regc_pg_locale.c File Reference

#include "catalog/pg_collation.h"
#include "common/unicode_case.h"
#include "common/unicode_category.h"
#include "utils/pg_locale.h"

Include dependency graph for regc_pg_locale.c:

This graph shows which files directly or indirectly include this file:

Go to the source code of this file.

Data Structures
struct	pg_ctype_cache

Macros
#define	PG_ISDIGIT   0x01
#define	PG_ISALPHA   0x02
#define	PG_ISALNUM   (PG_ISDIGIT | PG_ISALPHA)
#define	PG_ISUPPER   0x04
#define	PG_ISLOWER   0x08
#define	PG_ISGRAPH   0x10
#define	PG_ISPRINT   0x20
#define	PG_ISPUNCT   0x40
#define	PG_ISSPACE   0x80

Typedefs
typedef int(*	pg_wc_probefunc) (pg_wchar c)
typedef struct pg_ctype_cache	pg_ctype_cache

Functions
void	pg_set_regex_collation (Oid collation)
static int	pg_wc_isdigit (pg_wchar c)
static int	pg_wc_isalpha (pg_wchar c)
static int	pg_wc_isalnum (pg_wchar c)
static int	pg_wc_isword (pg_wchar c)
static int	pg_wc_isupper (pg_wchar c)
static int	pg_wc_islower (pg_wchar c)
static int	pg_wc_isgraph (pg_wchar c)
static int	pg_wc_isprint (pg_wchar c)
static int	pg_wc_ispunct (pg_wchar c)
static int	pg_wc_isspace (pg_wchar c)
static pg_wchar	pg_wc_toupper (pg_wchar c)
static pg_wchar	pg_wc_tolower (pg_wchar c)
static bool	store_match (pg_ctype_cache *pcc, pg_wchar chr1, int nchrs)
static struct cvec *	pg_ctype_get_cache (pg_wc_probefunc probefunc, int cclasscode)

Variables
static pg_locale_t	pg_regex_locale
static struct pg_locale_struct	dummy_c_locale
static const unsigned char	pg_char_properties [128]
static pg_ctype_cache *	pg_ctype_cache_list = NULL

Macro Definition Documentation

PG_ISALNUM

#define PG_ISALNUM   (PG_ISDIGIT | PG_ISALPHA)

Definition at line 35 of file regc_pg_locale.c.

PG_ISALPHA

#define PG_ISALPHA   0x02

Definition at line 34 of file regc_pg_locale.c.

PG_ISDIGIT

#define PG_ISDIGIT   0x01

Definition at line 33 of file regc_pg_locale.c.

PG_ISGRAPH

#define PG_ISGRAPH   0x10

Definition at line 38 of file regc_pg_locale.c.

PG_ISLOWER

#define PG_ISLOWER   0x08

Definition at line 37 of file regc_pg_locale.c.

PG_ISPRINT

#define PG_ISPRINT   0x20

Definition at line 39 of file regc_pg_locale.c.

PG_ISPUNCT

#define PG_ISPUNCT   0x40

Definition at line 40 of file regc_pg_locale.c.

PG_ISSPACE

#define PG_ISSPACE   0x80

Definition at line 41 of file regc_pg_locale.c.

PG_ISUPPER

#define PG_ISUPPER   0x04

Definition at line 36 of file regc_pg_locale.c.

Typedef Documentation

pg_ctype_cache

typedef struct pg_ctype_cache pg_ctype_cache

pg_wc_probefunc

typedef int(* pg_wc_probefunc) (pg_wchar c)

Definition at line 369 of file regc_pg_locale.c.

Function Documentation

pg_ctype_get_cache()

static struct cvec * pg_ctype_get_cache ( pg_wc_probefunc  probefunc, int  cclasscode  )

static

Definition at line 429 of file regc_pg_locale.c.

