016-02-06 05:37:29 +0000 |
|---|---|---|
| committer | duerst <duerst@b2dd03c8-39d4-4d8f-98ff-823fe69b080e> | 2016-02-06 05:37:29 +0000 |
| commit | 81515b2381dcd325ca57c0272a551bba4f112afd (patch) | |
| tree | bb427b2ff4bee977ae543f12cf6ed0436fc0007a /enc | |
| parent | 6ab70ad72aed0909868f59135f6f5d1e49877323 (diff) | |
* enc/unicode.c, test/ruby/enc/test_case_mapping.rb: Implemented :fold
option for String#downcase by using case folding data from
regular expression engine, and added a few simple tests.
(with Kimihito Matsui)
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@53747 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Diffstat (limited to 'enc')
| -rw-r--r-- | enc/unicode.c | 31 |
1 files changed, 26 insertions, 5 deletions
diff --git a/enc/unicode.c b/enc/unicode.c index 3f41ea3844..cb9b0a94b4 100644 --- a/enc/unicode.c +++ b/enc/unicode.c @@ -628,40 +628,61 @@ onigenc_unicode_case_map(OnigCaseFoldType* flagP, if (code<='z') { /* ASCII comes first */ if (code>='a' && code<='z') { if (flags&ONIGENC_CASE_UPCASE) { + MODIFIED; if (flags&ONIGENC_CASE_FOLD_TURKISH_AZERI && code==0x0069) /* i -> I WITH DOT ABOVE */ code = 0x0130; else code += 'A'-'a'; - MODIFIED; } } else if (code>='A' && code<='Z') { if (flags&ONIGENC_CASE_DOWNCASE) { + MODIFIED; if (flags&ONIGENC_CASE_FOLD_TURKISH_AZERI && code==0x0049) /* I -> DOTLESS i */ code = 0x0131; else code += 'a'-'A'; - MODIFIED; } } } else if (!(flags&ONIGENC_CASE_ASCII_ONLY) && code>=0x00C0) { /* deal with non-ASCII; nothing relevant below U+00C0 */ + const CodePointList3 *folded; + if (code==0x0130) { if (flags&ONIGENC_CASE_DOWNCASE) { + MODIFIED; if (flags&ONIGENC_CASE_FOLD_TURKISH_AZERI) code = 0x0069; /* I WITH DOT ABOVE -> i */ else { /* make dot above explicit */ to += ONIGENC_CODE_TO_MBC(enc, 0x0069, to); code = 0x0307; /* dot above */ } - MODIFIED; } } - /* the following case can be removed once we rely on data, + /* the following special case for DOTLESS i -> I + * can be removed once we rely on data, * because the mapping is always the same */ - else if (code==0x0131 && (flags&ONIGENC_CASE_UPCASE)) { /* DOTLESS i -> I */ + else if (code==0x0131 && (flags&ONIGENC_CASE_UPCASE)) { code = 0x0049; MODIFIED; } + else if ((folded = onigenc_unicode_fold_lookup(code)) != 0) { + if (flags&ONIGENC_CASE_FOLD) { + const OnigCodePoint *next = folded->code; + int count = OnigCodePointCount(folded->n); + MODIFIED; + if (count==1) + code = *next; + else if (count==2) { + to += ONIGENC_CODE_TO_MBC(enc, *next++, to); + code = *next; + } + else { /* count == 3 */ + to += ONIGENC_CODE_TO_MBC(enc, *next++, to); + to += ONIGENC_CODE_TO_MBC(enc, *next++, to); + code = *next; + } + } + } } to += ONIGENC_CODE_TO_MBC(enc, code, to); /* switch from titlecase to lowercase for capitalize */ |