factor/basis/io/encodings/iana/iana.factor

61 lines
1.5 KiB
Factor
Raw Normal View History

2008-04-30 23:06:13 -04:00
! Copyright (C) 2008 Daniel Ehrenberg
! See http://factorcode.org/license.txt for BSD license.
USING: kernel strings values io.files assocs
splitting sequences io namespaces sets
io.encodings.ascii io.encodings.utf8 io.encodings.utf16 ;
2008-04-30 22:04:57 -04:00
IN: io.encodings.iana
2008-04-30 23:06:13 -04:00
<PRIVATE
2009-02-03 18:32:05 -05:00
SYMBOL: n>e-table
SYMBOL: e>n-table
SYMBOL: aliases
2008-04-30 23:06:13 -04:00
PRIVATE>
2009-03-20 03:20:31 -04:00
: name>encoding ( name -- encoding )
n>e-table get-global at ;
2009-02-03 18:32:05 -05:00
2009-03-20 03:20:31 -04:00
: encoding>name ( encoding -- name )
e>n-table get-global at ;
2008-04-30 23:06:13 -04:00
<PRIVATE
: parse-iana ( file -- synonym-set )
utf8 file-lines { "" } split [
2008-04-30 22:04:57 -04:00
[ " " split ] map
[ first { "Name:" "Alias:" } member? ] filter
2010-05-17 23:20:46 -04:00
values { "None" } diff
2009-02-03 18:32:05 -05:00
] map harvest ;
: make-aliases ( file -- n>e )
2009-02-03 18:32:05 -05:00
parse-iana [ [ first ] [ ] bi ] H{ } map>assoc ;
2008-04-30 22:04:57 -04:00
2009-02-03 18:32:05 -05:00
: initial-n>e ( -- assoc )
H{
{ "UTF8" utf8 }
{ "utf8" utf8 }
{ "utf-8" utf8 }
2009-02-03 18:32:05 -05:00
{ "UTF-8" utf8 }
} clone ;
: initial-e>n ( -- assoc )
H{ { utf8 "UTF-8" } } clone ;
2008-04-30 23:06:13 -04:00
PRIVATE>
2009-02-15 20:53:21 -05:00
"vocab:io/encodings/iana/character-sets"
make-aliases aliases set-global
2009-02-03 18:32:05 -05:00
n>e-table [ initial-n>e ] initialize
e>n-table [ initial-e>n ] initialize
2009-02-03 18:32:05 -05:00
: register-encoding ( descriptor name -- )
[
aliases get at [
[ n>e-table get-global set-at ] with each
] [ "Bad encoding registration" throw ] if*
] [ swap e>n-table get-global set-at ] 2bi ;
ascii "ANSI_X3.4-1968" register-encoding
utf16be "UTF-16BE" register-encoding
utf16le "UTF-16LE" register-encoding
2009-11-12 17:38:21 -05:00
utf16 "UTF-16" register-encoding