Skip to content

MDEV-39520: Add MySQL 8.0 extended syntax for REGEXP_INSTR #5035

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
MohamedM216 wants to merge 1 commit into
base: main
Choose a base branch
from
Open

MDEV-39520: Add MySQL 8.0 extended syntax for REGEXP_INSTR #5035

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
277 changes: 277 additions & 0 deletions mysql-test/main/regexp_instr_mysql8.result
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,277 @@
#
# MDEV-39520: Improve REGEXP_INSTR for MySQL 8.0 compatibility
#
# 1. Basic 2-argument form
SELECT REGEXP_INSTR('abba', 'b{2}');
REGEXP_INSTR('abba', 'b{2}')
2
SELECT REGEXP_INSTR('abba', 'x');
REGEXP_INSTR('abba', 'x')
0
SELECT REGEXP_INSTR('hello world', 'world');
REGEXP_INSTR('hello world', 'world')
7
SELECT REGEXP_INSTR('hello', '');
REGEXP_INSTR('hello', '')
1
SELECT REGEXP_INSTR('', 'a');
REGEXP_INSTR('', 'a')
0
SELECT REGEXP_INSTR('', '');
REGEXP_INSTR('', '')
1
# 2. Three arguments: pos
SELECT REGEXP_INSTR('abba', 'b{2}', 2);
REGEXP_INSTR('abba', 'b{2}', 2)
2
SELECT REGEXP_INSTR('abba', 'b{2}', 3);
REGEXP_INSTR('abba', 'b{2}', 3)
0
SELECT REGEXP_INSTR('abbabba', 'b{2}', 1);
REGEXP_INSTR('abbabba', 'b{2}', 1)
2
SELECT REGEXP_INSTR('aabba', 'b', 3);
REGEXP_INSTR('aabba', 'b', 3)
3
SELECT REGEXP_INSTR('xyzabc', 'abc', 4);
REGEXP_INSTR('xyzabc', 'abc', 4)
4
SELECT REGEXP_INSTR('abc', 'c', 3);
REGEXP_INSTR('abc', 'c', 3)
3
SELECT REGEXP_INSTR('abc', 'c', 4);
REGEXP_INSTR('abc', 'c', 4)
0
# 3. Four arguments: occurrence
SELECT REGEXP_INSTR('abbabba', 'b{2}', 1, 1);
REGEXP_INSTR('abbabba', 'b{2}', 1, 1)
2
SELECT REGEXP_INSTR('abbabba', 'b{2}', 1, 2);
REGEXP_INSTR('abbabba', 'b{2}', 1, 2)
5
SELECT REGEXP_INSTR('abbabba', 'b{2}', 1, 3);
REGEXP_INSTR('abbabba', 'b{2}', 1, 3)
0
SELECT REGEXP_INSTR('abbabba', 'b{2}', 2, 1);
REGEXP_INSTR('abbabba', 'b{2}', 2, 1)
2
SELECT REGEXP_INSTR('abbabba', 'b{2}', 2, 2);
REGEXP_INSTR('abbabba', 'b{2}', 2, 2)
5
SELECT REGEXP_INSTR('abbabba', 'b{2}', 3, 2);
REGEXP_INSTR('abbabba', 'b{2}', 3, 2)
0
SELECT REGEXP_INSTR('aaa', 'a', 1, 1);
REGEXP_INSTR('aaa', 'a', 1, 1)
1
SELECT REGEXP_INSTR('aaa', 'a', 1, 2);
REGEXP_INSTR('aaa', 'a', 1, 2)
2
SELECT REGEXP_INSTR('aaa', 'a', 1, 3);
REGEXP_INSTR('aaa', 'a', 1, 3)
3
SELECT REGEXP_INSTR('aaa', 'a', 1, 4);
REGEXP_INSTR('aaa', 'a', 1, 4)
0
# 4. Five arguments: return_option
SELECT REGEXP_INSTR('abbabba', 'b{2}', 1, 1, 0);
REGEXP_INSTR('abbabba', 'b{2}', 1, 1, 0)
2
SELECT REGEXP_INSTR('abbabba', 'b{2}', 1, 1, 1);
REGEXP_INSTR('abbabba', 'b{2}', 1, 1, 1)
4
SELECT REGEXP_INSTR('abbabba', 'b{2}', 1, 2, 0);
REGEXP_INSTR('abbabba', 'b{2}', 1, 2, 0)
5
SELECT REGEXP_INSTR('abbabba', 'b{2}', 1, 2, 1);
