src/test/regress/expected/collate.utf8.out - master - githubmirror/postgres

src/test/regress/expected/collate.utf8.out
352 строки9 KB
Jeff Davis
Fix unintentional behavior change from 5a38104.
08 июл 2026, 04:04
08 июл 2026, 04:04d34ce77
/*
 * This test is for collations and character operations when using the
 * builtin provider with the C.UTF-8 locale.
 */
/* skip test if not UTF8 server encoding */
SELECT getdatabaseencoding() <> 'UTF8' AS skip_test \gset
\if :skip_test
\quit
\endif
SET client_encoding TO UTF8;
--
-- Test builtin "C"
--
CREATE COLLATION regress_builtin_c (
  provider = builtin, locale = 'C');
-- non-ASCII characters are unchanged
SELECT LOWER(U&'\00C1' COLLATE regress_builtin_c) = U&'\00C1';
 ?column? 
----------
 t
(1 row)

SELECT UPPER(U&'\00E1' COLLATE regress_builtin_c) = U&'\00E1';
 ?column? 
----------
 t
(1 row)

-- non-ASCII characters are not alphabetic
SELECT U&'\00C1\00E1' !~ '[[:alpha:]]' COLLATE regress_builtin_c;
 ?column? 
----------
 t
(1 row)

DROP COLLATION regress_builtin_c;
-- a '_' matches a multibyte character
SELECT 'café' ILIKE 'caf_' COLLATE "C";
 ?column? 
----------
 t
(1 row)

--
-- Test PG_C_UTF8
--
CREATE COLLATION regress_pg_c_utf8 (
  provider = builtin, locale = 'C_UTF8'); -- fails
ERROR:  invalid locale name "C_UTF8" for builtin provider
CREATE COLLATION regress_pg_c_utf8 (
  provider = builtin, locale = 'C.UTF8');
DROP COLLATION regress_pg_c_utf8;
CREATE COLLATION regress_pg_c_utf8 (
  provider = builtin, locale = 'C.UTF-8');
CREATE TABLE test_pg_c_utf8 (
  t TEXT COLLATE PG_C_UTF8
);
INSERT INTO test_pg_c_utf8 VALUES
  ('abc DEF 123abc'),
  ('ábc sßs ßss DÉF'),
  ('ǄxxǄ ǆxxǅ ǅxxǆ'),
  (U&'Λλ 1a \FF11a'),
  ('ȺȺȺ'),
  ('ⱥⱥⱥ'),
  ('ⱥȺ');
SELECT
    t, lower(t), initcap(t), upper(t),
    length(convert_to(t, 'UTF8')) AS t_bytes,
    length(convert_to(lower(t), 'UTF8')) AS lower_t_bytes,
    length(convert_to(initcap(t), 'UTF8')) AS initcap_t_bytes,
    length(convert_to(upper(t), 'UTF8')) AS upper_t_bytes
  FROM test_pg_c_utf8;
        t        |      lower      |     initcap     |      upper      | t_bytes | lower_t_bytes | initcap_t_bytes | upper_t_bytes 
-----------------+-----------------+-----------------+-----------------+---------+---------------+-----------------+---------------
 abc DEF 123abc  | abc def 123abc  | Abc Def 123abc  | ABC DEF 123ABC  |      14 |            14 |              14 |            14
 ábc sßs ßss DÉF | ábc sßs ßss déf | Ábc Sßs ßss Déf | ÁBC SßS ßSS DÉF |      19 |            19 |              19 |            19
 ǄxxǄ ǆxxǅ ǅxxǆ  | ǆxxǆ ǆxxǆ ǆxxǆ  | Ǆxxǆ Ǆxxǆ Ǆxxǆ  | ǄXXǄ ǄXXǄ ǄXXǄ  |      20 |            20 |              20 |            20
 Λλ 1a １a       | λλ 1a １a       | Λλ 1a １A       | ΛΛ 1A １A       |      12 |            12 |              12 |            12
 ȺȺȺ             | ⱥⱥⱥ             | Ⱥⱥⱥ             | ȺȺȺ             |       6 |             9 |               8 |             6
 ⱥⱥⱥ             | ⱥⱥⱥ             | Ⱥⱥⱥ             | ȺȺȺ             |       9 |             9 |               8 |             6
 ⱥȺ              | ⱥⱥ              | Ⱥⱥ              | ȺȺ              |       5 |             6 |               5 |             4
(7 rows)

