mirror of
https://github.com/postgres/postgres.git
synced 2025-05-28 00:03:23 -04:00
1d88a75c42
Non-backtracking flex parsers work faster than backtracking ones. So, this commit gets rid of backtracking in jsonpath_scan.l. That required explicit handling of some cases as well as manual backtracking for some cases. More regression tests for numerics are added. Discussion: https://mail.google.com/mail/u/0?ik=a20b091faa&view=om&permmsgid=msg-f%3A1628425344167939063 Author: John Naylor, Nikita Gluknov, Alexander Korotkov
250 lines
8.5 KiB
Plaintext
250 lines
8.5 KiB
Plaintext
-- encoding-sensitive tests for jsonpath
|
|
-- checks for double-quoted values
|
|
-- basic unicode input
|
|
SELECT '"\u"'::jsonpath; -- ERROR, incomplete escape
|
|
ERROR: bad jsonpath representation
|
|
LINE 1: SELECT '"\u"'::jsonpath;
|
|
^
|
|
DETAIL: Unicode sequence is invalid at or near "\u"
|
|
SELECT '"\u00"'::jsonpath; -- ERROR, incomplete escape
|
|
ERROR: bad jsonpath representation
|
|
LINE 1: SELECT '"\u00"'::jsonpath;
|
|
^
|
|
DETAIL: Unicode sequence is invalid at or near "\u00"
|
|
SELECT '"\u000g"'::jsonpath; -- ERROR, g is not a hex digit
|
|
ERROR: bad jsonpath representation
|
|
LINE 1: SELECT '"\u000g"'::jsonpath;
|
|
^
|
|
DETAIL: Unicode sequence is invalid at or near "\u000"
|
|
SELECT '"\u0000"'::jsonpath; -- OK, legal escape
|
|
ERROR: unsupported Unicode escape sequence
|
|
LINE 1: SELECT '"\u0000"'::jsonpath;
|
|
^
|
|
DETAIL: \u0000 cannot be converted to text.
|
|
SELECT '"\uaBcD"'::jsonpath; -- OK, uppercase and lower case both OK
|
|
jsonpath
|
|
----------
|
|
"ꯍ"
|
|
(1 row)
|
|
|
|
-- handling of unicode surrogate pairs
|
|
select '"\ud83d\ude04\ud83d\udc36"'::jsonpath as correct_in_utf8;
|
|
correct_in_utf8
|
|
-----------------
|
|
"😄🐶"
|
|
(1 row)
|
|
|
|
select '"\ud83d\ud83d"'::jsonpath; -- 2 high surrogates in a row
|
|
ERROR: invalid input syntax for type jsonpath
|
|
LINE 1: select '"\ud83d\ud83d"'::jsonpath;
|
|
^
|
|
DETAIL: Unicode high surrogate must not follow a high surrogate.
|
|
select '"\ude04\ud83d"'::jsonpath; -- surrogates in wrong order
|
|
ERROR: invalid input syntax for type jsonpath
|
|
LINE 1: select '"\ude04\ud83d"'::jsonpath;
|
|
^
|
|
DETAIL: Unicode low surrogate must follow a high surrogate.
|
|
select '"\ud83dX"'::jsonpath; -- orphan high surrogate
|
|
ERROR: invalid input syntax for type jsonpath
|
|
LINE 1: select '"\ud83dX"'::jsonpath;
|
|
^
|
|
DETAIL: Unicode low surrogate must follow a high surrogate.
|
|
select '"\ude04X"'::jsonpath; -- orphan low surrogate
|
|
ERROR: invalid input syntax for type jsonpath
|
|
LINE 1: select '"\ude04X"'::jsonpath;
|
|
^
|
|
DETAIL: Unicode low surrogate must follow a high surrogate.
|
|
--handling of simple unicode escapes
|
|
select '"the Copyright \u00a9 sign"'::jsonpath as correct_in_utf8;
|
|
correct_in_utf8
|
|
------------------------
|
|
"the Copyright © sign"
|
|
(1 row)
|
|
|
|
select '"dollar \u0024 character"'::jsonpath as correct_everywhere;
|
|
correct_everywhere
|
|
----------------------
|
|
"dollar $ character"
|
|
(1 row)
|
|
|
|
select '"dollar \\u0024 character"'::jsonpath as not_an_escape;
|
|
not_an_escape
|
|
----------------------------
|
|
"dollar \\u0024 character"
|
|
(1 row)
|
|
|
|
select '"null \u0000 escape"'::jsonpath as not_unescaped;
|
|
ERROR: unsupported Unicode escape sequence
|
|
LINE 1: select '"null \u0000 escape"'::jsonpath as not_unescaped;
|
|
^
|
|
DETAIL: \u0000 cannot be converted to text.
