! Copyright (C) 2006, 2008 Daniel Ehrenberg. ! See http://factorcode.org/license.txt for BSD license. USING: math kernel sequences sbufs vectors namespaces io.binary io.encodings combinators splitting io byte-arrays ; IN: io.encodings.utf16 SYMBOL: double SYMBOL: quad1 SYMBOL: quad2 SYMBOL: quad3 SYMBOL: ignore : do-ignore ( -- ch state ) 0 ignore ; : append-nums ( byte ch -- ch ) 8 shift bitor ; : end-multibyte ( buf byte ch -- buf ch state ) append-nums push-decoded ; : begin-utf16be ( buf byte -- buf ch state ) dup -3 shift BIN: 11011 number= [ dup BIN: 00000100 bitand zero? [ BIN: 11 bitand quad1 ] [ drop do-ignore ] if ] [ double ] if ; : handle-quad2be ( byte ch -- ch state ) swap dup -2 shift BIN: 110111 number= [ >r 2 shift r> BIN: 11 bitand bitor quad3 ] [ 2drop do-ignore ] if ; : decode-utf16be-step ( buf byte ch state -- buf ch state ) { { begin [ drop begin-utf16be ] } { double [ end-multibyte ] } { quad1 [ append-nums quad2 ] } { quad2 [ handle-quad2be ] } { quad3 [ append-nums HEX: 10000 + push-decoded ] } { ignore [ 2drop push-replacement ] } } case ; : decode-utf16be ( seq -- str ) [ decode-utf16be-step ] decode ; : handle-double ( buf byte ch -- buf ch state ) swap dup -3 shift BIN: 11011 = [ dup BIN: 100 bitand 0 number= [ BIN: 11 bitand 8 shift bitor quad2 ] [ 2drop push-replacement ] if ] [ end-multibyte ] if ; : handle-quad3le ( buf byte ch -- buf ch state ) swap dup -2 shift BIN: 110111 = [ BIN: 11 bitand append-nums HEX: 10000 + push-decoded ] [ 2drop push-replacement ] if ; : decode-utf16le-step ( buf byte ch state -- buf ch state ) { { begin [ drop double ] } { double [ handle-double ] } { quad1 [ append-nums quad2 ] } { quad2 [ 10 shift bitor quad3 ] } { quad3 [ handle-quad3le ] } } case ; : decode-utf16le ( seq -- str ) [ decode-utf16le-step ] decode ; : encode-first -10 shift dup -8 shift BIN: 11011000 bitor swap HEX: FF bitand ; : encode-second BIN: 1111111111 bitand dup -8 shift BIN: 11011100 bitor swap BIN: 11111111 bitand ; : char>utf16be ( char -- ) dup HEX: FFFF > [ HEX: 10000 - dup encode-first swap , , encode-second swap , , ] [ h>b/b , , ] if ; : encode-utf16be ( str -- seq ) [ [ char>utf16be ] each ] B{ } make ; : char>utf16le ( char -- ) dup HEX: FFFF > [ HEX: 10000 - dup encode-first , , encode-second , , ] [ h>b/b swap , , ] if ; : encode-utf16le ( str -- seq ) [ [ char>utf16le ] each ] B{ } make ; : bom-le B{ HEX: ff HEX: fe } ; inline : bom-be B{ HEX: fe HEX: ff } ; inline : encode-utf16 ( str -- seq ) encode-utf16le bom-le swap append ; : start-utf16le? ( seq1 -- seq2 ? ) bom-le ?head ; : start-utf16be? ( seq1 -- seq2 ? ) bom-be ?head ; : decode-utf16 ( seq -- str ) { { [ start-utf16le? ] [ decode-utf16le ] } { [ start-utf16be? ] [ decode-utf16be ] } { [ t ] [ decode-error ] } } cond ; TUPLE: utf16le ; M: utf16le encode-string drop encode-utf16le ; M: utf16le decode-step drop decode-utf16le-step ; TUPLE: utf16be ; M: utf16be encode-string drop encode-utf16be ; M: utf16be decode-step drop decode-utf16be-step ; TUPLE: utf16 encoding ; M: utf16 underlying-stream delegate dup delegate [ ] [ ] ?if ; ! necessary? M: utf16 set-underlying-stream delegate set-delegate ; ! necessary? M: utf16 encode-string >r encode-utf16le r> dup utf16-encoding [ drop ] [ t swap set-utf16-encoding bom-le swap append ] if ; : bom>le/be ( bom -- le/be ) dup bom-le sequence= [ drop utf16le ] [ bom-be sequence= [ utf16be ] [ decode-error ] if ] if ; : read-bom ( utf16 -- encoding ) 2 over delegate stream-read bom>le/be construct-empty [ swap set-utf16-encoding ] keep ; M: utf16 decode-step ! inefficient: checks if bom is done many times ! This should transform itself into utf16be or utf16le after reading BOM dup utf16-encoding [ ] [ read-bom ] ?if decode-step ;