Skip to content

Commit fc2660f

Browse files
tglsfdctglsfdc
committed
Fix LIKE's special-case code for % followed by _. I'm not entirely sure that
this case is worth a special code path, but a special code path that gets the boundary condition wrong is definitely no good. Per bug #4821 from Andrew Gierth. In passing, clean up some minor code formatting issues (excess parentheses and blank lines in odd places). Back-patch to 8.3, where the bug was introduced.
1 parent c6c458e commit fc2660f

File tree

3 files changed

+96
-62
lines changed

3 files changed

+96
-62
lines changed

‎src/backend/utils/adt/like_match.c‎

Lines changed: 58 additions & 62 deletions
Original file line numberDiff line numberDiff line change
@@ -9,7 +9,7 @@
99
* (UTF8 is a special case because we can use a much more efficient version
1010
* of NextChar than can be used for general multi-byte encodings.)
1111
*
12-
* Before the inclusion, we need to define following macros:
12+
* Before the inclusion, we need to define the following macros:
1313
*
1414
* NextChar
1515
* MatchText - to name of function wanted
@@ -19,47 +19,46 @@
1919
* Copyright (c) 1996-2009, PostgreSQL Global Development Group
2020
*
2121
* IDENTIFICATION
22-
* $PostgreSQL: pgsql/src/backend/utils/adt/like_match.c,v 1.24 2009/01/01 17:23:49 momjian Exp $
22+
* $PostgreSQL: pgsql/src/backend/utils/adt/like_match.c,v 1.25 2009/05/24 18:10:37 tgl Exp $
2323
*
2424
*-------------------------------------------------------------------------
2525
*/
2626

2727
/*
28-
** Originally written by Rich $alz, mirror!rs, Wed Nov 26 19:03:17 EST 1986.
29-
** Rich $alz is now <rsalz@bbn.com>.
30-
** Special thanks to Lars Mathiesen <thorinn@diku.dk> for the LABORT code.
31-
**
32-
** This code was shamelessly stolen from the "pql" code by myself and
33-
** slightly modified :)
34-
**
35-
** All references to the word "star" were replaced by "percent"
36-
** All references to the word "wild" were replaced by "like"
37-
**
38-
** All the nice shell RE matching stuff was replaced by just "_" and "%"
39-
**
40-
** As I don't have a copy of the SQL standard handy I wasn't sure whether
41-
** to leave in the '\' escape character handling.
42-
**
43-
** Keith Parks. <keith@mtcc.demon.co.uk>
44-
**
45-
** SQL92 lets you specify the escape character by saying
46-
** LIKE <pattern> ESCAPE <escape character>. We are a small operation
47-
** so we force you to use '\'. - ay 7/95
48-
**
49-
** Now we have the like_escape() function that converts patterns with
50-
** any specified escape character (or none at all) to the internal
51-
** default escape character, which is still '\'. - tgl 9/2000
52-
**
53-
** The code is rewritten to avoid requiring null-terminated strings,
54-
** which in turn allows us to leave out some memcpy() operations.
55-
** This code should be faster and take less memory, but no promises...
56-
** - thomas 2000-08-06
57-
**
58-
*/
28+
* Originally written by Rich $alz, mirror!rs, Wed Nov 26 19:03:17 EST 1986.
29+
* Rich $alz is now <rsalz@bbn.com>.
30+
* Special thanks to Lars Mathiesen <thorinn@diku.dk> for the LABORT code.
31+
*
32+
* This code was shamelessly stolen from the "pql" code by myself and
33+
* slightly modified :)
34+
*
35+
* All references to the word "star" were replaced by "percent"
36+
* All references to the word "wild" were replaced by "like"
37+
*
38+
* All the nice shell RE matching stuff was replaced by just "_" and "%"
39+
*
40+
* As I don't have a copy of the SQL standard handy I wasn't sure whether
41+
* to leave in the '\' escape character handling.
42+
*
43+
* Keith Parks. <keith@mtcc.demon.co.uk>
44+
*
45+
* SQL92 lets you specify the escape character by saying
46+
* LIKE <pattern> ESCAPE <escape character>. We are a small operation
47+
* so we force you to use '\'. - ay 7/95
48+
*
49+
* Now we have the like_escape() function that converts patterns with
50+
* any specified escape character (or none at all) to the internal
51+
* default escape character, which is still '\'. - tgl 9/2000
52+
*
53+
* The code is rewritten to avoid requiring null-terminated strings,
54+
* which in turn allows us to leave out some memcpy() operations.
55+
* This code should be faster and take less memory, but no promises...
56+
* - thomas 2000-08-06
57+
*/
5958

