Skip to content

Commit 769065c

Browse files
tglsfdctglsfdc
committed
Prefer pg_any_to_server/pg_server_to_any over pg_do_encoding_conversion.
A large majority of the callers of pg_do_encoding_conversion were specifying the database encoding as either source or target of the conversion, meaning that we can use the less general functions pg_any_to_server/pg_server_to_any instead. The main advantage of using the latter functions is that they can make use of a cached conversion-function lookup in the common case that the other encoding is the current client_encoding. It's notationally cleaner too in most cases, not least because of the historical artifact that the latter functions use "char *" rather than "unsigned char *" in their APIs. Note that pg_any_to_server will apply an encoding verification step in some cases where pg_do_encoding_conversion would have just done nothing. This seems to me to be a good idea at most of these call sites, though it partially negates the performance benefit. Per discussion of bug #9210.
1 parent 49c817e commit 769065c

File tree

12 files changed

+51
-100
lines changed

12 files changed

+51
-100
lines changed

‎contrib/pg_stat_statements/pg_stat_statements.c

Lines changed: 3 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -1458,11 +1458,9 @@ pg_stat_statements_internal(FunctionCallInfo fcinfo,
14581458
{
14591459
char *enc;
14601460

1461-
enc = (char *)
1462-
pg_do_encoding_conversion((unsigned char *) qstr,
1463-
entry->query_len,
1464-
entry->encoding,
1465-
GetDatabaseEncoding());
1461+
enc = pg_any_to_server(qstr,
1462+
entry->query_len,
1463+
entry->encoding);
14661464

14671465
values[i++] = CStringGetTextDatum(enc);
14681466

‎contrib/sslinfo/sslinfo.c

Lines changed: 2 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -158,10 +158,7 @@ ASN1_STRING_to_text(ASN1_STRING *str)
158158
nullterm = '\0';
159159
BIO_write(membuf, &nullterm, 1);
160160
size = BIO_get_mem_data(membuf, &sp);
161-
dp = (char *) pg_do_encoding_conversion((unsigned char *) sp,
162-
size - 1,
163-
PG_UTF8,
164-
GetDatabaseEncoding());
161+
dp = pg_any_to_server(sp, size - 1, PG_UTF8);
165162
result = cstring_to_text(dp);
166163
if (dp != sp)
167164
pfree(dp);
@@ -323,10 +320,7 @@ X509_NAME_to_text(X509_NAME *name)
323320
nullterm = '\0';
324321
BIO_write(membuf, &nullterm, 1);
325322
size = BIO_get_mem_data(membuf, &sp);
326-
dp = (char *) pg_do_encoding_conversion((unsigned char *) sp,
327-
size - 1,
328-
PG_UTF8,
329-
GetDatabaseEncoding());
323+
dp = pg_any_to_server(sp, size - 1, PG_UTF8);
330324
result = cstring_to_text(dp);
331325
if (dp != sp)
332326
pfree(dp);

‎src/backend/commands/extension.c

Lines changed: 2 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -635,7 +635,6 @@ read_extension_script_file(const ExtensionControlFile *control,
635635
const char *filename)
636636
{
637637
int src_encoding;
638-
int dest_encoding = GetDatabaseEncoding();
639638
bytea *content;
640639
char *src_str;
641640
char *dest_str;
@@ -645,7 +644,7 @@ read_extension_script_file(const ExtensionControlFile *control,
645644

646645
/* use database encoding if not given */
647646
if (control->encoding < 0)
648-
src_encoding = dest_encoding;
647+
src_encoding = GetDatabaseEncoding();
649648
else
650649
src_encoding = control->encoding;
651650

@@ -655,10 +654,7 @@ read_extension_script_file(const ExtensionControlFile *control,
655654
pg_verify_mbstr_len(src_encoding, src_str, len, false);
656655

657656
/* convert the encoding to the database encoding */
658-
dest_str = (char *) pg_do_encoding_conversion((unsigned char *) src_str,
659-
len,
660-
src_encoding,
661-
dest_encoding);
657+
dest_str = pg_any_to_server(src_str, len, src_encoding);
662658

663659
/* if no conversion happened, we have to arrange for null termination */
664660
if (dest_str == src_str)

