Skip to content

Commit 53874c5

Browse files
deanrasheeddeanrasheed
committed
Add pg_size_bytes() to parse human-readable size strings.
This will parse strings in the format produced by pg_size_pretty() and return sizes in bytes. This allows queries to be written with clauses like "pg_total_relation_size(oid) > pg_size_bytes('10 GB')". Author: Pavel Stehule with various improvements by Vitaly Burovoy Discussion: http://www.postgresql.org/message-id/CAFj8pRD-tGoDKnxdYgECzA4On01_uRqPrwF-8LdkSE-6bDHp0w@mail.gmail.com Reviewed-by: Vitaly Burovoy, Oleksandr Shulgin, Kyotaro Horiguchi, Michael Paquier and Robert Haas
1 parent 091b605 commit 53874c5

File tree

7 files changed

+331
-3
lines changed

7 files changed

+331
-3
lines changed

‎doc/src/sgml/func.sgml

Lines changed: 30 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -17766,6 +17766,9 @@ postgres=# SELECT * FROM pg_xlogfile_name_offset(pg_stop_backup());
1776617766
<indexterm>
1776717767
<primary>pg_relation_size</primary>
1776817768
</indexterm>
17769+
<indexterm>
17770+
<primary>pg_size_bytes</primary>
17771+
</indexterm>
1776917772
<indexterm>
1777017773
<primary>pg_size_pretty</primary>
1777117774
</indexterm>
@@ -17836,6 +17839,15 @@ postgres=# SELECT * FROM pg_xlogfile_name_offset(pg_stop_backup());
1783617839
Shorthand for <literal>pg_relation_size(..., 'main')</literal>
1783717840
</entry>
1783817841
</row>
17842+
<row>
17843+
<entry>
17844+
<literal><function>pg_size_bytes(<type>text</type>)</function></literal>
17845+
</entry>
17846+
<entry><type>bigint</type></entry>
17847+
<entry>
17848+
Converts a size in human-readable format with size units into bytes
17849+
</entry>
17850+
</row>
1783917851
<row>
1784017852
<entry>
1784117853
<literal><function>pg_size_pretty(<type>bigint</type>)</function></literal>
@@ -17968,10 +17980,26 @@ postgres=# SELECT * FROM pg_xlogfile_name_offset(pg_stop_backup());
1796817980

1796917981
<para>
1797017982
<function>pg_size_pretty</> can be used to format the result of one of
17971-
the other functions in a human-readable way, using kB, MB, GB or TB as
17972-
appropriate.
17983+
the other functions in a human-readable way, using bytes, kB, MB, GB or TB
17984+
as appropriate.
1797317985
</para>
1797417986

17987+
<para>
17988+
<function>pg_size_bytes</> can be used to get the size in bytes from a
17989+
string in human-readable format. The input may have units of bytes, kB,
17990+
MB, GB or TB, and is parsed case-insensitively. If no units are specified,
17991+
bytes are assumed.
17992+
</para>
17993+
17994+
<note>
17995+
<para>
17996+
The units kB, MB, GB and TB used by the functions
17997+
<function>pg_size_pretty</> and <function>pg_size_bytes</> are defined
17998+
using powers of 2 rather than powers of 10, so 1kB is 1024 bytes, 1MB is
17999+
1024<superscript>2</> = 1048576 bytes, and so on.
18000+
</para>
18001+
</note>
18002+
1797518003
<para>
1797618004
The functions above that operate on tables or indexes accept a
1797718005
<type>regclass</> argument, which is simply the OID of the table or index

‎src/backend/utils/adt/dbsize.c

Lines changed: 149 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -699,6 +699,155 @@ pg_size_pretty_numeric(PG_FUNCTION_ARGS)
699699
PG_RETURN_TEXT_P(cstring_to_text(result));
700700
}
701701

