Make sure chr(int) can't create invalid UTF8 sequences.

Several years ago we changed chr(int) so that if the database encoding is
UTF8, it would interpret its argument as a Unicode code point and expand it
into the appropriate multibyte sequence.  However, we weren't sufficiently
careful about checking validity of the input.  According to RFC3629, UTF8
disallows code points above U+10FFFF (note that the predecessor standard
RFC2279 was more liberal).  Also, both versions of the UTF8 spec agree
that Unicode surrogate-pair codes should never appear in UTF8.  Because
our encoding validity checks follow RFC3629, our failure to enforce these
restrictions in chr() means it could be used to produce text strings that
will be rejected when the database is dumped and reloaded.  To ensure
consistency with the input functions, let's actually apply
pg_utf8_islegal() to the proposed output of chr().

Per discussion, this seems like too much of a behavioral change to
back-patch, but it's not too late to squeeze it into 9.4.

1 files changed, 18 insertions, 7 deletions

diff --git a/src/backend/utils/adt/oracle_compat.c b/src/backend/utils/adt/oracle_compat.c
index 4dab45caf4f..ba3d5d6e139 100644
--- a/src/backend/utils/adt/oracle_compat.c
+++ b/src/backend/utils/adt/oracle_compat.c
@@ -932,10 +932,14 @@ chr			(PG_FUNCTION_ARGS)
 	{
 		/* for Unicode we treat the argument as a code point */
 		int			bytes;
-		char	   *wch;
+		unsigned char *wch;
 
-		/* We only allow valid Unicode code points */
-		if (cvalue > 0x001fffff)
+		/*
+		 * We only allow valid Unicode code points; per RFC3629 that stops at
+		 * U+10FFFF, even though 4-byte UTF8 sequences can hold values up to
+		 * U+1FFFFF.
+		 */
+		if (cvalue > 0x0010ffff)
 			ereport(ERROR,
 					(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
 					 errmsg("requested character too large for encoding: %d",
@@ -950,7 +954,7 @@ chr			(PG_FUNCTION_ARGS)
 
 		result = (text *) palloc(VARHDRSZ + bytes);
 		SET_VARSIZE(result, VARHDRSZ + bytes);
-		wch = VARDATA(result);
+		wch = (unsigned char *) VARDATA(result);
 
 		if (bytes == 2)
 		{
@@ -971,8 +975,17 @@ chr			(PG_FUNCTION_ARGS)
 			wch[3] = 0x80 | (cvalue & 0x3F);
 		}
 
+		/*
+		 * The preceding range check isn't sufficient, because UTF8 excludes
+		 * Unicode "surrogate pair" codes.  Make sure what we created is valid
+		 * UTF8.
+		 */
+		if (!pg_utf8_islegal(wch, bytes))
+			ereport(ERROR,
+					(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
+					 errmsg("requested character not valid for encoding: %d",
+							cvalue)));
 	}
-
 	else
 	{
 		bool		is_mb;
@@ -981,7 +994,6 @@ chr			(PG_FUNCTION_ARGS)
 		 * Error out on arguments that make no sense or that we can't validly
 		 * represent in the encoding.
 		 */
-
 		if (cvalue == 0)
 			ereport(ERROR,
 					(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
@@ -995,7 +1007,6 @@ chr			(PG_FUNCTION_ARGS)
 					 errmsg("requested character too large for encoding: %d",
 							cvalue)));
 
-
 		result = (text *) palloc(VARHDRSZ + 1);
 		SET_VARSIZE(result, VARHDRSZ + 1);
 		*VARDATA(result) = (char) cvalue;