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

Commitad004ce

Browse files
committed
Change ILIKE to invoke lower() and then do plain LIKE comparison when
working in a multibyte encoding. This fixes the problems exhibited inbug #1931 and other reports of ILIKE misbehavior in UTF8 encoding.It's a pretty grotty solution though --- should rethink how to do itafter we install better locale support, someday.
1 parent6dc920d commitad004ce

File tree

1 file changed

+100
-63
lines changed

1 file changed

+100
-63
lines changed

‎src/backend/utils/adt/like.c

Lines changed: 100 additions & 63 deletions
Original file line numberDiff line numberDiff line change
@@ -11,7 +11,7 @@
1111
* Portions Copyright (c) 1994, Regents of the University of California
1212
*
1313
* IDENTIFICATION
14-
*$PostgreSQL: pgsql/src/backend/utils/adt/like.c,v 1.64 2006/03/05 15:58:42 momjian Exp $
14+
*$PostgreSQL: pgsql/src/backend/utils/adt/like.c,v 1.65 2006/09/04 18:32:55 tgl Exp $
1515
*
1616
*-------------------------------------------------------------------------
1717
*/
@@ -64,50 +64,23 @@ wchareq(char *p1, char *p2)
6464
return1;
6565
}
6666

67-
/*--------------------
68-
* Support routine for MatchTextIC. Compares given multibyte streams
69-
* as wide characters ignoring case.
70-
* If they match, returns 1 otherwise returns 0.
71-
*--------------------
67+
/*
68+
* Formerly we had a routine iwchareq() here that tried to do case-insensitive
69+
* comparison of multibyte characters. It did not work at all, however,
70+
* because it relied on tolower() which has a single-byte API ... and
71+
* towlower() wouldn't be much better since we have no suitably cheap way
72+
* of getting a single character transformed to the system's wchar_t format.
73+
* So now, we just downcase the strings using lower() and apply regular LIKE
74+
* comparison. This should be revisited when we install better locale support.
75+
*
76+
* Note that MBMatchText and MBMatchTextIC do exactly the same thing now.
77+
* Is it worth refactoring to avoid duplicated code? They might become
78+
* different again in the future.
7279
*/
73-
#defineCHARMAX 0x80
74-
75-
staticint
76-
iwchareq(char*p1,char*p2)
77-
{
78-
pg_wcharc1[2],
79-
c2[2];
80-
intl;
81-
82-
/*
83-
* short cut. if *p1 and *p2 is lower than CHARMAX, then we could assume
84-
* they are ASCII
85-
*/
86-
if ((unsignedchar)*p1<CHARMAX&& (unsignedchar)*p2<CHARMAX)
87-
return (tolower((unsignedchar)*p1)==tolower((unsignedchar)*p2));
88-
89-
/*
90-
* if one of them is an ASCII while the other is not, then they must be
91-
* different characters
92-
*/
93-
elseif ((unsignedchar)*p1<CHARMAX|| (unsignedchar)*p2<CHARMAX)
94-
return0;
95-
96-
/*
97-
* ok, p1 and p2 are both > CHARMAX, then they must be multibyte
98-
* characters
99-
*/
100-
l=pg_mblen(p1);
101-
(void)pg_mb2wchar_with_len(p1,c1,l);
102-
c1[0]=tolower(c1[0]);
103-
l=pg_mblen(p2);
104-
(void)pg_mb2wchar_with_len(p2,c2,l);
105-
c2[0]=tolower(c2[0]);
106-
return (c1[0]==c2[0]);
107-
}
10880