6059

6160
/*--------------------
62-
* Match text and p, return LIKE_TRUE, LIKE_FALSE, or LIKE_ABORT.
61+
* Match text and pattern, return LIKE_TRUE, LIKE_FALSE, or LIKE_ABORT.
6362
*
6463
* LIKE_TRUE: they match
6564
* LIKE_FALSE: they don't match
@@ -80,19 +79,18 @@ static int
8079
MatchText(char *t, int tlen, char *p, int plen)
8180
{
8281
/* Fast path for match-everything pattern */
83-
if ((plen == 1) && (*p == '%'))
82+
if (plen == 1 && *p == '%')
8483
return LIKE_TRUE;
8584

8685
/*
8786
* In this loop, we advance by char when matching wildcards (and thus on
8887
* recursive entry to this function we are properly char-synced). On other
8988
* occasions it is safe to advance by byte, as the text and pattern will
90-
* be in lockstep. This allows us to perform all comparisons between the
89+
* be in lockstep. This allows us to perform all comparisons between the
9190
* text and pattern on a byte by byte basis, even for multi-byte
9291
* encodings.
9392
*/
94-
95-
while ((tlen > 0) && (plen > 0))
93+
while (tlen > 0 && plen > 0)
9694
{
9795
if (*p == '\\')
9896
{
@@ -116,7 +114,7 @@ MatchText(char *t, int tlen, char *p, int plen)
116114

117115
/* %% is the same as % according to the SQL standard */
118116
/* Advance past all %'s */
119-
while ((plen > 0) && (*p == '%'))
117+
while (plen > 0 && *p == '%')
120118
NextByte(p, plen);
121119
/* Trailing percent matches everything. */
122120
if (plen <= 0)
@@ -127,22 +125,24 @@ MatchText(char *t, int tlen, char *p, int plen)
127125
* rest of the pattern.
128126
*/
129127
if (*p == '_')
130-
131128
{
132129
/* %_ is the same as _% - avoid matching _ repeatedly */
133130

134-
NextChar(t, tlen);
135-
NextByte(p, plen);
136-
137-
if (tlen <= 0)
138-
{
139-
return (plen <= 0) ? LIKE_TRUE : LIKE_ABORT;
140-
}
141-
else if (plen <= 0)
131+
do
142132
{
143-
return LIKE_FALSE;
144-
}
145-
133+
NextChar(t, tlen);
134+
NextByte(p, plen);
135+
} while (tlen > 0 && plen > 0 && *p == '_');
136+
137+
/*
138+
* If we are at the end of the pattern, succeed: % followed
139+
* by n _'s matches any string of at least n characters, and
140+
* we have now found there are at least n characters.
141+
*/
142+
if (plen <= 0)
143+
return LIKE_TRUE;
144+
145+
/* Look for a place that matches the rest of the pattern */
146146
while (tlen > 0)
147147
{
148148
int matched = MatchText(t, tlen, p, plen);
@@ -155,7 +155,6 @@ MatchText(char *t, int tlen, char *p, int plen)
155155
}
156156
else
157157
{
158-
159158
char firstpat = TCHAR(*p);
160159

161160
if (*p == '\\')
@@ -180,7 +179,6 @@ MatchText(char *t, int tlen, char *p, int plen)
180179
}
181180