DROP TABLE test_pg_c_utf8;
-- negative test: Final_Sigma not used for builtin locale C.UTF-8
SELECT lower('ΑΣ' COLLATE PG_C_UTF8);
 lower 
-------
 ασ
(1 row)

SELECT lower('ΑͺΣͺ' COLLATE PG_C_UTF8);
 lower 
-------
 αͺσͺ
(1 row)

SELECT lower('Α΄Σ΄' COLLATE PG_C_UTF8);
 lower 
-------
 α΄σ΄
(1 row)

-- properties
SELECT 'xyz' ~ '[[:alnum:]]' COLLATE PG_C_UTF8;
 ?column? 
----------
 t
(1 row)

SELECT 'xyz' !~ '[[:upper:]]' COLLATE PG_C_UTF8;
 ?column? 
----------
 t
(1 row)

SELECT '@' !~ '[[:alnum:]]' COLLATE PG_C_UTF8;
 ?column? 
----------
 t
(1 row)

SELECT '=' ~ '[[:punct:]]' COLLATE PG_C_UTF8; -- symbols are punctuation in posix
 ?column? 
----------
 t
(1 row)

SELECT 'a8a' ~ '[[:digit:]]' COLLATE PG_C_UTF8;
 ?column? 
----------
 t
(1 row)

SELECT '൧' !~ '\d' COLLATE PG_C_UTF8; -- only 0-9 considered digits in posix
 ?column? 
----------
 t
(1 row)

-- case mapping
SELECT 'xYz' ~* 'XyZ' COLLATE PG_C_UTF8;
 ?column? 
----------
 t
(1 row)

SELECT 'xAb' ~* '[W-Y]' COLLATE PG_C_UTF8;
 ?column? 
----------
 t
(1 row)

SELECT 'xAb' !~* '[c-d]' COLLATE PG_C_UTF8;
 ?column? 
----------
 t
(1 row)

SELECT 'Δ' ~* '[γ-λ]' COLLATE PG_C_UTF8;
 ?column? 
----------
 t
(1 row)

SELECT 'δ' ~* '[Γ-Λ]' COLLATE PG_C_UTF8; -- same as above with cases reversed
 ?column? 
----------
 t
(1 row)

-- case folding
select casefold('AbCd 123 #$% ıiIİ ẞ ß Ǆǅǆ Σσς' collate PG_C_UTF8);
           casefold            
-------------------------------
 abcd 123 #$% ıiiİ ß ß ǆǆǆ σσσ
(1 row)

--
-- Test PG_UNICODE_FAST
--
CREATE COLLATION regress_pg_unicode_fast (
  provider = builtin, locale = 'unicode'); -- fails
ERROR:  invalid locale name "unicode" for builtin provider
CREATE COLLATION regress_pg_unicode_fast (
  provider = builtin, locale = 'PG_UNICODE_FAST');
CREATE TABLE test_pg_unicode_fast (
  t TEXT COLLATE PG_UNICODE_FAST
);
INSERT INTO test_pg_unicode_fast VALUES
  ('abc DEF 123abc'),
  ('ábc sßs ßss DÉF'),
  ('ǄxxǄ ǆxxǅ ǅxxǆ'),
  (U&'Λλ 1a \FF11a'),
  ('ȺȺȺ'),
  ('ⱥⱥⱥ'),
  ('ⱥȺ');
SELECT
    t, lower(t), initcap(t), upper(t),
    length(convert_to(t, 'UTF8')) AS t_bytes,
    length(convert_to(lower(t), 'UTF8')) AS lower_t_bytes,
    length(convert_to(initcap(t), 'UTF8')) AS initcap_t_bytes,
    length(convert_to(upper(t), 'UTF8')) AS upper_t_bytes
  FROM test_pg_unicode_fast;
        t        |      lower      |     initcap      |       upper       | t_bytes | lower_t_bytes | initcap_t_bytes | upper_t_bytes 
-----------------+-----------------+------------------+-------------------+---------+---------------+-----------------+---------------
 abc DEF 123abc  | abc def 123abc  | Abc Def 123abc   | ABC DEF 123ABC    |      14 |            14 |              14 |            14
 ábc sßs ßss DÉF | ábc sßs ßss déf | Ábc Sßs Ssss Déf | ÁBC SSSS SSSS DÉF |      19 |            19 |              19 |            19
 ǄxxǄ ǆxxǅ ǅxxǆ  | ǆxxǆ ǆxxǆ ǆxxǆ  | ǅxxǆ ǅxxǆ ǅxxǆ   | ǄXXǄ ǄXXǄ ǄXXǄ    |      20 |            20 |              20 |            20
 Λλ 1a １a       | λλ 1a １a       | Λλ 1a １a        | ΛΛ 1A １A         |      12 |            12 |              12 |            12
 ȺȺȺ             | ⱥⱥⱥ             | Ⱥⱥⱥ              | ȺȺȺ               |       6 |             9 |               8 |             6
 ⱥⱥⱥ             | ⱥⱥⱥ             | Ⱥⱥⱥ              | ȺȺȺ               |       9 |             9 |               8 |             6
 ⱥȺ              | ⱥⱥ              | Ⱥⱥ               | ȺȺ                |       5 |             6 |               5 |             4
(7 rows)