‎src/backend/snowball/dict_snowball.c

Lines changed: 2 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -255,10 +255,7 @@ dsnowball_lexize(PG_FUNCTION_ARGS)
255255
{
256256
char *recoded;
257257

258-
recoded = (char *) pg_do_encoding_conversion((unsigned char *) txt,
259-
strlen(txt),
260-
GetDatabaseEncoding(),
261-
PG_UTF8);
258+
recoded = pg_server_to_any(txt, strlen(txt), PG_UTF8);
262259
if (recoded != txt)
263260
{
264261
pfree(txt);
@@ -284,10 +281,7 @@ dsnowball_lexize(PG_FUNCTION_ARGS)
284281
{
285282
char *recoded;
286283

287-
recoded = (char *) pg_do_encoding_conversion((unsigned char *) txt,
288-
strlen(txt),
289-
PG_UTF8,
290-
GetDatabaseEncoding());
284+
recoded = pg_any_to_server(txt, strlen(txt), PG_UTF8);
291285
if (recoded != txt)
292286
{
293287
pfree(txt);

‎src/backend/tsearch/ts_locale.c

Lines changed: 1 addition & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -209,10 +209,7 @@ t_readline(FILE *fp)
209209
(void) pg_verify_mbstr(PG_UTF8, buf, len, false);
210210

211211
/* And convert */
212-
recoded = (char *) pg_do_encoding_conversion((unsigned char *) buf,
213-
len,
214-
PG_UTF8,
215-
GetDatabaseEncoding());
212+
recoded = pg_any_to_server(buf, len, PG_UTF8);
216213
if (recoded == buf)
217214
{
218215
/*

‎src/backend/utils/adt/pg_locale.c

Lines changed: 9 additions & 14 deletions
Original file line numberDiff line numberDiff line change
@@ -418,9 +418,7 @@ db_encoding_strdup(int encoding, const char *str)
418418
char *mstr;
419419

420420
/* convert the string to the database encoding */
421-
pstr = (char *) pg_do_encoding_conversion(
422-
(unsigned char *) str, strlen(str),
423-
encoding, GetDatabaseEncoding());
421+
pstr = pg_any_to_server(str, strlen(str), encoding);
424422
mstr = strdup(pstr);
425423
if (pstr != str)
426424
pfree(pstr);
@@ -581,35 +579,32 @@ strftime_win32(char *dst, size_t dstlen, const wchar_t *format, const struct tm
581579
{
582580
size_t len;
583581
wchar_t wbuf[MAX_L10N_DATA];
584-
int encoding;
585-
586-
encoding = GetDatabaseEncoding();
587582

588583
len = wcsftime(wbuf, MAX_L10N_DATA, format, tm);
589584
if (len == 0)
590-
585+
{
591586
/*
592587
* strftime call failed - return 0 with the contents of dst
593588
* unspecified
594589
*/
595590
return 0;
591+
}
596592

597593
len = WideCharToMultiByte(CP_UTF8, 0, wbuf, len, dst, dstlen, NULL, NULL);
598594
if (len == 0)
599-
elog(ERROR,
600-
"could not convert string to UTF-8: error code %lu", GetLastError());
595+
elog(ERROR, "could not convert string to UTF-8: error code %lu",
596+
GetLastError());
601597

602598
dst[len] = '\0';
603-
if (encoding != PG_UTF8)
599+
if (GetDatabaseEncoding() != PG_UTF8)
604600
{
605-
char *convstr =
606-
(char *) pg_do_encoding_conversion((unsigned char *) dst,
607-
len, PG_UTF8, encoding);
601+
char *convstr = pg_any_to_server(dst, len, PG_UTF8);
608602

609-
if (dst != convstr)
603+
if (convstr != dst)
610604
{
611605
strlcpy(dst, convstr, dstlen);
612606
len = strlen(dst);
607+
pfree(convstr);
613608
}
614609
}
615610

‎src/backend/utils/adt/xml.c

Lines changed: 9 additions & 18 deletions
Original file line numberDiff line numberDiff line change
@@ -345,10 +345,7 @@ xml_recv(PG_FUNCTION_ARGS)
345345
xmlFreeDoc(doc);
346346