182181
NextChar(t, tlen);
183-
184182
}
185183
}
186184

@@ -192,20 +190,20 @@ MatchText(char *t, int tlen, char *p, int plen)
192190
}
193191
else if (*p == '_')
194192
{
193+
/* _ matches any single character, and we know there is one */
195194
NextChar(t, tlen);
196195
NextByte(p, plen);
197196
continue;
198197
}
199-
else if (TCHAR(*t) != TCHAR(*p))
198+
else if (TCHAR(*p) != TCHAR(*t))
200199
{
201-
/*
202-
* Not the single-character wildcard and no explicit match? Then
203-
* time to quit...
204-
*/
200+
/* non-wildcard pattern char fails to match text char */
205201
return LIKE_FALSE;
206202
}
207203

208204
/*
205+
* Pattern and text match, so advance.
206+
*
209207
* It is safe to use NextByte instead of NextChar here, even for
210208
* multi-byte character sets, because we are not following immediately
211209
* after a wildcard character. If we are in the middle of a multibyte
@@ -222,9 +220,8 @@ MatchText(char *t, int tlen, char *p, int plen)
222220
if (tlen > 0)
223221
return LIKE_FALSE; /* end of pattern, but not of text */
224222

225-
/* End of input string. Do we have matching pattern remaining? */
226-
while ((plen > 0) && (*p == '%')) /* allow multiple %'s at end of
227-
* pattern */
223+
/* End of text string. Do we have matching pattern remaining? */
224+
while (plen > 0 && *p == '%') /* allow multiple %'s at end of pattern */
228225
NextByte(p, plen);
229226

230227
if (plen <= 0)
@@ -354,5 +351,4 @@ do_like_escape(text *pat, text *esc)
354351

355352
#ifdef MATCH_LOWER
356353
#undef MATCH_LOWER
357-
358354
#endif

‎src/test/regress/expected/strings.out‎

Lines changed: 27 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -902,6 +902,33 @@ SELECT 'Hawkeye' NOT ILIKE 'h%' AS "false";
902902
f
903903
(1 row)
904904

905+
--
906+
-- test %/_ combination cases, cf bug #4821
907+
--
908+
SELECT 'foo' LIKE '_%' as t, 'f' LIKE '_%' as t, '' LIKE '_%' as f;
909+
t | t | f
910+
---+---+---
911+
t | t | f
912+
(1 row)
913+
914+
SELECT 'foo' LIKE '%_' as t, 'f' LIKE '%_' as t, '' LIKE '%_' as f;
915+
t | t | f
916+
---+---+---
917+
t | t | f
918+
(1 row)
919+
920+
SELECT 'foo' LIKE '__%' as t, 'foo' LIKE '___%' as t, 'foo' LIKE '____%' as f;
921+
t | t | f
922+
---+---+---
923+
t | t | f
924+
(1 row)
925+
926+
SELECT 'foo' LIKE '%__' as t, 'foo' LIKE '%___' as t, 'foo' LIKE '%____' as f;
927+
t | t | f
928+
---+---+---
929+
t | t | f
930+
(1 row)
931+
905932
--
906933
-- test implicit type conversion
907934
--

‎src/test/regress/sql/strings.sql‎

Lines changed: 11 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -279,6 +279,17 @@ SELECT 'hawkeye' NOT ILIKE 'H%Eye' AS "false";
279279
SELECT 'Hawkeye' ILIKE 'h%' AS "true";
280280
SELECT 'Hawkeye' NOT ILIKE 'h%' AS "false";
281281

282+
--
283+
-- test %/_ combination cases, cf bug #4821
284+
--
285+
286+
SELECT 'foo' LIKE '_%' as t, 'f' LIKE '_%' as t, '' LIKE '_%' as f;
287+
SELECT 'foo' LIKE '%_' as t, 'f' LIKE '%_' as t, '' LIKE '%_' as f;
288+
289+
SELECT 'foo' LIKE '__%' as t, 'foo' LIKE '___%' as t, 'foo' LIKE '____%' as f;
290+
SELECT 'foo' LIKE '%__' as t, 'foo' LIKE '%___' as t, 'foo' LIKE '%____' as f;
291+
292+
282293
--
283294
-- test implicit type conversion
284295
--

0 commit comments

Comments
 (0)