database/postgres
1
0
Fork 0
mirror of https://github.com/postgres/postgres.git synced 2025-12-22 17:42:17 +03:00
Code Activity

Allow underscores in integer and numeric constants.

Browse Source 
This allows underscores to be used in integer and numeric literals,
and their corresponding type input functions, for visual grouping.
For example:

    1_500_000_000
    3.14159_26535_89793
    0xffff_ffff
    0b_1001_0001

A single underscore is allowed between any 2 digits, or immediately
after the base prefix indicator of non-decimal integers, per SQL:202x
draft.

Peter Eisentraut and Dean Rasheed

Discussion: https://postgr.es/m/84aae844-dc55-a4be-86d9-4f0fa405cc97%40enterprisedb.com
This commit is contained in:
Dean Rasheed
2023-02-04 09:48:51 +00:00
parent 1b6f632a35
commit faff8f8e47
22 changed files with 712 additions and 172 deletions
Download Patch File Download Diff File Expand all files Collapse all files

106
src/backend/utils/adt/numeric.c
View File

@@ -6968,10 +6968,7 @@ set_var_from_str(const char *str, const char *cp,
}
if (!isdigit((unsigned char) *cp))
ereturn(escontext, false,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
errmsg("invalid input syntax for type %s: \"%s\"",
"numeric", str)));
goto invalid_syntax;
decdigits = (unsigned char *) palloc(strlen(cp) + DEC_DIGITS * 2);
@@ -6992,12 +6989,19 @@ set_var_from_str(const char *str, const char *cp,
else if (*cp == '.')
{
if (have_dp)
ereturn(escontext, false,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
errmsg("invalid input syntax for type %s: \"%s\"",
"numeric", str)));
goto invalid_syntax;
have_dp = true;
cp++;
/* decimal point must not be followed by underscore */
if (*cp == '_')
goto invalid_syntax;
}
else if (*cp == '_')
{
/* underscore must be followed by more digits */
cp++;
if (!isdigit((unsigned char) *cp))
goto invalid_syntax;
}
else
break;
@@ -7010,17 +7014,8 @@ set_var_from_str(const char *str, const char *cp,
/* Handle exponent, if any */
if (*cp == 'e' || *cp == 'E')
{
long exponent;
char *endptr;
cp++;
exponent = strtol(cp, &endptr, 10);
if (endptr == cp)
ereturn(escontext, false,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
errmsg("invalid input syntax for type %s: \"%s\"",
"numeric", str)));
cp = endptr;
int64 exponent = 0;
bool neg = false;
/*
* At this point, dweight and dscale can't be more than about
@@ -7030,10 +7025,43 @@ set_var_from_str(const char *str, const char *cp,
* fit in storage format, make_result() will complain about it later;
* for consistency use the same ereport errcode/text as make_result().
*/
if (exponent >= INT_MAX / 2 || exponent <= -(INT_MAX / 2))
ereturn(escontext, false,
(errcode(ERRCODE_NUMERIC_VALUE_OUT_OF_RANGE),
errmsg("value overflows numeric format")));
/* exponent sign */
cp++;
if (*cp == '+')
cp++;
else if (*cp == '-')
{
neg = true;
cp++;
}
/* exponent digits */
if (!isdigit((unsigned char) *cp))
goto invalid_syntax;
while (*cp)
{
if (isdigit((unsigned char) *cp))
{
exponent = exponent * 10 + (*cp++ - '0');
if (exponent > PG_INT32_MAX / 2)
goto out_of_range;
}
else if (*cp == '_')
{
/* underscore must be followed by more digits */
cp++;
if (!isdigit((unsigned char) *cp))
goto invalid_syntax;
}
else
break;
}
if (neg)
exponent = -exponent;
dweight += (int) exponent;
dscale -= (int) exponent;
if (dscale < 0)
@@ -7085,6 +7113,17 @@ set_var_from_str(const char *str, const char *cp,
*endptr = cp;
return true;
out_of_range:
ereturn(escontext, false,
(errcode(ERRCODE_NUMERIC_VALUE_OUT_OF_RANGE),
errmsg("value overflows numeric format")));
invalid_syntax:
ereturn(escontext, false,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
errmsg("invalid input syntax for type %s: \"%s\"",
"numeric", str)));
}
@@ -7167,6 +7206,13 @@ set_var_from_non_decimal_integer_str(const char *str, const char *cp, int sign,
tmp = tmp * 16 + xdigit_value(*cp++);
mul = mul * 16;
}
else if (*cp == '_')
{
/* Underscore must be followed by more digits */
cp++;
if (!isxdigit((unsigned char) *cp))
goto invalid_syntax;
}
else
break;
}
@@ -7197,6 +7243,13 @@ set_var_from_non_decimal_integer_str(const char *str, const char *cp, int sign,
tmp = tmp * 8 + (*cp++ - '0');
mul = mul * 8;
}
else if (*cp == '_')
{
/* Underscore must be followed by more digits */
cp++;
if (*cp < '0' || *cp > '7')
goto invalid_syntax;
}
else
break;
}
@@ -7227,6 +7280,13 @@ set_var_from_non_decimal_integer_str(const char *str, const char *cp, int sign,
tmp = tmp * 2 + (*cp++ - '0');
mul = mul * 2;
}
else if (*cp == '_')
{
/* Underscore must be followed by more digits */
cp++;
if (*cp < '0' || *cp > '1')
goto invalid_syntax;
}
else
break;
}

