! Copyright (C) 2008 Daniel Ehrenberg. ! See http://factorcode.org/license.txt for BSD license. USING: accessors arrays assocs combinators combinators.short-circuit combinators.smart fry kernel make math math.order math.parser namespaces sequences simple-flat-file splitting strings unicode.data ; IN: unicode.collation ( primary secondary tertiary -- weight-levels> ) weight-levels new swap >>tertiary swap >>secondary swap >>primary ; inline : parse-weight ( string -- weight ) "]" split but-last [ weight-levels new swap rest unclip CHAR: * = swapd >>ignorable? swap "." split first3 [ hex> ] tri@ [ >>primary ] [ >>secondary ] [ >>tertiary ] tri* ] map ; : parse-keys ( string -- chars ) split-words [ hex> ] "" map-as ; : parse-ducet ( file -- ducet ) load-data-file [ [ parse-keys ] [ parse-weight ] bi* ] H{ } assoc-map-as ; "vocab:unicode/UCA/allkeys.txt" parse-ducet ducet set-global ! https://www.unicode.org/reports/tr10/tr10-41.html#Well_Formed_DUCET ! WF5 - Well-formedness 5 condition: ! https://www.unicode.org/reports/tr10/tr10-41.html#WF5 ! { "0CC6" "0CC2" "0CD5" } ! 0CD5 is not a non-starter, don't add 2-gram "0CC6" "0CC2"to ducet ! { "0DD9" "0DCF" "0DCA" } ! already in allkeys.txt file ! { "0FB2" "0F71" "0F80" } ! added below ! { "0FB3" "0F71" "0F80" } ! added below ! This breaks the unicode tests that ship in CollationTest_SHIFTED.txt ! but it's supposedly more correct. : fixup-ducet-for-tibetan ( -- ) { { { 0x0FB2 0x0F71 } ! CE(0FB2) CE(0F71) { T{ weight-levels { primary 12719 } { secondary 32 } { tertiary 2 } } T{ weight-levels { primary 12741 } { secondary 32 } { tertiary 2 } } } } { { 0x0FB3 0x0F71 } ! CE(0FB3) CE(0F71) { T{ weight-levels { primary 12722 } { secondary 32 } { tertiary 2 } } T{ weight-levels { primary 12741 } { secondary 32 } { tertiary 2 } } } } { { 0x0FB2 0x0F71 0x0F72 } ! CE(0FB2) CE(0F71 0F72) { T{ weight-levels { primary 12719 } { secondary 32 } { tertiary 2 } } T{ weight-levels { primary 12743 } { secondary 32 } { tertiary 2 } } } } { { 0x0FB2 0x0F73 } ! CE(0FB2) CE(0F71 0F72) { T{ weight-levels { primary 12719 } { secondary 32 } { tertiary 2 } } T{ weight-levels { primary 12743 } { secondary 32 } { tertiary 2 } } } } { { 0x0FB2 0x0F71 0x0F74 } ! CE(0FB2) CE(0F71 0F74) { T{ weight-levels { primary 12719 } { secondary 32 } { tertiary 2 } } T{ weight-levels { primary 12747 } { secondary 32 } { tertiary 2 } } } } { { 0x0FB2 0x0F75 } ! CE(0FB2) CE(0F71 0F74) { T{ weight-levels { primary 12719 } { secondary 32 } { tertiary 2 } } T{ weight-levels { primary 12747 } { secondary 32 } { tertiary 2 } } } } { { 0x0FB3 0x0F71 0x0F72 } ! CE(0FB3) CE(0F71 0F72) { T{ weight-levels { primary 12722 } { secondary 32 } { tertiary 2 } } T{ weight-levels { primary 12743 } { secondary 32 } { tertiary 2 } } } } { { 0x0FB3 0x0F73 } ! CE(0FB3) CE(0F71 0F72) { T{ weight-levels { primary 12722 } { secondary 32 } { tertiary 2 } } T{ weight-levels { primary 12743 } { secondary 32 } { tertiary 2 } } } } { { 0x0FB3 0x0F71 0x0F74 } ! CE(0FB3) CE(0F71 0F74) { T{ weight-levels { primary 12722 } { secondary 32 } { tertiary 2 } } T{ weight-levels { primary 12747 } { secondary 32 } { tertiary 2 } } } } { { 0x0FB3 0x0F75 } ! CE(0FB3) CE(0F71 0F74) { T{ weight-levels { primary 12722 } { secondary 32 } { tertiary 2 } } T{ weight-levels { primary 12747 } { secondary 32 } { tertiary 2 } } } } } ducet get-global '[ swap >string _ set-at ] assoc-each ; ! These values actually break the collation unit tests in CollationTest_SHIFTED.txt ! So we disable those tests in favor of supposedly better collation for Tibetan. ! https://www.unicode.org/reports/tr10/tr10-41.html#Well_Formed_DUCET fixup-ducet-for-tibetan : tangut-block? ( char -- ? ) { [ 0x17000 0x18AFF between? ] ! Tangut and Tangut Components [ 