aboutsummaryrefslogtreecommitdiff
path: root/src/test/modules/test_regex
diff options
context:
space:
mode:
Diffstat (limited to 'src/test/modules/test_regex')
-rw-r--r--src/test/modules/test_regex/expected/test_regex.out120
-rw-r--r--src/test/modules/test_regex/expected/test_regex_utf8.out106
-rw-r--r--src/test/modules/test_regex/sql/test_regex.sql32
-rw-r--r--src/test/modules/test_regex/sql/test_regex_utf8.sql18
4 files changed, 276 insertions, 0 deletions
diff --git a/src/test/modules/test_regex/expected/test_regex.out b/src/test/modules/test_regex/expected/test_regex.out
index 5a6cdf47c2f..6242d0baa9a 100644
--- a/src/test/modules/test_regex/expected/test_regex.out
+++ b/src/test/modules/test_regex/expected/test_regex.out
@@ -937,6 +937,34 @@ select * from test_regex('a[[=x=]]', 'az', '+Lb');
{0,REG_ULOCALE}
(1 row)
+-- expectMatch 9.9b &iL {a[[=Y=]]} ay ay
+select * from test_regex('a[[=Y=]]', 'ay', 'iL');
+ test_regex
+-----------------
+ {0,REG_ULOCALE}
+ {ay}
+(2 rows)
+
+select * from test_regex('a[[=Y=]]', 'ay', 'iLb');
+ test_regex
+-----------------
+ {0,REG_ULOCALE}
+ {ay}
+(2 rows)
+
+-- expectNomatch 9.9c &L {a[[=Y=]]} ay
+select * from test_regex('a[[=Y=]]', 'ay', 'L');
+ test_regex
+-----------------
+ {0,REG_ULOCALE}
+(1 row)
+
+select * from test_regex('a[[=Y=]]', 'ay', 'Lb');
+ test_regex
+-----------------
+ {0,REG_ULOCALE}
+(1 row)
+
-- expectError 9.10 & {a[0-[=x=]]} ERANGE
select * from test_regex('a[0-[=x=]]', '', '');
ERROR: invalid regular expression: invalid character range
@@ -2932,6 +2960,34 @@ select * from test_regex('a[^b-d]', 'aC', 'iMb');
{0,REG_UUNPORT}
(1 row)
+-- expectMatch 19.6 &iM {a[B-Z]} aC aC
+select * from test_regex('a[B-Z]', 'aC', 'iM');
+ test_regex
+-----------------
+ {0,REG_UUNPORT}
+ {aC}
+(2 rows)
+
+select * from test_regex('a[B-Z]', 'aC', 'iMb');
+ test_regex
+-----------------
+ {0,REG_UUNPORT}
+ {aC}
+(2 rows)
+
+-- expectNomatch 19.7 &iM {a[^B-Z]} aC
+select * from test_regex('a[^B-Z]', 'aC', 'iM');
+ test_regex
+-----------------
+ {0,REG_UUNPORT}
+(1 row)
+
+select * from test_regex('a[^B-Z]', 'aC', 'iMb');
+ test_regex
+-----------------
+ {0,REG_UUNPORT}
+(1 row)
+
-- doing 20 "directors and embedded options"
-- expectError 20.1 & ***? BADPAT
select * from test_regex('***?', '', '');
@@ -3850,6 +3906,14 @@ select * from test_regex('^([^/]+?)(?:/([^/]+?))(?:/([^/]+?))?$', 'foo/bar/baz',
{foo/bar/baz,foo,bar,baz}
(2 rows)
+-- expectMatch 24.14 PRT {^(.+?)(?:/(.+?))(?:/(.+?)\3)?$} {foo/bar/baz/quux} {foo/bar/baz/quux} {foo} {bar/baz/quux} {}
+select * from test_regex('^(.+?)(?:/(.+?))(?:/(.+?)\3)?$', 'foo/bar/baz/quux', 'PRT');
+ test_regex
+----------------------------------------------
+ {3,REG_UBACKREF,REG_UNONPOSIX,REG_USHORTEST}
+ {foo/bar/baz/quux,foo,bar/baz/quux,NULL}
+(2 rows)
+
-- doing 25 "mixed quantifiers"
-- # this is very incomplete as yet
-- # should include |
@@ -4926,3 +4990,59 @@ select * from test_regex('(\Y)+', 'foo', 'LNP');
{"",""}
(2 rows)
+-- and now, tests not from either Spencer or the Tcl project