249
src/backend/utils/adt/numutils.c
View File

@@ -141,48 +141,99 @@ pg_strtoint16_safe(const char *s, Node *escontext)
{
firstdigit = ptr += 2;
while (*ptr && isxdigit((unsigned char) *ptr))
while (*ptr)
{
if (unlikely(tmp > -(PG_INT16_MIN / 16)))
goto out_of_range;
if (isxdigit((unsigned char) *ptr))
{
if (unlikely(tmp > -(PG_INT16_MIN / 16)))
goto out_of_range;
tmp = tmp * 16 + hexlookup[(unsigned char) *ptr++];
tmp = tmp * 16 + hexlookup[(unsigned char) *ptr++];
}
else if (*ptr == '_')
{
/* underscore must be followed by more digits */
ptr++;
if (*ptr == '\0' || !isxdigit((unsigned char) *ptr))
goto invalid_syntax;
}
else
break;
}
}
else if (ptr[0] == '0' && (ptr[1] == 'o' || ptr[1] == 'O'))
{
firstdigit = ptr += 2;
while (*ptr && (*ptr >= '0' && *ptr <= '7'))
while (*ptr)
{
if (unlikely(tmp > -(PG_INT16_MIN / 8)))
goto out_of_range;
if (*ptr >= '0' && *ptr <= '7')
{
if (unlikely(tmp > -(PG_INT16_MIN / 8)))
goto out_of_range;
tmp = tmp * 8 + (*ptr++ - '0');
tmp = tmp * 8 + (*ptr++ - '0');
}
else if (*ptr == '_')
{
/* underscore must be followed by more digits */
ptr++;
if (*ptr == '\0' || *ptr < '0' || *ptr > '7')
goto invalid_syntax;
}
else
break;
}
}
else if (ptr[0] == '0' && (ptr[1] == 'b' || ptr[1] == 'B'))
{
firstdigit = ptr += 2;
while (*ptr && (*ptr >= '0' && *ptr <= '1'))
while (*ptr)
{
if (unlikely(tmp > -(PG_INT16_MIN / 2)))
goto out_of_range;
if (*ptr >= '0' && *ptr <= '1')
{
if (unlikely(tmp > -(PG_INT16_MIN / 2)))
goto out_of_range;
tmp = tmp * 2 + (*ptr++ - '0');
tmp = tmp * 2 + (*ptr++ - '0');
}
else if (*ptr == '_')
{
/* underscore must be followed by more digits */
ptr++;
if (*ptr == '\0' || *ptr < '0' || *ptr > '1')
goto invalid_syntax;
}
else
break;
}
}
else
{
firstdigit = ptr;
while (*ptr && isdigit((unsigned char) *ptr))
while (*ptr)
{
if (unlikely(tmp > -(PG_INT16_MIN / 10)))
goto out_of_range;
if (isdigit((unsigned char) *ptr))
{
if (unlikely(tmp > -(PG_INT16_MIN / 10)))
goto out_of_range;
tmp = tmp * 10 + (*ptr++ - '0');
tmp = tmp * 10 + (*ptr++ - '0');
}
else if (*ptr == '_')
{
/* underscore may not be first */
if (unlikely(ptr == firstdigit))
goto invalid_syntax;
/* and it must be followed by more digits */
ptr++;
if (*ptr == '\0' || !isdigit((unsigned char) *ptr))
