diff options
| author | Philipp Gesang <phg42.2a@gmail.com> | 2013-09-21 13:31:56 +0200 |
|---|---|---|
| committer | Philipp Gesang <phg42.2a@gmail.com> | 2013-09-21 13:31:56 +0200 |
| commit | d96d015887e6fcbe1ca8defcec44cb794421785d (patch) | |
| tree | 7e636bed27130e0b4846a7d41dfcf50e12a83492 /lualibs-unicode.lua | |
| parent | 08563510e0f4328d9e893f17fc1fd37bac00dbc7 (diff) | |
| download | lualibs-d96d015887e6fcbe1ca8defcec44cb794421785d.tar.gz | |
sync with Context as of 2013-09-21
Diffstat (limited to 'lualibs-unicode.lua')
| -rw-r--r-- | lualibs-unicode.lua | 31 |
1 files changed, 24 insertions, 7 deletions
diff --git a/lualibs-unicode.lua b/lualibs-unicode.lua index 3ce5bd3..7ada394 100644 --- a/lualibs-unicode.lua +++ b/lualibs-unicode.lua @@ -38,13 +38,14 @@ local replacer = lpeg.replacer local utfvalues = utf.values local utfgmatch = utf.gmatch -- not always present -local p_utftype = patterns.utftype -local p_utfoffset = patterns.utfoffset -local p_utf8char = patterns.utf8char -local p_utf8byte = patterns.utf8byte -local p_utfbom = patterns.utfbom -local p_newline = patterns.newline -local p_whitespace = patterns.whitespace +local p_utftype = patterns.utftype +local p_utfstricttype = patterns.utfstricttype +local p_utfoffset = patterns.utfoffset +local p_utf8char = patterns.utf8char +local p_utf8byte = patterns.utf8byte +local p_utfbom = patterns.utfbom +local p_newline = patterns.newline +local p_whitespace = patterns.whitespace if not unicode then @@ -976,6 +977,22 @@ function utf.xstring(s) return format("0x%05X",type(s) == "number" and s or utfbyte(s)) end +function utf.toeight(str) + if not str then + return nil + end + local utftype = lpegmatch(p_utfstricttype,str) + if utftype == "utf-8" then + return sub(str,4) + elseif utftype == "utf-16-le" then + return utf16_to_utf8_le(str) + elseif utftype == "utf-16-be" then + return utf16_to_utf8_ne(str) + else + return str + end +end + -- local p_nany = p_utf8char / "" |