+-- These cases exercise additional code paths in pushfwd()/push()/combine()
+select * from test_regex('a\Y(?=45)', 'a45', 'HLP');
+ test_regex
+-----------------------------------------------
+ {0,REG_ULOOKAROUND,REG_UNONPOSIX,REG_ULOCALE}
+ {a}
+(2 rows)
+
+select * from test_regex('a(?=.)c', 'ac', 'HP');
+ test_regex
+-----------------------------------
+ {0,REG_ULOOKAROUND,REG_UNONPOSIX}
+ {ac}
+(2 rows)
+
+select * from test_regex('a(?=.).*(?=3)3*', 'azz33', 'HP');
+ test_regex
+-----------------------------------
+ {0,REG_ULOOKAROUND,REG_UNONPOSIX}
+ {azz33}
+(2 rows)
+
+select * from test_regex('a(?=\w)\w*(?=.).*', 'az3%', 'HLP');
+ test_regex
+-----------------------------------------------
+ {0,REG_ULOOKAROUND,REG_UNONPOSIX,REG_ULOCALE}
+ {az3%}
+(2 rows)
+
+-- These exercise the bulk-arc-movement paths in moveins() and moveouts();
+-- you may need to make them longer if you change BULK_ARC_OP_USE_SORT()
+select * from test_regex('ABCDEFGHIJKLMNOPQRSTUVWXYZ(?:\w|a|b|c|d|e|f|0|1|2|3|4|5|6|Q)',
+ 'ABCDEFGHIJKLMNOPQRSTUVWXYZ3', 'LP');
+ test_regex
+-------------------------------
+ {0,REG_UNONPOSIX,REG_ULOCALE}
+ {ABCDEFGHIJKLMNOPQRSTUVWXYZ3}
+(2 rows)
+
+select * from test_regex('ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789(\Y\Y)+',
+ 'ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789Z', 'LP');
+ test_regex
+-------------------------------------------
+ {1,REG_UNONPOSIX,REG_ULOCALE}
+ {ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789,""}
+(2 rows)
+
+select * from test_regex('((x|xabcdefghijklmnopqrstuvwxyz0123456789)x*|[^y]z)$',
+ 'az', '');
+ test_regex
+--------------
+ {2}
+ {az,az,NULL}
+(2 rows)
+
diff --git a/src/test/modules/test_regex/expected/test_regex_utf8.out b/src/test/modules/test_regex/expected/test_regex_utf8.out
index 112698ac618..3b56f36c071 100644
--- a/src/test/modules/test_regex/expected/test_regex_utf8.out
+++ b/src/test/modules/test_regex/expected/test_regex_utf8.out
@@ -98,3 +98,109 @@ select * from test_regex('[[:alnum:]]*[[:upper:]]*[\u1000-\u2000]*\u1237',
{0,REG_UBBS,REG_UNONPOSIX,REG_UUNPORT,REG_ULOCALE}
(1 row)
+select * from test_regex('[[:alnum:]]*[[:upper:]]*[\u1000-\u2000]*\u1237',
+ E'\u1500\u1237', 'iELMP');
+ test_regex
+----------------------------------------------------
+ {0,REG_UBBS,REG_UNONPOSIX,REG_UUNPORT,REG_ULOCALE}
+ {ᔀሷ}
+(2 rows)
+
+-- systematically test char classes
+select * from test_regex('[[:alnum:]]+', E'x\u1500\u1237', 'L');
+ test_regex
+-----------------
+ {0,REG_ULOCALE}
+ {xᔀሷ}
+(2 rows)
+
+select * from test_regex('[[:alpha:]]+', E'x\u1500\u1237', 'L');
+ test_regex
+-----------------
+ {0,REG_ULOCALE}
+ {xᔀሷ}
+(2 rows)
+
+select * from test_regex('[[:ascii:]]+', E'x\u1500\u1237', 'L');
+ test_regex
+-----------------
+ {0,REG_ULOCALE}
+ {x}
+(2 rows)
+
+select * from test_regex('[[:blank:]]+', E'x \t\u1500\u1237', 'L');
+ test_regex
+-----------------
+ {0,REG_ULOCALE}
+ {" "}
+(2 rows)
+
+select * from test_regex('[[:cntrl:]]+', E'x\u1500\u1237', 'L');