81+
/* Set up to compile like_match.c for multibyte characters */
10982
#defineCHAREQ(p1,p2) wchareq(p1, p2)
110-
#defineICHAREQ(p1,p2)iwchareq(p1, p2)
83+
#defineICHAREQ(p1,p2)wchareq(p1, p2)
11184
#defineNextChar(p,plen) \
11285
do { int __l = pg_mblen(p); (p) +=__l; (plen) -=__l; } while (0)
11386
#defineCopyAdvChar(dst,src,srclen) \
@@ -120,7 +93,9 @@ iwchareq(char *p1, char *p2)
12093
#defineMatchTextMBMatchText
12194
#defineMatchTextIC MBMatchTextIC
12295
#definedo_like_escapeMB_do_like_escape
96+
12397
#include"like_match.c"
98+
12499
#undef CHAREQ
125100
#undef ICHAREQ
126101
#undef NextChar
@@ -129,15 +104,19 @@ iwchareq(char *p1, char *p2)
129104
#undef MatchTextIC
130105
#undef do_like_escape
131106

107+
/* Set up to compile like_match.c for single-byte characters */
132108
#defineCHAREQ(p1,p2) (*(p1) == *(p2))
133109
#defineICHAREQ(p1,p2) (tolower((unsigned char) *(p1)) == tolower((unsigned char) *(p2)))
134110
#defineNextChar(p,plen) ((p)++, (plen)--)
135111
#defineCopyAdvChar(dst,src,srclen) (*(dst)++ = *(src)++, (srclen)--)
136112

113+
#include"like_match.c"
114+
115+
/* And some support for BYTEA */
137116
#defineBYTEA_CHAREQ(p1,p2) (*(p1) == *(p2))
138117
#defineBYTEA_NextChar(p,plen) ((p)++, (plen)--)
139118
#defineBYTEA_CopyAdvChar(dst,src,srclen) (*(dst)++ = *(src)++, (srclen)--)
140-
#include"like_match.c"
119+
141120

142121
/*
143122
*interface routines called by the function manager
@@ -296,15 +275,32 @@ nameiclike(PG_FUNCTION_ARGS)
296275
intslen,
297276
plen;
298277

299-
s=NameStr(*str);
300-
slen=strlen(s);
301-
p=VARDATA(pat);
302-
plen= (VARSIZE(pat)-VARHDRSZ);
303-
304278
if (pg_database_encoding_max_length()==1)
279+
{
280+
s=NameStr(*str);
281+
slen=strlen(s);
282+
p=VARDATA(pat);
283+
plen= (VARSIZE(pat)-VARHDRSZ);
305284
result= (MatchTextIC(s,slen,p,plen)==LIKE_TRUE);
285+
}
306286
else
287+
{
288+
/* Force inputs to lower case to achieve case insensitivity */
289+
text*strtext;
290+
291+
strtext=DatumGetTextP(DirectFunctionCall1(name_text,
292+
NameGetDatum(str)));
293+
strtext=DatumGetTextP(DirectFunctionCall1(lower,
294+
PointerGetDatum(strtext)));
295+
pat=DatumGetTextP(DirectFunctionCall1(lower,
296+
PointerGetDatum(pat)));
297+
298+
s=VARDATA(strtext);
299+
slen= (VARSIZE(strtext)-VARHDRSZ);
300+
p=VARDATA(pat);
301+
plen= (VARSIZE(pat)-VARHDRSZ);
307302
result= (MBMatchTextIC(s,slen,p,plen)==LIKE_TRUE);
303+
}
308304

309305
PG_RETURN_BOOL(result);
310306
}
@@ -320,15 +316,32 @@ nameicnlike(PG_FUNCTION_ARGS)
320316
intslen,
321317
plen;
322318