{
    pg_ctype_cache *pcc;
    pg_wchar    max_chr;
    pg_wchar    cur_chr;
    int         nmatches;
    chr        *newchrs;
 
    /*
     * Do we already have the answer cached?
     */
    for (pcc = pg_ctype_cache_list; pcc != NULL; pcc = pcc->next)
    {
        if (pcc->probefunc == probefunc &&
            pcc->locale == pg_regex_locale)
            return &pcc->cv;
    }
 
    /*
     * Nope, so initialize some workspace ...
     */
    pcc = (pg_ctype_cache *) malloc(sizeof(pg_ctype_cache));
    if (pcc == NULL)
        return NULL;
    pcc->probefunc = probefunc;
    pcc->locale = pg_regex_locale;
    pcc->cv.nchrs = 0;
    pcc->cv.chrspace = 128;
    pcc->cv.chrs = (chr *) malloc(pcc->cv.chrspace * sizeof(chr));
    pcc->cv.nranges = 0;
    pcc->cv.rangespace = 64;
    pcc->cv.ranges = (chr *) malloc(pcc->cv.rangespace * sizeof(chr) * 2);
    if (pcc->cv.chrs == NULL || pcc->cv.ranges == NULL)
        goto out_of_memory;
    pcc->cv.cclasscode = cclasscode;
 
    /*
     * Decide how many character codes we ought to look through.  In general
     * we don't go past MAX_SIMPLE_CHR; chr codes above that are handled at
     * runtime using the "high colormap" mechanism.  However, in C locale
     * there's no need to go further than 127, and if we only have a 1-byte
     * <ctype.h> API there's no need to go further than that can handle.
     *
     * If it's not MAX_SIMPLE_CHR that's constraining the search, mark the
     * output cvec as not having any locale-dependent behavior, since there
     * will be no need to do any run-time locale checks.  (The #if's here
     * would always be true for production values of MAX_SIMPLE_CHR, but it's
     * useful to allow it to be small for testing purposes.)
     */
    if (pg_regex_locale->ctype_is_c)
    {
#if MAX_SIMPLE_CHR >= 127
        max_chr = (pg_wchar) 127;
        pcc->cv.cclasscode = -1;
#else
        max_chr = (pg_wchar) MAX_SIMPLE_CHR;
#endif
    }
    else
    {
        if (pg_regex_locale->ctype->max_chr != 0 &&
            pg_regex_locale->ctype->max_chr <= MAX_SIMPLE_CHR)
        {
            max_chr = pg_regex_locale->ctype->max_chr;
            pcc->cv.cclasscode = -1;
        }
        else
            max_chr = (pg_wchar) MAX_SIMPLE_CHR;
    }
 
    /*
     * And scan 'em ...
     */
    nmatches = 0;               /* number of consecutive matches */
 
    for (cur_chr = 0; cur_chr <= max_chr; cur_chr++)
    {
        if ((*probefunc) (cur_chr))
            nmatches++;
        else if (nmatches > 0)
        {
            if (!store_match(pcc, cur_chr - nmatches, nmatches))
                goto out_of_memory;
            nmatches = 0;
        }
    }
 
    if (nmatches > 0)
        if (!store_match(pcc, cur_chr - nmatches, nmatches))
            goto out_of_memory;
 
    /*
     * We might have allocated more memory than needed, if so free it
     */
    if (pcc->cv.nchrs == 0)
    {
        free(pcc->cv.chrs);
        pcc->cv.chrs = NULL;
        pcc->cv.chrspace = 0;
    }
    else if (pcc->cv.nchrs < pcc->cv.chrspace)
    {
        newchrs = (chr *) realloc(pcc->cv.chrs,
                                  pcc->cv.nchrs * sizeof(chr));
        if (newchrs == NULL)
            goto out_of_memory;
        pcc->cv.chrs = newchrs;
        pcc->cv.chrspace = pcc->cv.nchrs;
    }
    if (pcc->cv.nranges == 0)
    {
        free(pcc->cv.ranges);
        pcc->cv.ranges = NULL;
        pcc->cv.rangespace = 0;
    }
    else if (pcc->cv.nranges < pcc->cv.rangespace)
    {
        newchrs = (chr *) realloc(pcc->cv.ranges,
                                  pcc->cv.nranges * sizeof(chr) * 2);
        if (newchrs == NULL)
            goto out_of_memory;
        pcc->cv.ranges = newchrs;
        pcc->cv.rangespace = pcc->cv.nranges;
    }
 
    /*
     * Success, link it into cache chain
     */
    pcc->next = pg_ctype_cache_list;
    pg_ctype_cache_list = pcc;
 
    return &pcc->cv;
 
    /*
     * Failure, clean up
     */
out_of_memory:
    free(pcc->cv.chrs);
    free(pcc->cv.ranges);
    free(pcc);
 
    return NULL;
}

References cvec::cclasscode, cvec::chrs, cvec::chrspace, pg_locale_struct::ctype, pg_locale_struct::ctype_is_c, pg_ctype_cache::cv, free, pg_ctype_cache::locale, malloc, ctype_methods::max_chr, MAX_SIMPLE_CHR, cvec::nchrs, pg_ctype_cache::next, cvec::nranges, pg_ctype_cache_list, pg_regex_locale, pg_ctype_cache::probefunc, cvec::ranges, cvec::rangespace, realloc, and store_match().