DROP TABLE test_pg_unicode_fast;
-- test Final_Sigma
SELECT lower('ΑΣ' COLLATE PG_UNICODE_FAST); -- 0391 03A3
 lower 
-------
 ας
(1 row)

SELECT lower('ΑΣ0' COLLATE PG_UNICODE_FAST); -- 0391 03A3 0030
 lower 
-------
 ας0
(1 row)

SELECT lower('ἈΣ̓' COLLATE PG_UNICODE_FAST); -- 0391 0343 03A3 0343
 lower 
-------
 ἀς̓
(1 row)

SELECT lower('ᾼΣͅ' COLLATE PG_UNICODE_FAST); -- 0391 0345 03A3 0345
 lower 
-------
 ᾳςͅ
(1 row)

-- test !Final_Sigma
SELECT lower('Σ' COLLATE PG_UNICODE_FAST); -- 03A3
 lower 
-------
 σ
(1 row)

SELECT lower('0Σ' COLLATE PG_UNICODE_FAST); -- 0030 03A3
 lower 
-------
 0σ
(1 row)

SELECT lower('ΑΣΑ' COLLATE PG_UNICODE_FAST); -- 0391 03A3 0391
 lower 
-------
 ασα
(1 row)

SELECT lower('ἈΣ̓Α' COLLATE PG_UNICODE_FAST); -- 0391 0343 03A3 0343 0391
 lower 
-------
 ἀσ̓α
(1 row)

SELECT lower('ᾼΣͅΑ' COLLATE PG_UNICODE_FAST); -- 0391 0345 03A3 0345 0391
 lower 
-------
 ᾳσͅα
(1 row)

SELECT lower(U&'\0300\03A3' COLLATE PG_UNICODE_FAST);
 lower 
-------
 ̀σ
(1 row)

-- properties
SELECT 'xyz' ~ '[[:alnum:]]' COLLATE PG_UNICODE_FAST;
 ?column? 
----------
 t
(1 row)

SELECT 'xyz' !~ '[[:upper:]]' COLLATE PG_UNICODE_FAST;
 ?column? 
----------
 t
(1 row)

SELECT '@' !~ '[[:alnum:]]' COLLATE PG_UNICODE_FAST;
 ?column? 
----------
 t
(1 row)

SELECT '=' !~ '[[:punct:]]' COLLATE PG_UNICODE_FAST; -- symbols are not punctuation
 ?column? 
----------
 t
(1 row)

SELECT 'a8a' ~ '[[:digit:]]' COLLATE PG_UNICODE_FAST;
 ?column? 
----------
 t
(1 row)

SELECT '൧' ~ '\d' COLLATE PG_UNICODE_FAST;
 ?column? 
----------
 t
(1 row)

-- case mapping
SELECT 'xYz' ~* 'XyZ' COLLATE PG_UNICODE_FAST;
 ?column? 
----------
 t
(1 row)

SELECT 'xAb' ~* '[W-Y]' COLLATE PG_UNICODE_FAST;
 ?column? 
----------
 t
(1 row)

SELECT 'xAb' !~* '[c-d]' COLLATE PG_UNICODE_FAST;
 ?column? 
----------
 t
(1 row)

SELECT 'Δ' ~* '[γ-λ]' COLLATE PG_UNICODE_FAST;
 ?column? 
----------
 t
(1 row)

SELECT 'δ' ~* '[Γ-Λ]' COLLATE PG_UNICODE_FAST; -- same as above with cases reversed
 ?column? 
----------
 t
(1 row)

-- case folding
select casefold('AbCd 123 #$% ıiIİ ẞ ß Ǆǅǆ Σσς' collate PG_UNICODE_FAST);
            casefold             
---------------------------------
 abcd 123 #$% ıiii̇ ss ss ǆǆǆ σσσ
(1 row)