323-
s=NameStr(*str);
324-
slen=strlen(s);
325-
p=VARDATA(pat);
326-
plen= (VARSIZE(pat)-VARHDRSZ);
327-
328319
if (pg_database_encoding_max_length()==1)
320+
{
321+
s=NameStr(*str);
322+
slen=strlen(s);
323+
p=VARDATA(pat);
324+
plen= (VARSIZE(pat)-VARHDRSZ);
329325
result= (MatchTextIC(s,slen,p,plen)!=LIKE_TRUE);
326+
}
330327
else
328+
{
329+
/* Force inputs to lower case to achieve case insensitivity */
330+
text*strtext;
331+
332+
strtext=DatumGetTextP(DirectFunctionCall1(name_text,
333+
NameGetDatum(str)));
334+
strtext=DatumGetTextP(DirectFunctionCall1(lower,
335+
PointerGetDatum(strtext)));
336+
pat=DatumGetTextP(DirectFunctionCall1(lower,
337+
PointerGetDatum(pat)));
338+
339+
s=VARDATA(strtext);
340+
slen= (VARSIZE(strtext)-VARHDRSZ);
341+
p=VARDATA(pat);
342+
plen= (VARSIZE(pat)-VARHDRSZ);
331343
result= (MBMatchTextIC(s,slen,p,plen)!=LIKE_TRUE);
344+
}
332345

333346
PG_RETURN_BOOL(result);
334347
}
@@ -344,15 +357,27 @@ texticlike(PG_FUNCTION_ARGS)
344357
intslen,
345358
plen;
346359

347-
s=VARDATA(str);
348-
slen= (VARSIZE(str)-VARHDRSZ);
349-
p=VARDATA(pat);
350-
plen= (VARSIZE(pat)-VARHDRSZ);
351-
352360
if (pg_database_encoding_max_length()==1)
361+
{
362+
s=VARDATA(str);
363+
slen= (VARSIZE(str)-VARHDRSZ);
364+
p=VARDATA(pat);
365+
plen= (VARSIZE(pat)-VARHDRSZ);
353366
result= (MatchTextIC(s,slen,p,plen)==LIKE_TRUE);
367+
}
354368
else
369+
{
370+
/* Force inputs to lower case to achieve case insensitivity */
371+
str=DatumGetTextP(DirectFunctionCall1(lower,
372+
PointerGetDatum(str)));
373+
pat=DatumGetTextP(DirectFunctionCall1(lower,
374+
PointerGetDatum(pat)));
375+
s=VARDATA(str);
376+
slen= (VARSIZE(str)-VARHDRSZ);
377+
p=VARDATA(pat);
378+
plen= (VARSIZE(pat)-VARHDRSZ);
355379
result= (MBMatchTextIC(s,slen,p,plen)==LIKE_TRUE);
380+
}
356381

357382
PG_RETURN_BOOL(result);
358383
}
@@ -368,15 +393,27 @@ texticnlike(PG_FUNCTION_ARGS)
368393
intslen,
369394
plen;
370395

371-
s=VARDATA(str);
372-
slen= (VARSIZE(str)-VARHDRSZ);
373-
p=VARDATA(pat);
374-
plen= (VARSIZE(pat)-VARHDRSZ);
375-
376396
if (pg_database_encoding_max_length()==1)
397+
{
398+
s=VARDATA(str);
399+
slen= (VARSIZE(str)-VARHDRSZ);
400+
p=VARDATA(pat);
401+
plen= (VARSIZE(pat)-VARHDRSZ);
377402
result= (MatchTextIC(s,slen,p,plen)!=LIKE_TRUE);
403+
}
378404
else
405+
{
406+
/* Force inputs to lower case to achieve case insensitivity */
407+
str=DatumGetTextP(DirectFunctionCall1(lower,
408+
PointerGetDatum(str)));
409+
pat=DatumGetTextP(DirectFunctionCall1(lower,
410+
PointerGetDatum(pat)));
411+
s=VARDATA(str);
412+
slen= (VARSIZE(str)-VARHDRSZ);
413+
p=VARDATA(pat);
414+
plen= (VARSIZE(pat)-VARHDRSZ);
379415
result= (MBMatchTextIC(s,slen,p,plen)!=LIKE_TRUE);
416+
}
380417

381418
PG_RETURN_BOOL(result);
382419
}

0 commit comments

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