59 lines
1.5 KiB
Factor
59 lines
1.5 KiB
Factor
! 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.utf8 ;
|
|
IN: io.encodings.iana
|
|
|
|
<PRIVATE
|
|
SYMBOL: n>e-table
|
|
SYMBOL: e>n-table
|
|
SYMBOL: aliases
|
|
PRIVATE>
|
|
|
|
ERROR: missing-encoding name ;
|
|
|
|
: name>encoding ( name -- encoding )
|
|
dup n>e-table get-global at [ ] [ missing-encoding ] ?if ;
|
|
|
|
ERROR: missing-name encoding ;
|
|
|
|
: encoding>name ( encoding -- name )
|
|
dup e>n-table get-global at [ ] [ missing-name ] ?if ;
|
|
|
|
<PRIVATE
|
|
: parse-iana ( stream -- synonym-set )
|
|
lines { "" } split [
|
|
[ " " split ] map
|
|
[ first { "Name:" "Alias:" } member? ] filter
|
|
[ second ] map { "None" } diff
|
|
] map harvest ;
|
|
|
|
: make-aliases ( stream -- n>e )
|
|
parse-iana [ [ first ] [ ] bi ] H{ } map>assoc ;
|
|
|
|
: initial-n>e ( -- assoc )
|
|
H{
|
|
{ "UTF8" utf8 }
|
|
{ "utf8" utf8 }
|
|
{ "utf-8" utf8 }
|
|
{ "UTF-8" utf8 }
|
|
} clone ;
|
|
|
|
: initial-e>n ( -- assoc )
|
|
H{ { utf8 "UTF-8" } } clone ;
|
|
|
|
PRIVATE>
|
|
|
|
"resource:basis/io/encodings/iana/character-sets"
|
|
utf8 <file-reader> make-aliases aliases set-global
|
|
|
|
n>e-table global [ initial-n>e or ] change-at
|
|
e>n-table global [ initial-e>n or ] change-at
|
|
|
|
: 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 ;
|