Introduce t_isalnum() to replace t_isalpha() || t_isdigit() tests.

ts_locale.c omitted support for "isalnum" tests, perhaps on the
grounds that there were initially no use-cases for that.  However,
both ltree and pg_trgm need such tests, and we do also have one
use-case now in the core backend.  The workaround of testing
isalpha and isdigit separately seems quite inefficient, especially
when dealing with multibyte characters; so let's fill in the
missing support.

Discussion: https://postgr.es/m/2548310.1664999615@sss.pgh.pa.us
pull/104/head
Tom Lane 3 years ago
parent 5757141cae
commit ca71131eeb
  1. 2
      contrib/ltree/ltree.h
  2. 2
      contrib/pg_trgm/trgm.h
  3. 16
      src/backend/tsearch/ts_locale.c
  4. 2
      src/backend/utils/adt/tsquery.c
  5. 1
      src/include/tsearch/ts_locale.h

@ -126,7 +126,7 @@ typedef struct
#define LQUERY_HASNOT 0x01
#define ISALNUM(x) ( t_isalpha(x) || t_isdigit(x) || ( pg_mblen(x) == 1 && t_iseq((x), '_') ) )
#define ISALNUM(x) ( t_isalnum(x) || ( pg_mblen(x) == 1 && t_iseq((x), '_') ) )
/* full text query */

@ -52,7 +52,7 @@ typedef char trgm[3];
} while(0)
#ifdef KEEPONLYALNUM
#define ISWORDCHR(c) (t_isalpha(c) || t_isdigit(c))
#define ISWORDCHR(c) (t_isalnum(c))
#define ISPRINTABLECHAR(a) ( isascii( *(unsigned char*)(a) ) && (isalnum( *(unsigned char*)(a) ) || *(unsigned char*)(a)==' ') )
#else
#define ISWORDCHR(c) (!t_isspace(c))

@ -81,6 +81,22 @@ t_isalpha(const char *ptr)
return iswalpha((wint_t) character[0]);
}
int
t_isalnum(const char *ptr)
{
int clen = pg_mblen(ptr);
wchar_t character[WC_BUF_LEN];
Oid collation = DEFAULT_COLLATION_OID; /* TODO */
pg_locale_t mylocale = 0; /* TODO */
if (clen == 1 || lc_ctype_is_c(collation))
return isalnum(TOUCHAR(ptr));
char2wchar(character, WC_BUF_LEN, ptr, clen, mylocale);
return iswalnum((wint_t) character[0]);
}
int
t_isprint(const char *ptr)
{

@ -248,7 +248,7 @@ parse_or_operator(TSQueryParserState pstate)
return false;
/* it shouldn't be a part of any word */
if (t_iseq(ptr, '-') || t_iseq(ptr, '_') || t_isalpha(ptr) || t_isdigit(ptr))
if (t_iseq(ptr, '-') || t_iseq(ptr, '_') || t_isalnum(ptr))
return false;
for (;;)

@ -42,6 +42,7 @@ typedef struct
extern int t_isdigit(const char *ptr);
extern int t_isspace(const char *ptr);
extern int t_isalpha(const char *ptr);
extern int t_isalnum(const char *ptr);
extern int t_isprint(const char *ptr);
extern char *lowerstr(const char *str);

Loading…
Cancel
Save