347347
/* Now that we know what we're dealing with, convert to server encoding */
348-
newstr = (char *) pg_do_encoding_conversion((unsigned char *) str,
349-
nbytes,
350-
encoding,
351-
GetDatabaseEncoding());
348+
newstr = pg_any_to_server(str, nbytes, encoding);
352349

353350
if (newstr != str)
354351
{
@@ -1793,10 +1790,8 @@ sqlchar_to_unicode(char *s)
17931790
char *utf8string;
17941791
pg_wchar ret[2]; /* need space for trailing zero */
17951792

1796-
utf8string = (char *) pg_do_encoding_conversion((unsigned char *) s,
1797-
pg_mblen(s),
1798-
GetDatabaseEncoding(),
1799-
PG_UTF8);
1793+
/* note we're not assuming s is null-terminated */
1794+
utf8string = pg_server_to_any(s, pg_mblen(s), PG_UTF8);
18001795

18011796
pg_encoding_mb2wchar_with_len(PG_UTF8, utf8string, ret,
18021797
pg_encoding_mblen(PG_UTF8, utf8string));
@@ -1892,19 +1887,15 @@ map_sql_identifier_to_xml_name(char *ident, bool fully_escaped,
18921887
static char *
18931888
unicode_to_sqlchar(pg_wchar c)
18941889
{
1895-
unsigned char utf8string[5]; /* need room for trailing zero */
1890+
char utf8string[8]; /* need room for trailing zero */
18961891
char *result;
18971892

18981893
memset(utf8string, 0, sizeof(utf8string));
1899-
unicode_to_utf8(c, utf8string);
1900-
1901-
result = (char *) pg_do_encoding_conversion(utf8string,
1902-
pg_encoding_mblen(PG_UTF8,
1903-
(char *) utf8string),
1904-
PG_UTF8,
1905-
GetDatabaseEncoding());
1906-
/* if pg_do_encoding_conversion didn't strdup, we must */
1907-
if (result == (char *) utf8string)
1894+
unicode_to_utf8(c, (unsigned char *) utf8string);
1895+
1896+
result = pg_any_to_server(utf8string, strlen(utf8string), PG_UTF8);
1897+
/* if pg_any_to_server didn't strdup, we must */
1898+
if (result == utf8string)
19081899
result = pstrdup(result);
19091900
return result;
19101901
}

‎src/backend/utils/mb/mbutils.c

Lines changed: 3 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1077,7 +1077,9 @@ pgwin32_message_to_UTF16(const char *str, int len, int *utf16len)
10771077
char *utf8;
10781078

10791079
utf8 = (char *) pg_do_encoding_conversion((unsigned char *) str,
1080-
len, GetMessageEncoding(), PG_UTF8);
1080+
len,
1081+
GetMessageEncoding(),
1082+
PG_UTF8);
10811083
if (utf8 != str)
10821084
len = strlen(utf8);
10831085

‎src/pl/plperl/plperl.c

Lines changed: 2 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -3811,9 +3811,7 @@ hv_store_string(HV *hv, const char *key, SV *val)
38113811
char *hkey;
38123812
SV **ret;
38133813

3814-
hkey = (char *)
3815-
pg_do_encoding_conversion((unsigned char *) key, strlen(key),
3816-
GetDatabaseEncoding(), PG_UTF8);
3814+
hkey = pg_server_to_any(key, strlen(key), PG_UTF8);
38173815

38183816
/*
38193817
* This seems nowhere documented, but under Perl 5.8.0 and up, hv_store()
@@ -3841,9 +3839,7 @@ hv_fetch_string(HV *hv, const char *key)
38413839
char *hkey;
38423840
SV **ret;
38433841

3844-
hkey = (char *)
3845-
pg_do_encoding_conversion((unsigned char *) key, strlen(key),
3846-
GetDatabaseEncoding(), PG_UTF8);
3842+
hkey = pg_server_to_any(key, strlen(key), PG_UTF8);
38473843

38483844
/* See notes in hv_store_string */
38493845
hlen = -(int) strlen(hkey);