|
|
select '"null \\u0000 escape"'::jsonpath as not_an_escape;
|
|
not_an_escape
|
|
-----------------------
|
|
"null \\u0000 escape"
|
|
(1 row)
|
|
|
|
-- checks for single-quoted values
|
|
-- basic unicode input
|
|
SELECT E'\'\u\''::jsonpath; -- ERROR, incomplete escape
|
|
ERROR: invalid Unicode escape
|
|
LINE 1: SELECT E'\'\u\''::jsonpath;
|
|
^
|
|
HINT: Unicode escapes must be \uXXXX or \UXXXXXXXX.
|
|
SELECT E'\'\u00\''::jsonpath; -- ERROR, incomplete escape
|
|
ERROR: invalid Unicode escape
|
|
LINE 1: SELECT E'\'\u00\''::jsonpath;
|
|
^
|
|
HINT: Unicode escapes must be \uXXXX or \UXXXXXXXX.
|
|
SELECT E'\'\u000g\''::jsonpath; -- ERROR, g is not a hex digit
|
|
ERROR: invalid Unicode escape
|
|
LINE 1: SELECT E'\'\u000g\''::jsonpath;
|
|
^
|
|
HINT: Unicode escapes must be \uXXXX or \UXXXXXXXX.
|
|
SELECT E'\'\u0000\''::jsonpath; -- OK, legal escape
|
|
ERROR: invalid Unicode escape value at or near "E'\'\u0000"
|
|
LINE 1: SELECT E'\'\u0000\''::jsonpath;
|
|
^
|
|
SELECT E'\'\uaBcD\''::jsonpath; -- OK, uppercase and lower case both OK
|
|
jsonpath
|
|
----------
|
|
"ꯍ"
|
|
(1 row)
|
|
|
|
-- handling of unicode surrogate pairs
|
|
select E'\'\ud83d\ude04\ud83d\udc36\''::jsonpath as correct_in_utf8;
|
|
correct_in_utf8
|
|
-----------------
|
|
"😄🐶"
|
|
(1 row)
|
|
|
|
select E'\'\ud83d\ud83d\''::jsonpath; -- 2 high surrogates in a row
|
|
ERROR: invalid Unicode surrogate pair at or near "E'\'\ud83d\ud83d"
|
|
LINE 1: select E'\'\ud83d\ud83d\''::jsonpath;
|
|
^
|
|
select E'\'\ude04\ud83d\''::jsonpath; -- surrogates in wrong order
|
|
ERROR: invalid Unicode surrogate pair at or near "E'\'\ude04"
|
|
LINE 1: select E'\'\ude04\ud83d\''::jsonpath;
|
|
^
|
|
select E'\'\ud83dX\''::jsonpath; -- orphan high surrogate
|
|
ERROR: invalid Unicode surrogate pair at or near "E'\'\ud83dX"
|
|
LINE 1: select E'\'\ud83dX\''::jsonpath;
|
|
^
|
|
select E'\'\ude04X\''::jsonpath; -- orphan low surrogate
|
|
ERROR: invalid Unicode surrogate pair at or near "E'\'\ude04"
|
|
LINE 1: select E'\'\ude04X\''::jsonpath;
|
|
^
|
|
--handling of simple unicode escapes
|
|
select E'\'the Copyright \u00a9 sign\''::jsonpath as correct_in_utf8;
|
|
correct_in_utf8
|
|
------------------------
|
|
"the Copyright © sign"
|
|
(1 row)
|
|
|
|
select E'\'dollar \u0024 character\''::jsonpath as correct_everywhere;
|
|
correct_everywhere
|
|
----------------------
|
|
"dollar $ character"
|
|
(1 row)
|
|
|
|
select E'\'dollar \\u0024 character\''::jsonpath as not_an_escape;
|
|
not_an_escape
|
|
----------------------
|
|
"dollar $ character"
|
|
(1 row)
|
|
|
|
select E'\'null \u0000 escape\''::jsonpath as not_unescaped;
|
|
ERROR: invalid Unicode escape value at or near "E'\'null \u0000"
|
|
LINE 1: select E'\'null \u0000 escape\''::jsonpath as not_unescaped;
|
|
^
|
|
select E'\'null \\u0000 escape\''::jsonpath as not_an_escape;
|
|
ERROR: unsupported Unicode escape sequence
|
|
LINE 1: select E'\'null \\u0000 escape\''::jsonpath as not_an_escape...