goto invalid_syntax;
}
else
break;
}
}
@@ -268,48 +319,99 @@ pg_strtoint32_safe(const char *s, Node *escontext)
{
firstdigit = ptr += 2;
while (*ptr && isxdigit((unsigned char) *ptr))
while (*ptr)
{
if (unlikely(tmp > -(PG_INT32_MIN / 16)))
goto out_of_range;
if (isxdigit((unsigned char) *ptr))
{
if (unlikely(tmp > -(PG_INT32_MIN / 16)))
goto out_of_range;
tmp = tmp * 16 + hexlookup[(unsigned char) *ptr++];
tmp = tmp * 16 + hexlookup[(unsigned char) *ptr++];
}
else if (*ptr == '_')
{
/* underscore must be followed by more digits */
ptr++;
if (*ptr == '\0' || !isxdigit((unsigned char) *ptr))
goto invalid_syntax;
}
else
break;
}
}
else if (ptr[0] == '0' && (ptr[1] == 'o' || ptr[1] == 'O'))
{
firstdigit = ptr += 2;
while (*ptr && (*ptr >= '0' && *ptr <= '7'))
while (*ptr)
{
if (unlikely(tmp > -(PG_INT32_MIN / 8)))
goto out_of_range;
if (*ptr >= '0' && *ptr <= '7')
{
if (unlikely(tmp > -(PG_INT32_MIN / 8)))
goto out_of_range;
tmp = tmp * 8 + (*ptr++ - '0');
tmp = tmp * 8 + (*ptr++ - '0');
}
else if (*ptr == '_')
{
/* underscore must be followed by more digits */
ptr++;
if (*ptr == '\0' || *ptr < '0' || *ptr > '7')
goto invalid_syntax;
}
else
break;
}
}
else if (ptr[0] == '0' && (ptr[1] == 'b' || ptr[1] == 'B'))
{
firstdigit = ptr += 2;
while (*ptr && (*ptr >= '0' && *ptr <= '1'))
while (*ptr)
{
if (unlikely(tmp > -(PG_INT32_MIN / 2)))
goto out_of_range;
if (*ptr >= '0' && *ptr <= '1')
{
if (unlikely(tmp > -(PG_INT32_MIN / 2)))
goto out_of_range;
tmp = tmp * 2 + (*ptr++ - '0');
tmp = tmp * 2 + (*ptr++ - '0');
}
else if (*ptr == '_')
{
/* underscore must be followed by more digits */
ptr++;
if (*ptr == '\0' || *ptr < '0' || *ptr > '1')
goto invalid_syntax;
}
else
break;
}
}
else
{
firstdigit = ptr;
while (*ptr && isdigit((unsigned char) *ptr))
while (*ptr)
{
if (unlikely(tmp > -(PG_INT32_MIN / 10)))
goto out_of_range;
if (isdigit((unsigned char) *ptr))
{
if (unlikely(tmp > -(PG_INT32_MIN / 10)))
goto out_of_range;
tmp = tmp * 10 + (*ptr++ - '0');
tmp = tmp * 10 + (*ptr++ - '0');
}
else if (*ptr == '_')
{
/* underscore may not be first */
if (unlikely(ptr == firstdigit))
goto invalid_syntax;
/* and it must be followed by more digits */
ptr++;
if (*ptr == '\0' || !isdigit((unsigned char) *ptr))
goto invalid_syntax;
}