‎src/pl/plperl/plperl_helpers.h

Lines changed: 7 additions & 18 deletions
Original file line numberDiff line numberDiff line change
@@ -9,24 +9,11 @@
99
static inline char *
1010
utf_u2e(char *utf8_str, size_t len)
1111
{
12-
int enc = GetDatabaseEncoding();
1312
char *ret;
1413

15-
/*
16-
* When we are in a PG_UTF8 or SQL_ASCII database
17-
* pg_do_encoding_conversion() will not do any conversion (which is good)
18-
* or verification (not so much), so we need to run the verification step
19-
* separately.
20-
*/
21-
if (enc == PG_UTF8 || enc == PG_SQL_ASCII)
22-
{
23-
pg_verify_mbstr_len(enc, utf8_str, len, false);
24-
ret = utf8_str;
25-
}
26-
else
27-
ret = (char *) pg_do_encoding_conversion((unsigned char *) utf8_str,
28-
len, PG_UTF8, enc);
14+
ret = pg_any_to_server(utf8_str, len, PG_UTF8);
2915

16+
/* ensure we have a copy even if no conversion happened */
3017
if (ret == utf8_str)
3118
ret = pstrdup(ret);
3219

@@ -41,12 +28,14 @@ utf_u2e(char *utf8_str, size_t len)
4128
static inline char *
4229
utf_e2u(const char *str)
4330
{
44-
char *ret =
45-
(char *) pg_do_encoding_conversion((unsigned char *) str, strlen(str),
46-
GetDatabaseEncoding(), PG_UTF8);
31+
char *ret;
4732

33+
ret = pg_server_to_any(str, strlen(str), PG_UTF8);
34+
35+
/* ensure we have a copy even if no conversion happened */
4836
if (ret == str)
4937
ret = pstrdup(ret);
38+
5039
return ret;
5140
}
5241

‎src/pl/plpython/plpy_util.c

Lines changed: 5 additions & 10 deletions
Original file line numberDiff line numberDiff line change
@@ -90,11 +90,9 @@ PLyUnicode_Bytes(PyObject *unicode)
9090
{
9191
PG_TRY();
9292
{
93-
encoded = (char *) pg_do_encoding_conversion(
94-
(unsigned char *) utf8string,
95-
strlen(utf8string),
96-
PG_UTF8,
97-
GetDatabaseEncoding());
93+
encoded = pg_any_to_server(utf8string,
94+
strlen(utf8string),
95+
PG_UTF8);
9896
}
9997
PG_CATCH();
10098
{
@@ -109,7 +107,7 @@ PLyUnicode_Bytes(PyObject *unicode)
109107
/* finally, build a bytes object in the server encoding */
110108
rv = PyBytes_FromStringAndSize(encoded, strlen(encoded));
111109

112-
/* if pg_do_encoding_conversion allocated memory, free it now */
110+
/* if pg_any_to_server allocated memory, free it now */
113111
if (utf8string != encoded)
114112
pfree(encoded);
115113

@@ -149,10 +147,7 @@ PLyUnicode_FromString(const char *s)
149147
char *utf8string;
150148
PyObject *o;
151149

152-
utf8string = (char *) pg_do_encoding_conversion((unsigned char *) s,
153-
strlen(s),
154-
GetDatabaseEncoding(),
155-
PG_UTF8);
150+
utf8string = pg_server_to_any(s, strlen(s), PG_UTF8);
156151

157152
o = PyUnicode_FromString(utf8string);
158153

‎src/pl/tcl/pltcl.c

Lines changed: 6 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -63,13 +63,17 @@
6363
static unsigned char *
6464
utf_u2e(unsigned char *src)
6565
{
66-
return pg_do_encoding_conversion(src, strlen(src), PG_UTF8, GetDatabaseEncoding());
66+
return (unsigned char *) pg_any_to_server((char *) src,
67+
strlen(src),
68+
PG_UTF8);
6769
}
6870

6971
static unsigned char *
7072
utf_e2u(unsigned char *src)
7173
{
72-
return pg_do_encoding_conversion(src, strlen(src), GetDatabaseEncoding(), PG_UTF8);
74+
return (unsigned char *) pg_server_to_any((char *) src,
75+
strlen(src),
76+
PG_UTF8);
7377
}
7478

7579
#define PLTCL_UTF

0 commit comments

Comments
 (0)