Make replace(), split_part(), and string_to_array() behave somewhat sanely

when handed an invalidly-encoded pattern.  The previous coding could get
into an infinite loop if pg_mb2wchar_with_len() returned a zero-length
string after we'd tested for nonempty pattern; which is exactly what it
will do if the string consists only of an incomplete multibyte character.
This led to either an out-of-memory error or a backend crash depending
on platform.  Per report from Wiktor Wodecki.

src/backend/utils/adt/varlena.c

@@ -8,7 +8,7 @@
  *
  *
  * IDENTIFICATION
- *	  $PostgreSQL: pgsql/src/backend/utils/adt/varlena.c,v 1.139.2.4 2006/10/07 00:11:59 tgl Exp $
+ *	  $PostgreSQL: pgsql/src/backend/utils/adt/varlena.c,v 1.139.2.5 2007/07/19 20:34:34 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -2076,8 +2076,8 @@ replace_text(PG_FUNCTION_ARGS)
 	text	   *src_text = PG_GETARG_TEXT_P(0);
 	text	   *from_sub_text = PG_GETARG_TEXT_P(1);
 	text	   *to_sub_text = PG_GETARG_TEXT_P(2);
-	int			src_text_len = TEXTLEN(src_text);
-	int			from_sub_text_len = TEXTLEN(from_sub_text);
+	int			src_text_len;
+	int			from_sub_text_len;
 	TextPositionState state;
 	text	   *chunk_text;
 	text	   *ret_text;
@@ -2085,11 +2085,22 @@ replace_text(PG_FUNCTION_ARGS)
 	int			curr_posn;
 	StringInfoData str;
 
-	if (src_text_len == 0 || from_sub_text_len == 0)
-		PG_RETURN_TEXT_P(src_text);
-
 	text_position_setup(src_text, from_sub_text, &state);
 
+	/*
+	 * Note: we check the converted string length, not the original, because
+	 * they could be different if the input contained invalid encoding.
+	 */
+	src_text_len = state.len1;
+	from_sub_text_len = state.len2;
+
+	/* Return unmodified source string if empty source or pattern */
+	if (src_text_len < 1 || from_sub_text_len < 1)
+	{
+		text_position_cleanup(&state);
+		PG_RETURN_TEXT_P(src_text);
+	}
+
 	start_posn = 1;
 	curr_posn = text_position_next(1, &state);
 
@@ -2104,6 +2115,9 @@ replace_text(PG_FUNCTION_ARGS)
 
 	do
 	{
+		CHECK_FOR_INTERRUPTS();
+
+		/* copy the data skipped over by last text_position_next() */
 		chunk_text = text_substring(PointerGetDatum(src_text),
 									start_posn,
 									curr_posn - start_posn,
@@ -2396,8 +2410,8 @@ split_text(PG_FUNCTION_ARGS)
 	text	   *inputstring = PG_GETARG_TEXT_P(0);
 	text	   *fldsep = PG_GETARG_TEXT_P(1);
 	int			fldnum = PG_GETARG_INT32(2);
-	int			inputstring_len = TEXTLEN(inputstring);
-	int			fldsep_len = TEXTLEN(fldsep);
+	int			inputstring_len;
+	int			fldsep_len;
 	TextPositionState state;
 	int			start_posn;
 	int			end_posn;
@@ -2409,13 +2423,26 @@ split_text(PG_FUNCTION_ARGS)
 				(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
 				 errmsg("field position must be greater than zero")));
 
+	text_position_setup(inputstring, fldsep, &state);
+
+	/*
+	 * Note: we check the converted string length, not the original, because
+	 * they could be different if the input contained invalid encoding.
+	 */
+	inputstring_len = state.len1;
+	fldsep_len = state.len2;
+
 	/* return empty string for empty input string */
 	if (inputstring_len < 1)
+	{
+		text_position_cleanup(&state);
 		PG_RETURN_TEXT_P(PG_STR_GET_TEXT(""));
+	}
 
 	/* empty field separator */
 	if (fldsep_len < 1)
 	{
+		text_position_cleanup(&state);
 		/* if first field, return input string, else empty string */
 		if (fldnum == 1)
 			PG_RETURN_TEXT_P(inputstring);
@@ -2423,8 +2450,6 @@ split_text(PG_FUNCTION_ARGS)
 			PG_RETURN_TEXT_P(PG_STR_GET_TEXT(""));
 	}
 
-	text_position_setup(inputstring, fldsep, &state);
-
 	/* identify bounds of first field */
 	start_posn = 1;
 	end_posn = text_position_next(1, &state);
@@ -2484,8 +2509,8 @@ text_to_array(PG_FUNCTION_ARGS)
 {
 	text	   *inputstring = PG_GETARG_TEXT_P(0);
 	text	   *fldsep = PG_GETARG_TEXT_P(1);
-	int			inputstring_len = TEXTLEN(inputstring);
-	int			fldsep_len = TEXTLEN(fldsep);
+	int			inputstring_len;
+	int			fldsep_len;
 	TextPositionState state;
 	int			fldnum;
 	int			start_posn;
@@ -2493,23 +2518,38 @@ text_to_array(PG_FUNCTION_ARGS)
 	text	   *result_text;
 	ArrayBuildState *astate = NULL;
 
+	text_position_setup(inputstring, fldsep, &state);
+
+	/*
+	 * Note: we check the converted string length, not the original, because
+	 * they could be different if the input contained invalid encoding.
+	 */
+	inputstring_len = state.len1;
+	fldsep_len = state.len2;
+
 	/* return NULL for empty input string */
 	if (inputstring_len < 1)
+	{
+		text_position_cleanup(&state);
 		PG_RETURN_NULL();
+	}
 
 	/*
 	 * empty field separator return one element, 1D, array using the input
 	 * string
 	 */
 	if (fldsep_len < 1)
+	{
+		text_position_cleanup(&state);
 		PG_RETURN_ARRAYTYPE_P(create_singleton_array(fcinfo, TEXTOID,
 										   PointerGetDatum(inputstring), 1));
-
-	text_position_setup(inputstring, fldsep, &state);
+	}
 
 	start_posn = 1;
 	for (fldnum = 1;; fldnum++) /* field number is 1 based */
 	{
+		CHECK_FOR_INTERRUPTS();
+
 		end_posn = text_position_next(start_posn, &state);
 
 		if (end_posn == 0)