702+
/*
703+
* Convert a human-readable size to a size in bytes
704+
*/
705+
Datum
706+
pg_size_bytes(PG_FUNCTION_ARGS)
707+
{
708+
text *arg = PG_GETARG_TEXT_PP(0);
709+
char *str,
710+
*strptr,
711+
*endptr;
712+
char saved_char;
713+
Numeric num;
714+
int64 result;
715+
bool have_digits = false;
716+
717+
str = text_to_cstring(arg);
718+
719+
/* Skip leading whitespace */
720+
strptr = str;
721+
while (isspace((unsigned char) *strptr))
722+
strptr++;
723+
724+
/* Check that we have a valid number and determine where it ends */
725+
endptr = strptr;
726+
727+
/* Part (1): sign */
728+
if (*endptr == '-' || *endptr == '+')
729+
endptr++;
730+
731+
/* Part (2): main digit string */
732+
if (isdigit((unsigned char) *endptr))
733+
{
734+
have_digits = true;
735+
do
736+
endptr++;
737+
while (isdigit((unsigned char) *endptr));
738+
}
739+
740+
/* Part (3): optional decimal point and fractional digits */
741+
if (*endptr == '.')
742+
{
743+
endptr++;
744+
if (isdigit((unsigned char) *endptr))
745+
{
746+
have_digits = true;
747+
do
748+
endptr++;
749+
while (isdigit((unsigned char) *endptr));
750+
}
751+
}
752+
753+
/* Complain if we don't have a valid number at this point */
754+
if (!have_digits)
755+
ereport(ERROR,
756+
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
757+
errmsg("invalid size: \"%s\"", str)));
758+
759+
/* Part (4): optional exponent */
760+
if (*endptr == 'e' || *endptr == 'E')
761+
{
762+
long exponent;
763+
char *cp;
764+
765+
/*
766+
* Note we might one day support EB units, so if what follows isn't a
767+
* number, just treat it all as a unit to be parsed.
768+
*/
769+
exponent = strtol(endptr + 1, &cp, 10);
770+
if (cp > endptr + 1)
771+
{
772+
if (exponent > NUMERIC_MAX_PRECISION ||
773+
exponent < -NUMERIC_MAX_PRECISION)
774+
ereport(ERROR,
775+
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
776+
errmsg("invalid size: \"%s\"", str)));
777+
endptr = cp;
778+
}
779+
}
780+
781+
/*
782+
* Parse the number, saving the next character, which may be the first
783+
* character of the unit string.
784+
*/
785+
saved_char = *endptr;
786+
*endptr = '\0';
787+
788+
num = DatumGetNumeric(DirectFunctionCall3(numeric_in,
789+
CStringGetDatum(strptr),
790+
ObjectIdGetDatum(InvalidOid),
791+
Int32GetDatum(-1)));
792+
793+
*endptr = saved_char;
794+
795+
/* Skip whitespace between number and unit */
796+
strptr = endptr;
797+
while (isspace((unsigned char) *strptr))
798+
strptr++;
799+
800+
/* Handle possible unit */
801+
if (*strptr != '\0')
802+
{
803+
int64 multiplier = 0;
804+
805+
/* Trim any trailing whitespace */
806+
endptr = str + VARSIZE_ANY_EXHDR(arg) - 1;
807+
808+
while (isspace((unsigned char) *endptr))
809+
endptr--;
810+
811+
endptr++;
812+
*endptr = '\0';
813+
814+
/* Parse the unit case-insensitively */
815+
if (pg_strcasecmp(strptr, "bytes") == 0)
816+
multiplier = 1;
817+
else if (pg_strcasecmp(strptr, "kb") == 0)
818+
multiplier = 1024;
819+
else if (pg_strcasecmp(strptr, "mb") == 0)
820+
multiplier = 1024 * 1024;
821+
else if (pg_strcasecmp(strptr, "gb") == 0)
822+
multiplier = 1024 * 1024 * 1024;
823+
else if (pg_strcasecmp(strptr, "tb") == 0)
824+
multiplier = 1024 * 1024 * 1024 * 1024L;
825+
else
826+
ereport(ERROR,
827+
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
828+
errmsg("invalid size: \"%s\"", text_to_cstring(arg)),
829+
errdetail("Invalid size unit: \"%s\".", strptr),
830+
errhint("Valid units are \"bytes\", \"kB\", \"MB\", \"GB\", and \"TB\".")));
831+
832+
if (multiplier > 1)
833+
{
834+
Numeric mul_num;
835+
836+
mul_num = DatumGetNumeric(DirectFunctionCall1(int8_numeric,
837+
Int64GetDatum(multiplier)));
838+
839+
num = DatumGetNumeric(DirectFunctionCall2(numeric_mul,
840+
NumericGetDatum(mul_num),
841+
NumericGetDatum(num)));
842+
}
843+
}
844+
845+
result = DatumGetInt64(DirectFunctionCall1(numeric_int8,
846+
NumericGetDatum(num)));
847+
848+
PG_RETURN_INT64(result);
849+
}
850+
702851
/*
703852
* Get the filenode of a relation
704853
*