0x18D00 0x18D8F between? ] ! Tangut Supplement } 1|| ; inline : nushu-block? ( char -- ? ) 0x1b170 0x1B2FF between? ; inline : khitan-block? ( char -- ? ) 0x18b00 0x18cff between? ; inline ! https://wiki.computercraft.cc/Module:Unicode_data ! Unicode TR10 - Computing Implicit Weights : base ( char -- base ) { { [ dup 0x03400 0x04DBF between? ] [ drop 0xFB80 ] } ! Extension A { [ dup 0x20000 0x2A6DF between? ] [ drop 0xFB80 ] } ! Extension B { [ dup 0x2A700 0x2B739 between? ] [ drop 0xFB80 ] } ! Extension C { [ dup 0x2B740 0x2B81D between? ] [ drop 0xFB80 ] } ! Extension D { [ dup 0x2B820 0x2CEA1 between? ] [ drop 0xFB80 ] } ! Extension E { [ dup 0x2CEB0 0x2EBE0 between? ] [ drop 0xFB80 ] } ! Extension F { [ dup 0x30000 0x3134A between? ] [ drop 0xFB80 ] } ! Extension G { [ dup 0x31350 0x323AF between? ] [ drop 0xFB80 ] } ! Extension H { [ dup 0x2F800 0x2FA1D between? ] [ drop 0xFB80 ] } ! CJK Compatibility { [ dup 0x04E00 0x09FFF between? ] [ drop 0xFB40 ] } ! CJK { [ dup 0x0F900 0x0FAD9 between? ] [ drop 0xFB40 ] } ! CJK [ drop 0xFBC0 ] ! Other } cond ; : tangut-AAAA ( char -- weight-levels ) drop 0xfb00 0x0020 0x0002 ; inline : tangut-BBBB ( char -- weight-levels ) 0x17000 - 0x8000 bitor 0 0 ; inline : nushu-AAAA ( char -- weight-levels ) drop 0xfb01 0x0020 0x0002 ; inline : nushu-BBBB ( char -- weight-levels ) 0x1B170 - 0x8000 bitor 0 0 ; inline : khitan-AAAA ( char -- weight-levels ) drop 0xfb02 0x0020 0x0002 ; inline : khitan-BBBB ( char -- weight-levels ) 0x18b00 - 0x8000 bitor 0 0 ; inline : AAAA ( char -- weight-levels ) [ base ] [ -15 shift ] bi + 0x0020 0x0002 ; inline : BBBB ( char -- weight-levels ) 0x7FFF bitand 0x8000 bitor 0 0 ; inline : derive-weight ( 1string -- weight-levels-pair ) first { { [ dup tangut-block? ] [ [ tangut-AAAA ] [ tangut-BBBB ] bi 2array ] } { [ dup nushu-block? ] [ [ nushu-AAAA ] [ nushu-BBBB ] bi 2array ] } { [ dup khitan-block? ] [ [ khitan-AAAA ] [ khitan-BBBB ] bi 2array ] } [ [ AAAA ] [ BBBB ] bi 2array ] } cond ; : building-last ( -- char ) building get [ 0 ] [ last last ] if-empty ; ! https://www.unicode.org/reports/tr10/tr10-41.html#Collation_Graphemes : blocked? ( char -- ? ) combining-class dup { 0 f } member? [ drop building-last non-starter? ] [ building-last combining-class = ] if ; : possible-bases ( -- slice-of-building ) building get dup [ first non-starter? not ] find-last drop [ 0 ] unless* tail-slice ; :: ?combine ( char slice i -- ? ) i slice nth char suffix :> str str ducet get-global key? dup [ str i slice set-nth ] when ; : add ( char -- ) dup blocked? [ 1string , ] [ dup possible-bases dup length [ ?combine ] 2with any? [ drop ] [ 1string , ] if ] if ; : string>graphemes ( string -- graphemes ) [ [ add ] each ] { } make ; : char>weight-levels ( 1string -- weight-levels ) ducet get-global ?at [ derive-weight ] unless ; inline : graphemes>weights ( graphemes -- weights ) [ dup weight-levels? [ 1array ] ! From tailoring [ char>weight-levels ] if ] { } map-as concat ; : append-weights ( weight-levels quot -- seq ) [ [ ignorable?>> ] reject ] dip map [ zero? ] reject ; inline : variable-weight ( weight-levels -- obj ) dup ignorable?>> [ primary>> ] [ drop 0xFFFF ] if ; : weights>bytes ( weights -- array ) [ { [ [ primary>> ] append-weights { 0 } ] [ [ secondary>> ] append-weights { 0 } ] [ [ tertiary>> ] append-weights { 0 } ] [ [ [ secondary>> ] [ tertiary>> ] bi [ zero? ] both? ] reject [ variable-weight ] map ] } cleave ] { } append-outputs-as ; PRIVATE> : completely-ignorable? ( weight -- ? ) { [ primary>> zero? ] [ secondary>> zero? ] [ tertiary>> zero? ] } 1&& ; : filter-ignorable ( weights -- weights' ) f swap [ [ nip ] [ primary>> zero? and ] 2bi [ swap ignorable?>> or ] [ swap completely-ignorable? or not ] 2bi ] filter nip ;