Referenced by cclasscvec().

pg_set_regex_collation()

void pg_set_regex_collation

(

Oid

collation

)

Definition at line 183 of file regc_pg_locale.c.

{
    pg_locale_t locale = 0;
 
    if (!OidIsValid(collation))
    {
        /*
         * This typically means that the parser could not resolve a conflict
         * of implicit collations, so report it that way.
         */
        ereport(ERROR,
                (errcode(ERRCODE_INDETERMINATE_COLLATION),
                 errmsg("could not determine which collation to use for regular expression"),
                 errhint("Use the COLLATE clause to set the collation explicitly.")));
    }
 
    if (collation == C_COLLATION_OID)
    {
        /*
         * Some callers expect regexes to work for C_COLLATION_OID before
         * catalog access is available, so we can't call
         * pg_newlocale_from_collation().
         */
        locale = &dummy_c_locale;
    }
    else
    {
        locale = pg_newlocale_from_collation(collation);
 
        if (!locale->deterministic)
            ereport(ERROR,
                    (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
                     errmsg("nondeterministic collations are not supported for regular expressions")));
 
        if (locale->ctype_is_c)
        {
            /*
             * C/POSIX collations use this path regardless of database
             * encoding
             */
            locale = &dummy_c_locale;
        }
    }
 
    pg_regex_locale = locale;
}

References dummy_c_locale, ereport, errcode(), errhint(), errmsg(), ERROR, locale, OidIsValid, pg_newlocale_from_collation(), and pg_regex_locale.

Referenced by pg_regcomp(), pg_regexec(), and pg_regprefix().

pg_wc_isalnum()

static int pg_wc_isalnum ( pg_wchar  c )

static

Definition at line 251 of file regc_pg_locale.c.

{
    if (pg_regex_locale->ctype_is_c)
        return (c <= (pg_wchar) 127 &&
                (pg_char_properties[c] & PG_ISALNUM));
    else
        return pg_regex_locale->ctype->wc_isalnum(c, pg_regex_locale);
}

References pg_locale_struct::ctype, pg_locale_struct::ctype_is_c, pg_char_properties, PG_ISALNUM, pg_regex_locale, and ctype_methods::wc_isalnum.

Referenced by cclass_column_index(), cclasscvec(), and pg_wc_isword().

pg_wc_isalpha()

static int pg_wc_isalpha ( pg_wchar  c )

static

Definition at line 241 of file regc_pg_locale.c.

{
    if (pg_regex_locale->ctype_is_c)
        return (c <= (pg_wchar) 127 &&
                (pg_char_properties[c] & PG_ISALPHA));
    else
        return pg_regex_locale->ctype->wc_isalpha(c, pg_regex_locale);
}

References pg_locale_struct::ctype, pg_locale_struct::ctype_is_c, pg_char_properties, PG_ISALPHA, pg_regex_locale, and ctype_methods::wc_isalpha.

Referenced by cclass_column_index(), and cclasscvec().

pg_wc_isdigit()

static int pg_wc_isdigit ( pg_wchar  c )

static

Definition at line 231 of file regc_pg_locale.c.

{
    if (pg_regex_locale->ctype_is_c)
        return (c <= (pg_wchar) 127 &&
                (pg_char_properties[c] & PG_ISDIGIT));
    else
        return pg_regex_locale->ctype->wc_isdigit(c, pg_regex_locale);
}

References pg_locale_struct::ctype, pg_locale_struct::ctype_is_c, pg_char_properties, PG_ISDIGIT, pg_regex_locale, and ctype_methods::wc_isdigit.

Referenced by cclass_column_index(), and cclasscvec().

pg_wc_isgraph()

static int pg_wc_isgraph ( pg_wchar  c )

static

Definition at line 290 of file regc_pg_locale.c.