+ test_regex
+-----------------
+ {0,REG_ULOCALE}
+(1 row)
+
+select * from test_regex('[[:digit:]]+', E'x9\u1500\u1237', 'L');
+ test_regex
+-----------------
+ {0,REG_ULOCALE}
+ {9}
+(2 rows)
+
+select * from test_regex('[[:graph:]]+', E'x\u1500\u1237', 'L');
+ test_regex
+-----------------
+ {0,REG_ULOCALE}
+ {xᔀሷ}
+(2 rows)
+
+select * from test_regex('[[:lower:]]+', E'x\u1500\u1237', 'L');
+ test_regex
+-----------------
+ {0,REG_ULOCALE}
+ {x}
+(2 rows)
+
+select * from test_regex('[[:print:]]+', E'x\u1500\u1237', 'L');
+ test_regex
+-----------------
+ {0,REG_ULOCALE}
+ {xᔀሷ}
+(2 rows)
+
+select * from test_regex('[[:punct:]]+', E'x.\u1500\u1237', 'L');
+ test_regex
+-----------------
+ {0,REG_ULOCALE}
+ {.}
+(2 rows)
+
+select * from test_regex('[[:space:]]+', E'x \t\u1500\u1237', 'L');
+ test_regex
+-----------------
+ {0,REG_ULOCALE}
+ {" "}
+(2 rows)
+
+select * from test_regex('[[:upper:]]+', E'xX\u1500\u1237', 'L');
+ test_regex
+-----------------
+ {0,REG_ULOCALE}
+ {X}
+(2 rows)
+
+select * from test_regex('[[:xdigit:]]+', E'xa9\u1500\u1237', 'L');
+ test_regex
+-----------------
+ {0,REG_ULOCALE}
+ {a9}
+(2 rows)
+
+select * from test_regex('[[:word:]]+', E'x_\u1500\u1237', 'L');
+ test_regex
+-----------------
+ {0,REG_ULOCALE}
+ {x_ᔀሷ}
+(2 rows)
+
diff --git a/src/test/modules/test_regex/sql/test_regex.sql b/src/test/modules/test_regex/sql/test_regex.sql
index 3419564203a..389b8b61b3b 100644
--- a/src/test/modules/test_regex/sql/test_regex.sql
+++ b/src/test/modules/test_regex/sql/test_regex.sql
@@ -304,6 +304,12 @@ select * from test_regex('a[[=x=]]', 'ay', '+Lb');
-- expectNomatch 9.9 &+L {a[[=x=]]} az
select * from test_regex('a[[=x=]]', 'az', '+L');
select * from test_regex('a[[=x=]]', 'az', '+Lb');
+-- expectMatch 9.9b &iL {a[[=Y=]]} ay ay
+select * from test_regex('a[[=Y=]]', 'ay', 'iL');
+select * from test_regex('a[[=Y=]]', 'ay', 'iLb');
+-- expectNomatch 9.9c &L {a[[=Y=]]} ay
+select * from test_regex('a[[=Y=]]', 'ay', 'L');
+select * from test_regex('a[[=Y=]]', 'ay', 'Lb');
-- expectError 9.10 & {a[0-[=x=]]} ERANGE
select * from test_regex('a[0-[=x=]]', '', '');
select * from test_regex('a[0-[=x=]]', '', 'b');
@@ -864,6 +870,12 @@ select * from test_regex('a[b-d]', 'aC', 'iMb');
-- expectNomatch 19.5 &iM {a[^b-d]} aC
select * from test_regex('a[^b-d]', 'aC', 'iM');
select * from test_regex('a[^b-d]', 'aC', 'iMb');
+-- expectMatch 19.6 &iM {a[B-Z]} aC aC
+select * from test_regex('a[B-Z]', 'aC', 'iM');
+select * from test_regex('a[B-Z]', 'aC', 'iMb');
+-- expectNomatch 19.7 &iM {a[^B-Z]} aC
+select * from test_regex('a[^B-Z]', 'aC', 'iM');
+select * from test_regex('a[^B-Z]', 'aC', 'iMb');
-- doing 20 "directors and embedded options"
@@ -1171,6 +1183,8 @@ select * from test_regex('z*4', '123zzzz456', '-');
select * from test_regex('z*?4', '123zzzz456', 'PT');
-- expectMatch 24.13 PT {^([^/]+?)(?:/([^/]+?))(?:/([^/]+?))?$} {foo/bar/baz} {foo/bar/baz} {foo} {bar} {baz}
select * from test_regex('^([^/]+?)(?:/([^/]+?))(?:/([^/]+?))?$', 'foo/bar/baz', 'PT');