REGEXP_INSTR('abbabba', 'b{2}', 1, 2, 1)
7
SELECT REGEXP_INSTR('abcabc', 'b', 1, 1, 0);
REGEXP_INSTR('abcabc', 'b', 1, 1, 0)
2
SELECT REGEXP_INSTR('abcabc', 'b', 1, 1, 1);
REGEXP_INSTR('abcabc', 'b', 1, 1, 1)
3
SELECT REGEXP_INSTR('abcabc', 'b', 1, 2, 0);
REGEXP_INSTR('abcabc', 'b', 1, 2, 0)
5
SELECT REGEXP_INSTR('abcabc', 'b', 1, 2, 1);
REGEXP_INSTR('abcabc', 'b', 1, 2, 1)
6
SELECT REGEXP_INSTR('abcabc', 'z', 1, 1, 1);
REGEXP_INSTR('abcabc', 'z', 1, 1, 1)
0
# 5. Six arguments: match_type
SELECT REGEXP_INSTR('Abba', 'ABBA', 1, 1, 0, 'i');
REGEXP_INSTR('Abba', 'ABBA', 1, 1, 0, 'i')
1
SELECT REGEXP_INSTR('Abba', 'ABBA', 1, 1, 0, 'c');
REGEXP_INSTR('Abba', 'ABBA', 1, 1, 0, 'c')
0
SELECT REGEXP_INSTR('Abba', 'ABBA', 1, 1, 0, 'ci');
REGEXP_INSTR('Abba', 'ABBA', 1, 1, 0, 'ci')
1
SELECT REGEXP_INSTR('Abba', 'ABBA', 1, 1, 0, 'ic');
REGEXP_INSTR('Abba', 'ABBA', 1, 1, 0, 'ic')
0
SELECT REGEXP_INSTR('a\nb\nc', '^b$', 1, 1, 0, 'm');
REGEXP_INSTR('a\nb\nc', '^b$', 1, 1, 0, 'm')
3
SELECT REGEXP_INSTR('a\nb\nc', '^b$', 1, 1, 0, '');
REGEXP_INSTR('a\nb\nc', '^b$', 1, 1, 0, '')
0
SELECT REGEXP_INSTR('a\nb\nc', 'a.b', 1, 1, 0, 'n');
REGEXP_INSTR('a\nb\nc', 'a.b', 1, 1, 0, 'n')
1
SELECT REGEXP_INSTR('a\nb\nc', 'a.b', 1, 1, 0, '');
REGEXP_INSTR('a\nb\nc', 'a.b', 1, 1, 0, '')
0
SELECT REGEXP_INSTR('a\nb', '^b$', 1, 1, 0, 'mu');
REGEXP_INSTR('a\nb', '^b$', 1, 1, 0, 'mu')
3
SELECT REGEXP_INSTR('a\nB\nc', '^b$', 1, 1, 0, 'im');
REGEXP_INSTR('a\nB\nc', '^b$', 1, 1, 0, 'im')
3
# 6. Multibyte characters
SET NAMES utf8mb4;
SELECT REGEXP_INSTR('áéí', 'é');
REGEXP_INSTR('áéí', 'é')
2
SELECT REGEXP_INSTR('áéí', 'í');
REGEXP_INSTR('áéí', 'í')
3
SELECT REGEXP_INSTR('αβγδ', 'γ');
REGEXP_INSTR('αβγδ', 'γ')
3
SELECT REGEXP_INSTR('áéíó', 'í', 2);
REGEXP_INSTR('áéíó', 'í', 2)
3
SELECT REGEXP_INSTR('αβγδ', 'β', 2);
REGEXP_INSTR('αβγδ', 'β', 2)
2
SELECT REGEXP_INSTR('αβγδ', 'β', 1, 1, 1);
REGEXP_INSTR('αβγδ', 'β', 1, 1, 1)
3
# 7. NULL propagation
SELECT REGEXP_INSTR(NULL, 'a');
REGEXP_INSTR(NULL, 'a')
NULL
SELECT REGEXP_INSTR('abc', NULL);
REGEXP_INSTR('abc', NULL)
NULL
SELECT REGEXP_INSTR('abc', 'a', NULL);
REGEXP_INSTR('abc', 'a', NULL)
NULL
SELECT REGEXP_INSTR('abc', 'a', 1, NULL);
REGEXP_INSTR('abc', 'a', 1, NULL)
NULL
SELECT REGEXP_INSTR('abc', 'a', 1, 1, NULL);
REGEXP_INSTR('abc', 'a', 1, 1, NULL)
NULL
SELECT REGEXP_INSTR('abc', 'a', 1, 1, 0, NULL);
REGEXP_INSTR('abc', 'a', 1, 1, 0, NULL)
NULL
# 8. Edge cases
SELECT REGEXP_INSTR('abc', '', 1, 1);
REGEXP_INSTR('abc', '', 1, 1)
1
SELECT REGEXP_INSTR('abc', '', 1, 2);
REGEXP_INSTR('abc', '', 1, 2)
2
SELECT REGEXP_INSTR('abc', '', 1, 3);
REGEXP_INSTR('abc', '', 1, 3)
3
SELECT REGEXP_INSTR('abc', '', 1, 4);
