Movatterモバイル変換

[0]ホーム
URL:

Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

 Sign up
Appearance settings

Commit78ed8e0

Browse files
committed
Fix unescaping of JSON Unicode escapes, especially for non-UTF8.
Per discussion on -hackers. We treat Unicode escapes when unescapingthem similarly to the way we treat them in PostgreSQL string literals.Escapes in the ASCII range are always accepted, no matter what thedatabase encoding. Escapes for higher code points are only processed inUTF8 databases, and attempts to process them in other databases willresult in an error. \u0000 is never unescaped, since it would result inan impermissible null byte.
1 parentc1d729b commit78ed8e0

File tree

5 files changed

+1033
-11
lines changed

5 files changed

+1033
-11
lines changed

‎doc/src/sgml/func.sgml

Lines changed: 11 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -10159,6 +10159,17 @@ table2-mapping
1015910159
</para>
1016010160
</note>
1016110161

10162+
<note>
10163+
<para>
10164+
Many of these functions and operators will convert Unicode escapes
10165+
in the JSON text to the appropriate UTF8 character when the database encoding is UTF8. In
10166+
other encodings the escape sequence must be for an ASCII character, and any other code point
10167+
in a Unicode escape sequence will result in an error.
10168+
In general, it is best to avoid mixing Unicode escapes in JSON with a non-UTF8 database
10169+
encoding, if possible.
10170+
</para>
10171+
</note>
10172+
1016210173
<note>
1016310174
<para>
1016410175
The <xref linkend="hstore"> extension has a cast from <type>hstore</type> to

‎src/backend/utils/adt/json.c

Lines changed: 34 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -717,7 +717,6 @@ json_lex_string(JsonLexContext *lex)
717717
{
718718
charutf8str[5];
719719
intutf8len;
720-
char*converted;
721720

722721
if (ch >=0xd800&&ch <=0xdbff)
723722
{
@@ -749,13 +748,40 @@ json_lex_string(JsonLexContext *lex)
749748
errdetail("low order surrogate must follow a high order surrogate."),
750749
report_json_context(lex)));
751750

752-
unicode_to_utf8(ch, (unsignedchar*)utf8str);
753-
utf8len=pg_utf_mblen((unsignedchar*)utf8str);
754-
utf8str[utf8len]='\0';
755-
converted=pg_any_to_server(utf8str,utf8len,PG_UTF8);
756-
appendStringInfoString(lex->strval,converted);
757-
if (converted!=utf8str)
758-
pfree(converted);
751+
/*
752+
* For UTF8, replace the escape sequence by the actual utf8
753+
* character in lex->strval. Do this also for other encodings
754+
* if the escape designates an ASCII character, otherwise
755+
* raise an error. We don't ever unescape a \u0000, since that
756+
* would result in an impermissible nul byte.
757+
*/
758+
759+
if (ch==0)
760+
{
761+
appendStringInfoString(lex->strval,"\\u0000");
762+
}
763+
elseif (GetDatabaseEncoding()==PG_UTF8)
764+
{
765+
unicode_to_utf8(ch, (unsignedchar*)utf8str);
766+
utf8len=pg_utf_mblen((unsignedchar*)utf8str);
767+
appendBinaryStringInfo(lex->strval,utf8str,utf8len);
768+
}
769+
elseif (ch <=0x007f)
770+
{
771+
/*
772+
* This is the only way to designate things like a form feed
773+
* character in JSON, so it's useful in all encodings.
774+
*/
775+
appendStringInfoChar(lex->strval, (char)ch);
776+
}
777+
else
778+
{
779+
ereport(ERROR,
780+
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
781+
errmsg("invalid input syntax for type json"),
782+
errdetail("Unicode escape for code points higher than U+007F not permitted in non-UTF8 encoding"),
783+
report_json_context(lex)));
784+
}
759785

760786
}
761787
}

‎src/test/regress/expected/json.out

Lines changed: 21 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -921,8 +921,8 @@ ERROR: cannot call json_populate_recordset on a nested object
921921
select * from json_populate_recordset(row('def',99,null)::jpop,'[{"c":[100,200,300],"x":43.2},{"a":{"z":true},"b":3,"c":"2012-01-20 10:42:53"}]') q;
922922
ERROR: cannot call json_populate_recordset on a nested object
923923
-- handling of unicode surrogate pairs
924-
select json '{ "a": "\ud83d\ude04\ud83d\udc36" }' -> 'a' ascorrect;
925-
correct
924+
select json '{ "a": "\ud83d\ude04\ud83d\udc36" }' -> 'a' ascorrect_in_utf8;
925+
correct_in_utf8
926926
----------------------------
927927
"\ud83d\ude04\ud83d\udc36"
928928
(1 row)
@@ -943,3 +943,22 @@ select json '{ "a": "\ude04X" }' -> 'a'; -- orphan low surrogate
943943
ERROR: invalid input syntax for type json
944944
DETAIL: low order surrogate must follow a high order surrogate.
945945
CONTEXT: JSON data, line 1: { "a":...
946+
--handling of simple unicode escapes
947+
select json '{ "a": "the Copyright \u00a9 sign" }' ->> 'a' as correct_in_utf8;
948+
correct_in_utf8
949+
----------------------
950+
the Copyright © sign
951+
(1 row)
952+
953+
select json '{ "a": "dollar \u0024 character" }' ->> 'a' as correct_everywhere;
954+
correct_everywhere
955+
--------------------
956+
dollar $ character
957+
(1 row)
958+
959+
select json '{ "a": "null \u0000 escape" }' ->> 'a' as not_unescaped;
960+
not_unescaped
961+
--------------------
962+
null \u0000 escape
963+
(1 row)
964+

0 commit comments

Comments
 (0)
[8]ページ先頭
©2009-2025 Movatter.jp