1 USING: arrays assocs fry grouping hash-sets io.encodings.utf8
2 io.files kernel math math.order math.parser sequences sets
3 splitting strings tools.test unicode ;
4 IN: unicode.collation.tests
6 : test-equality ( str1 str2 -- ? ? ? ? )
7 { primary= secondary= tertiary= quaternary= }
8 [ execute( a b -- ? ) ] 2with map
11 { f f f f } [ "hello" "hi" test-equality ] unit-test
12 { t f f f } [ "hello" "h\u0000e9llo" test-equality ] unit-test
13 { t t f f } [ "hello" "HELLO" test-equality ] unit-test
14 { t t t f } [ "hello" "h e l l o." test-equality ] unit-test
15 { t t t t } [ "hello" "\0hello\0" test-equality ] unit-test
16 { { "good bye" "goodbye" "hello" "HELLO" } }
17 [ { "HELLO" "goodbye" "good bye" "hello" } sort-strings ] unit-test
19 : collation-test-lines ( -- lines )
20 "vocab:unicode/UCA/CollationTest/CollationTest_SHIFTED.txt" utf8 file-lines
21 [ "#" head? ] reject harvest ;
23 : parse-collation-test-shifted ( -- lines )
25 [ ";" split first " " split [ hex> ] "" map-as ] map ;
27 : tail-from-last ( string char -- string' )
28 '[ _ = ] dupd find-last drop 1 + tail ; inline
30 : line>test-weights ( string -- pair )
32 " " split [ hex> ] map
34 "#" split1 nip CHAR: [ tail-from-last
37 [ " " split harvest [ hex> ] map ] map
40 ! These tests actually would pass if I didn't fix up
41 ! the ducet table for Tibetan. It took me way too long to realize
42 ! that the Unicode committee recommends fixing Tibetan collation
43 ! yet ships tests that collation fails if you fix it.
44 ! (Specifically the ducet entries for { 0x0FB2 0x0F71 } and { 0x0FB3 0x0F71 }
45 ! cause these tests to fail)
46 : xfailed-collation-tests ( -- seq )
50 { 4018 820 3953 3968 }
51 { 4018 820 3968 3953 }
52 { 4018 3953 1 3968 97 }
56 { 4019 820 3953 3968 }
57 { 4019 820 3968 3953 }
58 { 4019 3953 820 3968 }
59 { 4019 3953 1 3968 97 }
62 : parse-collation-test-weights ( -- weights )
64 [ line>test-weights ] map
65 [ first xfailed-collation-tests in? ] reject ;
67 : calculate-collation ( chars collation -- collation-calculated collation-answer )
68 [ >string collation-key/nfd drop ] [ { 0 } join ] bi* ;
70 : find-bad-collations ( pairs -- seq )
71 [ first2 calculate-collation sequence= ] reject ;
74 [ parse-collation-test-weights find-bad-collations ] unit-test
77 parse-collation-test-shifted
80 ! Remove these two expected-fail Tibetan collation comparison tests
81 ! They are bad tests once you fix up the ducet table with { 0x0FB2 0x0F71 } and { 0x0FB3 0x0F71 }
83 { { 4018 820 3969 } { 3959 33 } }
84 { { 4019 3953 820 3968 } { 3961 33 } }
85 { { 4019 98 } { 4019 3953 1 3968 97 } }
86 { { 4028 98 } { 4018 3953 1 3968 97 } }
87 } [ [ >string ] bi@ ] assoc-map >hash-set diff members
89 [ string<=> { +lt+ +eq+ } member? ] assoc-reject
92 ! XXX: Once again, these tests pass if you don't
93 ! fix up the ducet table for { 0x0FB2 0x0F71 } and { 0x0FB3 0x0F71 }
94 ! { +lt+ } [ { 4018 820 3969 } { 3959 33 } [ >string ] bi@ string<=> ] unit-test
95 ! { +lt+ } [ { 4019 3953 820 3968 } { 3961 33 } [ >string ] bi@ string<=> ] unit-test