{
    if (pg_regex_locale->ctype_is_c)
        return (c <= (pg_wchar) 127 &&
                (pg_char_properties[c] & PG_ISGRAPH));
    else
        return pg_regex_locale->ctype->wc_isgraph(c, pg_regex_locale);
}

References pg_locale_struct::ctype, pg_locale_struct::ctype_is_c, pg_char_properties, PG_ISGRAPH, pg_regex_locale, and ctype_methods::wc_isgraph.

Referenced by cclass_column_index(), and cclasscvec().

pg_wc_islower()

static int pg_wc_islower ( pg_wchar  c )

static

Definition at line 280 of file regc_pg_locale.c.

{
    if (pg_regex_locale->ctype_is_c)
        return (c <= (pg_wchar) 127 &&
                (pg_char_properties[c] & PG_ISLOWER));
    else
        return pg_regex_locale->ctype->wc_islower(c, pg_regex_locale);
}

References pg_locale_struct::ctype, pg_locale_struct::ctype_is_c, pg_char_properties, PG_ISLOWER, pg_regex_locale, and ctype_methods::wc_islower.

Referenced by cclass_column_index(), and cclasscvec().

pg_wc_isprint()

static int pg_wc_isprint ( pg_wchar  c )

static

Definition at line 300 of file regc_pg_locale.c.

{
    if (pg_regex_locale->ctype_is_c)
        return (c <= (pg_wchar) 127 &&
                (pg_char_properties[c] & PG_ISPRINT));
    else
        return pg_regex_locale->ctype->wc_isprint(c, pg_regex_locale);
}

References pg_locale_struct::ctype, pg_locale_struct::ctype_is_c, pg_char_properties, PG_ISPRINT, pg_regex_locale, and ctype_methods::wc_isprint.

Referenced by cclass_column_index(), and cclasscvec().

pg_wc_ispunct()

static int pg_wc_ispunct ( pg_wchar  c )

static

Definition at line 310 of file regc_pg_locale.c.

{
    if (pg_regex_locale->ctype_is_c)
        return (c <= (pg_wchar) 127 &&
                (pg_char_properties[c] & PG_ISPUNCT));
    else
        return pg_regex_locale->ctype->wc_ispunct(c, pg_regex_locale);
}

References pg_locale_struct::ctype, pg_locale_struct::ctype_is_c, pg_char_properties, PG_ISPUNCT, pg_regex_locale, and ctype_methods::wc_ispunct.

Referenced by cclass_column_index(), and cclasscvec().

pg_wc_isspace()

static int pg_wc_isspace ( pg_wchar  c )

static

Definition at line 320 of file regc_pg_locale.c.

{
    if (pg_regex_locale->ctype_is_c)
        return (c <= (pg_wchar) 127 &&
                (pg_char_properties[c] & PG_ISSPACE));
    else
        return pg_regex_locale->ctype->wc_isspace(c, pg_regex_locale);
}

References pg_locale_struct::ctype, pg_locale_struct::ctype_is_c, pg_char_properties, PG_ISSPACE, pg_regex_locale, and ctype_methods::wc_isspace.

Referenced by cclass_column_index(), and cclasscvec().

pg_wc_isupper()

static int pg_wc_isupper ( pg_wchar  c )

static

Definition at line 270 of file regc_pg_locale.c.

{
    if (pg_regex_locale->ctype_is_c)
        return (c <= (pg_wchar) 127 &&
                (pg_char_properties[c] & PG_ISUPPER));
    else
        return pg_regex_locale->ctype->wc_isupper(c, pg_regex_locale);
}

References pg_locale_struct::ctype, pg_locale_struct::ctype_is_c, pg_char_properties, PG_ISUPPER, pg_regex_locale, and ctype_methods::wc_isupper.

Referenced by cclass_column_index(), and cclasscvec().

pg_wc_isword()

static int pg_wc_isword ( pg_wchar  c )

static

Definition at line 261 of file regc_pg_locale.c.

{
    /* We define word characters as alnum class plus underscore */
    if (c == CHR('_'))
        return 1;
    return pg_wc_isalnum(c);
}

References CHR, and pg_wc_isalnum().