else
break;
}
}
@@ -395,48 +497,99 @@ pg_strtoint64_safe(const char *s, Node *escontext)
{
firstdigit = ptr += 2;
while (*ptr && isxdigit((unsigned char) *ptr))
while (*ptr)
{
if (unlikely(tmp > -(PG_INT64_MIN / 16)))
goto out_of_range;
if (isxdigit((unsigned char) *ptr))
{
if (unlikely(tmp > -(PG_INT64_MIN / 16)))
goto out_of_range;
tmp = tmp * 16 + hexlookup[(unsigned char) *ptr++];
tmp = tmp * 16 + hexlookup[(unsigned char) *ptr++];
}
else if (*ptr == '_')
{
/* underscore must be followed by more digits */
ptr++;
if (*ptr == '\0' || !isxdigit((unsigned char) *ptr))
goto invalid_syntax;
}
else
break;
}
}
else if (ptr[0] == '0' && (ptr[1] == 'o' || ptr[1] == 'O'))
{
firstdigit = ptr += 2;
while (*ptr && (*ptr >= '0' && *ptr <= '7'))
while (*ptr)
{
if (unlikely(tmp > -(PG_INT64_MIN / 8)))
goto out_of_range;
if (*ptr >= '0' && *ptr <= '7')
{
if (unlikely(tmp > -(PG_INT64_MIN / 8)))
goto out_of_range;
tmp = tmp * 8 + (*ptr++ - '0');
tmp = tmp * 8 + (*ptr++ - '0');
}
else if (*ptr == '_')
{
/* underscore must be followed by more digits */
ptr++;
if (*ptr == '\0' || *ptr < '0' || *ptr > '7')
goto invalid_syntax;
}
else
break;
}
}
else if (ptr[0] == '0' && (ptr[1] == 'b' || ptr[1] == 'B'))
{
firstdigit = ptr += 2;
while (*ptr && (*ptr >= '0' && *ptr <= '1'))
while (*ptr)
{
if (unlikely(tmp > -(PG_INT64_MIN / 2)))
goto out_of_range;
if (*ptr >= '0' && *ptr <= '1')
{
if (unlikely(tmp > -(PG_INT64_MIN / 2)))
goto out_of_range;
tmp = tmp * 2 + (*ptr++ - '0');
tmp = tmp * 2 + (*ptr++ - '0');
}
else if (*ptr == '_')
{
/* underscore must be followed by more digits */
ptr++;
if (*ptr == '\0' || *ptr < '0' || *ptr > '1')
goto invalid_syntax;
}
else
break;
}
}
else
{
firstdigit = ptr;
while (*ptr && isdigit((unsigned char) *ptr))
while (*ptr)
{
if (unlikely(tmp > -(PG_INT64_MIN / 10)))
goto out_of_range;
if (isdigit((unsigned char) *ptr))
{
if (unlikely(tmp > -(PG_INT64_MIN / 10)))
goto out_of_range;
tmp = tmp * 10 + (*ptr++ - '0');
tmp = tmp * 10 + (*ptr++ - '0');
}
else if (*ptr == '_')
{
/* underscore may not be first */
if (unlikely(ptr == firstdigit))
goto invalid_syntax;
/* and it must be followed by more digits */
ptr++;
if (*ptr == '\0' || !isdigit((unsigned char) *ptr))
goto invalid_syntax;
}
else
break;
}
}