‎src/include/catalog/catversion.h

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -53,6 +53,6 @@
5353
*/
5454

5555
/* yyyymmddN */
56-
#define CATALOG_VERSION_NO 201602171
56+
#define CATALOG_VERSION_NO 201602201
5757

5858
#endif

‎src/include/catalog/pg_proc.h

Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -3601,6 +3601,8 @@ DATA(insert OID = 2288 ( pg_size_pretty PGNSP PGUID 12 1 0 0 0 f f f f t f i s
36013601
DESCR("convert a long int to a human readable text using size units");
36023602
DATA(insert OID = 3166 ( pg_size_pretty PGNSP PGUID 12 1 0 0 0 f f f f t f i s 1 0 25 "1700" _null_ _null_ _null_ _null_ _null_ pg_size_pretty_numeric _null_ _null_ _null_ ));
36033603
DESCR("convert a numeric to a human readable text using size units");
3604+
DATA(insert OID = 3334 ( pg_size_bytes PGNSP PGUID 12 1 0 0 0 f f f f t f i s 1 0 20 "25" _null_ _null_ _null_ _null_ _null_ pg_size_bytes _null_ _null_ _null_ ));
3605+
DESCR("convert a size in human-readable format with size units into bytes");
36043606
DATA(insert OID = 2997 ( pg_table_size PGNSP PGUID 12 1 0 0 0 f f f f t f v s 1 0 20 "2205" _null_ _null_ _null_ _null_ _null_ pg_table_size _null_ _null_ _null_ ));
36053607
DESCR("disk space usage for the specified table, including TOAST, free space and visibility map");
36063608
DATA(insert OID = 2998 ( pg_indexes_size PGNSP PGUID 12 1 0 0 0 f f f f t f v s 1 0 20 "2205" _null_ _null_ _null_ _null_ _null_ pg_indexes_size _null_ _null_ _null_ ));

‎src/include/utils/builtins.h

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -473,6 +473,7 @@ extern Datum pg_relation_size(PG_FUNCTION_ARGS);
473473
extern Datum pg_total_relation_size(PG_FUNCTION_ARGS);
474474
extern Datum pg_size_pretty(PG_FUNCTION_ARGS);
475475
extern Datum pg_size_pretty_numeric(PG_FUNCTION_ARGS);
476+
extern Datum pg_size_bytes(PG_FUNCTION_ARGS);
476477
extern Datum pg_table_size(PG_FUNCTION_ARGS);
477478
extern Datum pg_indexes_size(PG_FUNCTION_ARGS);
478479
extern Datum pg_relation_filenode(PG_FUNCTION_ARGS);

‎src/test/regress/expected/dbsize.out

Lines changed: 109 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -35,3 +35,112 @@ SELECT size, pg_size_pretty(size), pg_size_pretty(-1 * size) FROM
3535
1000000000000000.5 | 909 TB | -909 TB
3636
(12 rows)
3737