Referenced by cclass_column_index(), and cclasscvec().

pg_wc_tolower()

static pg_wchar pg_wc_tolower ( pg_wchar  c )

static

Definition at line 343 of file regc_pg_locale.c.

{
    if (pg_regex_locale->ctype_is_c)
    {
        if (c <= (pg_wchar) 127)
            return pg_ascii_tolower((unsigned char) c);
        return c;
    }
    else
        return pg_regex_locale->ctype->wc_tolower(c, pg_regex_locale);
}

References pg_locale_struct::ctype, pg_locale_struct::ctype_is_c, pg_ascii_tolower(), pg_regex_locale, and ctype_methods::wc_tolower.

Referenced by allcases(), casecmp(), and range().

pg_wc_toupper()

static pg_wchar pg_wc_toupper ( pg_wchar  c )

static

Definition at line 330 of file regc_pg_locale.c.

{
    if (pg_regex_locale->ctype_is_c)
    {
        if (c <= (pg_wchar) 127)
            return pg_ascii_toupper((unsigned char) c);
        return c;
    }
    else
        return pg_regex_locale->ctype->wc_toupper(c, pg_regex_locale);
}

References pg_locale_struct::ctype, pg_locale_struct::ctype_is_c, pg_ascii_toupper(), pg_regex_locale, and ctype_methods::wc_toupper.

Referenced by allcases(), and range().

store_match()

static bool store_match ( pg_ctype_cache *  pcc, pg_wchar  chr1, int  nchrs  )

static

Definition at line 385 of file regc_pg_locale.c.

{
    chr        *newchrs;
 
    if (nchrs > 1)
    {
        if (pcc->cv.nranges >= pcc->cv.rangespace)
        {
            pcc->cv.rangespace *= 2;
            newchrs = (chr *) realloc(pcc->cv.ranges,
                                      pcc->cv.rangespace * sizeof(chr) * 2);
            if (newchrs == NULL)
                return false;
            pcc->cv.ranges = newchrs;
        }
        pcc->cv.ranges[pcc->cv.nranges * 2] = chr1;
        pcc->cv.ranges[pcc->cv.nranges * 2 + 1] = chr1 + nchrs - 1;
        pcc->cv.nranges++;
    }
    else
    {
        assert(nchrs == 1);
        if (pcc->cv.nchrs >= pcc->cv.chrspace)
        {
            pcc->cv.chrspace *= 2;
            newchrs = (chr *) realloc(pcc->cv.chrs,
                                      pcc->cv.chrspace * sizeof(chr));
            if (newchrs == NULL)
                return false;
            pcc->cv.chrs = newchrs;
        }
        pcc->cv.chrs[pcc->cv.nchrs++] = chr1;
    }
    return true;
}

References assert, cvec::chrs, cvec::chrspace, pg_ctype_cache::cv, cvec::nchrs, cvec::nranges, cvec::ranges, cvec::rangespace, and realloc.

Referenced by pg_ctype_get_cache().

Variable Documentation

dummy_c_locale

struct pg_locale_struct dummy_c_locale

static

Initial value:

= {
    .collate_is_c = true,
    .ctype_is_c = true,
}

Definition at line 25 of file regc_pg_locale.c.

Referenced by pg_set_regex_collation().

pg_char_properties

const unsigned char pg_char_properties[128]

static

Definition at line 43 of file regc_pg_locale.c.

Referenced by pg_wc_isalnum(), pg_wc_isalpha(), pg_wc_isdigit(), pg_wc_isgraph(), pg_wc_islower(), pg_wc_isprint(), pg_wc_ispunct(), pg_wc_isspace(), and pg_wc_isupper().

pg_ctype_cache_list

pg_ctype_cache* pg_ctype_cache_list = NULL

static

Definition at line 379 of file regc_pg_locale.c.

Referenced by pg_ctype_get_cache().

pg_regex_locale

pg_locale_t pg_regex_locale

static

Definition at line 23 of file regc_pg_locale.c.

Referenced by pg_ctype_get_cache(), pg_set_regex_collation(), pg_wc_isalnum(), pg_wc_isalpha(), pg_wc_isdigit(), pg_wc_isgraph(), pg_wc_islower(), pg_wc_isprint(), pg_wc_ispunct(), pg_wc_isspace(), pg_wc_isupper(), pg_wc_tolower(), and pg_wc_toupper().