|
|
^
|
|
DETAIL: \u0000 cannot be converted to text.
|
|
-- checks for quoted key names
|
|
-- basic unicode input
|
|
SELECT '$."\u"'::jsonpath; -- ERROR, incomplete escape
|
|
ERROR: bad jsonpath representation
|
|
LINE 1: SELECT '$."\u"'::jsonpath;
|
|
^
|
|
DETAIL: Unicode sequence is invalid at or near "\u"
|
|
SELECT '$."\u00"'::jsonpath; -- ERROR, incomplete escape
|
|
ERROR: bad jsonpath representation
|
|
LINE 1: SELECT '$."\u00"'::jsonpath;
|
|
^
|
|
DETAIL: Unicode sequence is invalid at or near "\u00"
|
|
SELECT '$."\u000g"'::jsonpath; -- ERROR, g is not a hex digit
|
|
ERROR: bad jsonpath representation
|
|
LINE 1: SELECT '$."\u000g"'::jsonpath;
|
|
^
|
|
DETAIL: Unicode sequence is invalid at or near "\u000"
|
|
SELECT '$."\u0000"'::jsonpath; -- OK, legal escape
|
|
ERROR: unsupported Unicode escape sequence
|
|
LINE 1: SELECT '$."\u0000"'::jsonpath;
|
|
^
|
|
DETAIL: \u0000 cannot be converted to text.
|
|
SELECT '$."\uaBcD"'::jsonpath; -- OK, uppercase and lower case both OK
|
|
jsonpath
|
|
----------
|
|
$."ꯍ"
|
|
(1 row)
|
|
|
|
-- handling of unicode surrogate pairs
|
|
select '$."\ud83d\ude04\ud83d\udc36"'::jsonpath as correct_in_utf8;
|
|
correct_in_utf8
|
|
-----------------
|
|
$."😄🐶"
|
|
(1 row)
|
|
|
|
select '$."\ud83d\ud83d"'::jsonpath; -- 2 high surrogates in a row
|
|
ERROR: invalid input syntax for type jsonpath
|
|
LINE 1: select '$."\ud83d\ud83d"'::jsonpath;
|
|
^
|
|
DETAIL: Unicode high surrogate must not follow a high surrogate.
|
|
select '$."\ude04\ud83d"'::jsonpath; -- surrogates in wrong order
|
|
ERROR: invalid input syntax for type jsonpath
|
|
LINE 1: select '$."\ude04\ud83d"'::jsonpath;
|
|
^
|
|
DETAIL: Unicode low surrogate must follow a high surrogate.
|
|
select '$."\ud83dX"'::jsonpath; -- orphan high surrogate
|
|
ERROR: invalid input syntax for type jsonpath
|
|
LINE 1: select '$."\ud83dX"'::jsonpath;
|
|
^
|
|
DETAIL: Unicode low surrogate must follow a high surrogate.
|
|
select '$."\ude04X"'::jsonpath; -- orphan low surrogate
|
|
ERROR: invalid input syntax for type jsonpath
|
|
LINE 1: select '$."\ude04X"'::jsonpath;
|
|
^
|
|
DETAIL: Unicode low surrogate must follow a high surrogate.
|
|
--handling of simple unicode escapes
|
|
select '$."the Copyright \u00a9 sign"'::jsonpath as correct_in_utf8;
|
|
correct_in_utf8
|
|
--------------------------
|
|
$."the Copyright © sign"
|
|
(1 row)
|
|
|
|
select '$."dollar \u0024 character"'::jsonpath as correct_everywhere;
|
|
correct_everywhere
|
|
------------------------
|
|
$."dollar $ character"
|
|
(1 row)
|
|
|
|
select '$."dollar \\u0024 character"'::jsonpath as not_an_escape;
|
|
not_an_escape
|
|
------------------------------
|
|
$."dollar \\u0024 character"
|
|
(1 row)
|
|
|
|
select '$."null \u0000 escape"'::jsonpath as not_unescaped;
|
|
ERROR: unsupported Unicode escape sequence
|
|
LINE 1: select '$."null \u0000 escape"'::jsonpath as not_unescaped;
|
|
^
|
|
DETAIL: \u0000 cannot be converted to text.
|
|
select '$."null \\u0000 escape"'::jsonpath as not_an_escape;
|
|
not_an_escape
|
|
-------------------------
|
|
$."null \\u0000 escape"
|
|
(1 row)
|
|
|