38+
SELECT size, pg_size_bytes(size) FROM
39+
(VALUES ('1'), ('123bytes'), ('1kB'), ('1MB'), (' 1 GB'), ('1.5 GB '),
40+
('1TB'), ('3000 TB'), ('1e6 MB')) x(size);
41+
size | pg_size_bytes
42+
----------+------------------
43+
1 | 1
44+
123bytes | 123
45+
1kB | 1024
46+
1MB | 1048576
47+
1 GB | 1073741824
48+
1.5 GB | 1610612736
49+
1TB | 1099511627776
50+
3000 TB | 3298534883328000
51+
1e6 MB | 1048576000000
52+
(9 rows)
53+
54+
-- case-insensitive units are supported
55+
SELECT size, pg_size_bytes(size) FROM
56+
(VALUES ('1'), ('123bYteS'), ('1kb'), ('1mb'), (' 1 Gb'), ('1.5 gB '),
57+
('1tb'), ('3000 tb'), ('1e6 mb')) x(size);
58+
size | pg_size_bytes
59+
----------+------------------
60+
1 | 1
61+
123bYteS | 123
62+
1kb | 1024
63+
1mb | 1048576
64+
1 Gb | 1073741824
65+
1.5 gB | 1610612736
66+
1tb | 1099511627776
67+
3000 tb | 3298534883328000
68+
1e6 mb | 1048576000000
69+
(9 rows)
70+
71+
-- negative numbers are supported
72+
SELECT size, pg_size_bytes(size) FROM
73+
(VALUES ('-1'), ('-123bytes'), ('-1kb'), ('-1mb'), (' -1 Gb'), ('-1.5 gB '),
74+
('-1tb'), ('-3000 TB'), ('-10e-1 MB')) x(size);
75+
size | pg_size_bytes
76+
-----------+-------------------
77+
-1 | -1
78+
-123bytes | -123
79+
-1kb | -1024
80+
-1mb | -1048576
81+
-1 Gb | -1073741824
82+
-1.5 gB | -1610612736
83+
-1tb | -1099511627776
84+
-3000 TB | -3298534883328000
85+
-10e-1 MB | -1048576
86+
(9 rows)
87+
88+
-- different cases with allowed points
89+
SELECT size, pg_size_bytes(size) FROM
90+
(VALUES ('-1.'), ('-1.kb'), ('-1. kb'), ('-0. gb'),
91+
('-.1'), ('-.1kb'), ('-.1 kb'), ('-.0 gb')) x(size);
92+
size | pg_size_bytes
93+
--------+---------------
94+
-1. | -1
95+
-1.kb | -1024
96+
-1. kb | -1024
97+
-0. gb | 0
98+
-.1 | 0
99+
-.1kb | -102
100+
-.1 kb | -102
101+
-.0 gb | 0
102+
(8 rows)
103+
104+
-- invalid inputs
105+
SELECT pg_size_bytes('1 AB');
106+
ERROR: invalid size: "1 AB"
107+
DETAIL: Invalid size unit: "AB".
108+
HINT: Valid units are "bytes", "kB", "MB", "GB", and "TB".
109+
SELECT pg_size_bytes('1 AB A');
110+
ERROR: invalid size: "1 AB A"
111+
DETAIL: Invalid size unit: "AB A".
112+
HINT: Valid units are "bytes", "kB", "MB", "GB", and "TB".
113+
SELECT pg_size_bytes('1 AB A ');
114+
ERROR: invalid size: "1 AB A "
115+
DETAIL: Invalid size unit: "AB A".
116+
HINT: Valid units are "bytes", "kB", "MB", "GB", and "TB".
117+
SELECT pg_size_bytes('9223372036854775807.9');
118+
ERROR: bigint out of range
119+
SELECT pg_size_bytes('1e100');
120+
ERROR: bigint out of range
121+
SELECT pg_size_bytes('1e1000000000000000000');
122+
ERROR: invalid size: "1e1000000000000000000"
123+
SELECT pg_size_bytes('1 byte'); -- the singular "byte" is not supported
124+
ERROR: invalid size: "1 byte"
125+
DETAIL: Invalid size unit: "byte".
126+
HINT: Valid units are "bytes", "kB", "MB", "GB", and "TB".
127+
SELECT pg_size_bytes('');
128+
ERROR: invalid size: ""
129+
SELECT pg_size_bytes('kb');
130+
ERROR: invalid size: "kb"
131+
SELECT pg_size_bytes('..');
132+
ERROR: invalid size: ".."
133+
SELECT pg_size_bytes('-.');
134+
ERROR: invalid size: "-."
135+
SELECT pg_size_bytes('-.kb');
136+
ERROR: invalid size: "-.kb"
137+
SELECT pg_size_bytes('-. kb');
138+
ERROR: invalid size: "-. kb"
139+
SELECT pg_size_bytes('.+912');
140+
ERROR: invalid size: ".+912"
141+
SELECT pg_size_bytes('+912+ kB');
142+
ERROR: invalid size: "+912+ kB"
143+
DETAIL: Invalid size unit: "+ kB".
144+
HINT: Valid units are "bytes", "kB", "MB", "GB", and "TB".
145+
SELECT pg_size_bytes('++123 kB');
146+
ERROR: invalid size: "++123 kB"