REGEXP_INSTR('abc', '', 1, 4)
4
SELECT REGEXP_INSTR('abc', '', 1, 5);
REGEXP_INSTR('abc', '', 1, 5)
0
SELECT REGEXP_INSTR('abcabc', '^abc');
REGEXP_INSTR('abcabc', '^abc')
1
SELECT REGEXP_INSTR('abcabc', 'abc$');
REGEXP_INSTR('abcabc', 'abc$')
4
SELECT REGEXP_INSTR('abcabc', '^abc$');
REGEXP_INSTR('abcabc', '^abc$')
0
SELECT REGEXP_INSTR('foo bar baz', 'bar|baz', 1, 1);
REGEXP_INSTR('foo bar baz', 'bar|baz', 1, 1)
5
SELECT REGEXP_INSTR('foo bar baz', 'bar|baz', 1, 2);
REGEXP_INSTR('foo bar baz', 'bar|baz', 1, 2)
9
SELECT REGEXP_INSTR('aabbaabb', '(a+)(b+)\\1', 1, 1);
REGEXP_INSTR('aabbaabb', '(a+)(b+)\\1', 1, 1)
1
SELECT REGEXP_INSTR('aabbaabb', '(a+)(b+)\\1', 1, 2);
REGEXP_INSTR('aabbaabb', '(a+)(b+)\\1', 1, 2)
0
SELECT REGEXP_INSTR(REPEAT('x', 1000), 'x{5}', 1, 1);
REGEXP_INSTR(REPEAT('x', 1000), 'x{5}', 1, 1)
1
SELECT REGEXP_INSTR(REPEAT('x', 1000), 'x{5}', 1, 200);
REGEXP_INSTR(REPEAT('x', 1000), 'x{5}', 1, 200)
996
SELECT REGEXP_INSTR('Test-abc-abc-abc', 'AB', 1, 3, 0);
REGEXP_INSTR('Test-abc-abc-abc', 'AB', 1, 3, 0)
14
SELECT REGEXP_INSTR('Test-abc-abc-abc', 'AB', 1, 3, 0, 'c');
REGEXP_INSTR('Test-abc-abc-abc', 'AB', 1, 3, 0, 'c')
0
# Non-constant match_type with constant pattern
CREATE TABLE foo (cond VARCHAR(50));
INSERT INTO foo VALUES ('c'), ('ic');
SELECT REGEXP_INSTR('Abba', 'ABBA', 1, 1, 0, cond) FROM foo;
REGEXP_INSTR('Abba', 'ABBA', 1, 1, 0, cond)
0
0
DROP TABLE foo;
# Prepared statements with changing parameters
PREPARE stmt FROM
"SELECT REGEXP_INSTR(?, ?, 1, ?)";
SET @subject = 'abc abc abc';
SET @pattern = 'abc';
SET @occ = 1;
EXECUTE stmt USING @subject, @pattern, @occ;
REGEXP_INSTR(?, ?, 1, ?)
1
SET @occ = 2;
EXECUTE stmt USING @subject, @pattern, @occ;
REGEXP_INSTR(?, ?, 1, ?)
5
PREPARE stmt FROM
"SELECT REGEXP_INSTR(?, ?, 1, 1, 0, ?)";
SET @subject = 'Abc abc';
SET @pattern = 'abc';
SET @match = 'c';
EXECUTE stmt USING @subject, @pattern, @match;
REGEXP_INSTR(?, ?, 1, 1, 0, ?)
5
SET @match = 'i';
EXECUTE stmt USING @subject, @pattern, @match;
REGEXP_INSTR(?, ?, 1, 1, 0, ?)
1
# match_type from 1-row table
CREATE TABLE t1 (cond VARCHAR(10));
INSERT INTO t1 VALUES ('c');
SELECT REGEXP_INSTR('Abba', 'ABBA', 1, 1, 0, cond) FROM t1;
REGEXP_INSTR('Abba', 'ABBA', 1, 1, 0, cond)
0
DROP TABLE t1;
CREATE TABLE t1 (
subject VARCHAR(20),
pattern VARCHAR(20),
mt VARCHAR(10)
);
INSERT INTO t1 VALUES ('Abc abc', 'abc', 'c');
SELECT REGEXP_INSTR(subject, pattern, 1, 1, 0, mt) FROM t1;
REGEXP_INSTR(subject, pattern, 1, 1, 0, mt)
5
UPDATE t1 SET mt='i';
SELECT REGEXP_INSTR(subject, pattern, 1, 1, 0, mt) FROM t1;
REGEXP_INSTR(subject, pattern, 1, 1, 0, mt)
1
DROP TABLE t1;
End of 13.0 tests
Loading
Loading