+-- expectMatch 24.14 PRT {^(.+?)(?:/(.+?))(?:/(.+?)\3)?$} {foo/bar/baz/quux} {foo/bar/baz/quux} {foo} {bar/baz/quux} {}
+select * from test_regex('^(.+?)(?:/(.+?))(?:/(.+?)\3)?$', 'foo/bar/baz/quux', 'PRT');
-- doing 25 "mixed quantifiers"
-- # this is very incomplete as yet
@@ -1741,3 +1755,21 @@ select * from test_regex(repeat('x*y*z*', 200), 'x', 'N');
-- regexp {(\Y)+} foo
-- } 1
select * from test_regex('(\Y)+', 'foo', 'LNP');
+
+
+-- and now, tests not from either Spencer or the Tcl project
+
+-- These cases exercise additional code paths in pushfwd()/push()/combine()
+select * from test_regex('a\Y(?=45)', 'a45', 'HLP');
+select * from test_regex('a(?=.)c', 'ac', 'HP');
+select * from test_regex('a(?=.).*(?=3)3*', 'azz33', 'HP');
+select * from test_regex('a(?=\w)\w*(?=.).*', 'az3%', 'HLP');
+
+-- These exercise the bulk-arc-movement paths in moveins() and moveouts();
+-- you may need to make them longer if you change BULK_ARC_OP_USE_SORT()
+select * from test_regex('ABCDEFGHIJKLMNOPQRSTUVWXYZ(?:\w|a|b|c|d|e|f|0|1|2|3|4|5|6|Q)',
+ 'ABCDEFGHIJKLMNOPQRSTUVWXYZ3', 'LP');
+select * from test_regex('ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789(\Y\Y)+',
+ 'ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789Z', 'LP');
+select * from test_regex('((x|xabcdefghijklmnopqrstuvwxyz0123456789)x*|[^y]z)$',
+ 'az', '');
diff --git a/src/test/modules/test_regex/sql/test_regex_utf8.sql b/src/test/modules/test_regex/sql/test_regex_utf8.sql
index cfd9396194f..f23907162e4 100644
--- a/src/test/modules/test_regex/sql/test_regex_utf8.sql
+++ b/src/test/modules/test_regex/sql/test_regex_utf8.sql
@@ -58,3 +58,21 @@ select * from test_regex('[[:alnum:]]*[[:upper:]]*[\u1000-\u2000]*\u1237',
E'\u1500\u1237', 'ELMP');
select * from test_regex('[[:alnum:]]*[[:upper:]]*[\u1000-\u2000]*\u1237',
E'A\u1239', 'ELMP');
+select * from test_regex('[[:alnum:]]*[[:upper:]]*[\u1000-\u2000]*\u1237',
+ E'\u1500\u1237', 'iELMP');
+
+-- systematically test char classes
+select * from test_regex('[[:alnum:]]+', E'x\u1500\u1237', 'L');
+select * from test_regex('[[:alpha:]]+', E'x\u1500\u1237', 'L');
+select * from test_regex('[[:ascii:]]+', E'x\u1500\u1237', 'L');
+select * from test_regex('[[:blank:]]+', E'x \t\u1500\u1237', 'L');
+select * from test_regex('[[:cntrl:]]+', E'x\u1500\u1237', 'L');
+select * from test_regex('[[:digit:]]+', E'x9\u1500\u1237', 'L');
+select * from test_regex('[[:graph:]]+', E'x\u1500\u1237', 'L');
+select * from test_regex('[[:lower:]]+', E'x\u1500\u1237', 'L');
+select * from test_regex('[[:print:]]+', E'x\u1500\u1237', 'L');
+select * from test_regex('[[:punct:]]+', E'x.\u1500\u1237', 'L');
+select * from test_regex('[[:space:]]+', E'x \t\u1500\u1237', 'L');
+select * from test_regex('[[:upper:]]+', E'xX\u1500\u1237', 'L');
+select * from test_regex('[[:xdigit:]]+', E'xa9\u1500\u1237', 'L');
+select * from test_regex('[[:word:]]+', E'x_\u1500\u1237', 'L');
generated by cgit v1.2.3 (git 2.25.1) at 2025-05-31 07:56:34 +0000