‎src/test/regress/sql/dbsize.sql

Lines changed: 39 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -10,3 +10,42 @@ SELECT size, pg_size_pretty(size), pg_size_pretty(-1 * size) FROM
1010
(10.5::numeric), (1000.5::numeric), (1000000.5::numeric),
1111
(1000000000.5::numeric), (1000000000000.5::numeric),
1212
(1000000000000000.5::numeric)) x(size);
13+
14+
SELECT size, pg_size_bytes(size) FROM
15+
(VALUES ('1'), ('123bytes'), ('1kB'), ('1MB'), (' 1 GB'), ('1.5 GB '),
16+
('1TB'), ('3000 TB'), ('1e6 MB')) x(size);
17+
18+
-- case-insensitive units are supported
19+
SELECT size, pg_size_bytes(size) FROM
20+
(VALUES ('1'), ('123bYteS'), ('1kb'), ('1mb'), (' 1 Gb'), ('1.5 gB '),
21+
('1tb'), ('3000 tb'), ('1e6 mb')) x(size);
22+
23+
-- negative numbers are supported
24+
SELECT size, pg_size_bytes(size) FROM
25+
(VALUES ('-1'), ('-123bytes'), ('-1kb'), ('-1mb'), (' -1 Gb'), ('-1.5 gB '),
26+
('-1tb'), ('-3000 TB'), ('-10e-1 MB')) x(size);
27+
28+
-- different cases with allowed points
29+
SELECT size, pg_size_bytes(size) FROM
30+
(VALUES ('-1.'), ('-1.kb'), ('-1. kb'), ('-0. gb'),
31+
('-.1'), ('-.1kb'), ('-.1 kb'), ('-.0 gb')) x(size);
32+
33+
-- invalid inputs
34+
SELECT pg_size_bytes('1 AB');
35+
SELECT pg_size_bytes('1 AB A');
36+
SELECT pg_size_bytes('1 AB A ');
37+
SELECT pg_size_bytes('9223372036854775807.9');
38+
SELECT pg_size_bytes('1e100');
39+
SELECT pg_size_bytes('1e1000000000000000000');
40+
SELECT pg_size_bytes('1 byte'); -- the singular "byte" is not supported
41+
SELECT pg_size_bytes('');
42+
43+
SELECT pg_size_bytes('kb');
44+
SELECT pg_size_bytes('..');
45+
SELECT pg_size_bytes('-.');
46+
SELECT pg_size_bytes('-.kb');
47+
SELECT pg_size_bytes('-. kb');
48+
49+
SELECT pg_size_bytes('.+912');
50+
SELECT pg_size_bytes('+912+ kB');
51+
SELECT pg_size_bytes('++123 kB');

0 commit comments

Comments
 (0)