From 2efd906f8f4ae21d76dfa1c9e2e37468dec68560 Mon Sep 17 00:00:00 2001 From: David Christensen Date: Wed, 28 Apr 2021 10:38:37 -0500 Subject: [PATCH 2/2] Expand the supported units in pg_size_bytes to cover all units --- src/backend/utils/adt/dbsize.c | 63 ++++++++++----- src/include/catalog/pg_proc.dat | 2 +- src/test/regress/expected/dbsize.out | 117 ++++++++++++++++----------- src/test/regress/sql/dbsize.sql | 14 ++-- 4 files changed, 120 insertions(+), 76 deletions(-) diff --git a/src/backend/utils/adt/dbsize.c b/src/backend/utils/adt/dbsize.c index f25ec5c1af..ffa6488da8 100644 --- a/src/backend/utils/adt/dbsize.c +++ b/src/backend/utils/adt/dbsize.c @@ -708,7 +708,6 @@ pg_size_bytes(PG_FUNCTION_ARGS) *endptr; char saved_char; Numeric num; - int64 result; bool have_digits = false; str = text_to_cstring(arg); @@ -760,8 +759,9 @@ pg_size_bytes(PG_FUNCTION_ARGS) char *cp; /* - * Note we might one day support EB units, so if what follows 'E' - * isn't a number, just treat it all as a unit to be parsed. + * If what follows 'e' isn't a number, we just treat it all as a unit + * to be parsed; this allows us to support both exponential notation + * and EB units. */ exponent = strtol(endptr + 1, &cp, 10); (void) exponent; /* Silence -Wunused-result warnings */ @@ -791,7 +791,20 @@ pg_size_bytes(PG_FUNCTION_ARGS) /* Handle possible unit */ if (*strptr != '\0') { - int64 multiplier = 0; + int64 multiplier = 1; + int i; + int unit_count = 9; /* sizeof units table */ + char *units[] = { + "bytes", + "kb", + "mb", + "gb", + "tb", + "pb", + "eb", + "zb", + "yb", + }; /* Trim any trailing whitespace */ endptr = str + VARSIZE_ANY_EXHDR(arg) - 1; @@ -802,26 +815,26 @@ pg_size_bytes(PG_FUNCTION_ARGS) endptr++; *endptr = '\0'; - /* Parse the unit case-insensitively */ - if (pg_strcasecmp(strptr, "bytes") == 0) - multiplier = (int64) 1; - else if (pg_strcasecmp(strptr, "kb") == 0) - multiplier = (int64) 1024; - else if (pg_strcasecmp(strptr, "mb") == 0) - multiplier = ((int64) 1024) * 1024; - - else if (pg_strcasecmp(strptr, "gb") == 0) - multiplier = ((int64) 1024) * 1024 * 1024; - - else if (pg_strcasecmp(strptr, "tb") == 0) - multiplier = ((int64) 1024) * 1024 * 1024 * 1024; + for (i = 0; i < unit_count; i++) { + printf("strptr: %s units: %s", strptr, units[i]); + if (pg_strcasecmp(strptr, units[i]) == 0) + break; + /* + * Note: int64 isn't large enough to store the full multiplier + * going past ~ 9EB, but since this is a fixed value, we can apply + * it twice, thus storing use 2 ** 5 = 32 here, but 2 ** 10 = 1024 + * on actual conversion to numeric. + */ + multiplier *= 32; + } - else + if (i == unit_count) ereport(ERROR, (errcode(ERRCODE_INVALID_PARAMETER_VALUE), errmsg("invalid size: \"%s\"", text_to_cstring(arg)), errdetail("Invalid size unit: \"%s\".", strptr), - errhint("Valid units are \"bytes\", \"kB\", \"MB\", \"GB\", and \"TB\"."))); + errhint("Valid units are \"bytes\", \"kB\", \"MB\", \"GB\", \"TB\", " + "\"PB\", \"EB\", \"ZB\", and \"YB\"."))); if (multiplier > 1) { @@ -832,13 +845,19 @@ pg_size_bytes(PG_FUNCTION_ARGS) num = DatumGetNumeric(DirectFunctionCall2(numeric_mul, NumericGetDatum(mul_num), NumericGetDatum(num))); + + /* second application to get around int64 limitations in unit multipliers */ + num = DatumGetNumeric(DirectFunctionCall2(numeric_mul, + NumericGetDatum(mul_num), + NumericGetDatum(num))); } } - result = DatumGetInt64(DirectFunctionCall1(numeric_int8, - NumericGetDatum(num))); + /* now finally truncate, since this is always in integer-like units */ + num = DatumGetNumeric(DirectFunctionCall1(numeric_ceil, + NumericGetDatum(num))); - PG_RETURN_INT64(result); + PG_RETURN_NUMERIC(num); } /* diff --git a/src/include/catalog/pg_proc.dat b/src/include/catalog/pg_proc.dat index acbcae4607..628dcd69a8 100644 --- a/src/include/catalog/pg_proc.dat +++ b/src/include/catalog/pg_proc.dat @@ -7192,7 +7192,7 @@ prosrc => 'pg_size_pretty_numeric' }, { oid => '3334', descr => 'convert a size in human-readable format with size units into bytes', - proname => 'pg_size_bytes', prorettype => 'int8', proargtypes => 'text', + proname => 'pg_size_bytes', prorettype => 'numeric', proargtypes => 'text', prosrc => 'pg_size_bytes' }, { oid => '2997', descr => 'disk space usage for the specified table, including TOAST, free space and visibility map', diff --git a/src/test/regress/expected/dbsize.out b/src/test/regress/expected/dbsize.out index b93d40fb02..8993522904 100644 --- a/src/test/regress/expected/dbsize.out +++ b/src/test/regress/expected/dbsize.out @@ -64,53 +64,65 @@ SELECT size, pg_size_pretty(size), pg_size_pretty(-1 * size) FROM SELECT size, pg_size_bytes(size) FROM (VALUES ('1'), ('123bytes'), ('1kB'), ('1MB'), (' 1 GB'), ('1.5 GB '), - ('1TB'), ('3000 TB'), ('1e6 MB')) x(size); - size | pg_size_bytes -----------+------------------ - 1 | 1 - 123bytes | 123 - 1kB | 1024 - 1MB | 1048576 - 1 GB | 1073741824 - 1.5 GB | 1610612736 - 1TB | 1099511627776 - 3000 TB | 3298534883328000 - 1e6 MB | 1048576000000 -(9 rows) + ('1TB'), ('3000 TB'), ('1e6 MB'), ('99 PB'), ('45 EB'), ('5.1ZB'), + ('1.17 YB')) x(size); + size | pg_size_bytes +----------+--------------------------- + 1 | 1 + 123bytes | 123 + 1kB | 1024 + 1MB | 1048576 + 1 GB | 1073741824 + 1.5 GB | 1610612736 + 1TB | 1099511627776 + 3000 TB | 3298534883328000 + 1e6 MB | 1048576000000 + 99 PB | 111464090777419776 + 45 EB | 51881467707308113920 + 5.1ZB | 6021017265658797647463 + 1.17 YB | 1414443208949116134406226 +(13 rows) -- case-insensitive units are supported SELECT size, pg_size_bytes(size) FROM (VALUES ('1'), ('123bYteS'), ('1kb'), ('1mb'), (' 1 Gb'), ('1.5 gB '), - ('1tb'), ('3000 tb'), ('1e6 mb')) x(size); - size | pg_size_bytes -----------+------------------ - 1 | 1 - 123bYteS | 123 - 1kb | 1024 - 1mb | 1048576 - 1 Gb | 1073741824 - 1.5 gB | 1610612736 - 1tb | 1099511627776 - 3000 tb | 3298534883328000 - 1e6 mb | 1048576000000 -(9 rows) + ('1tb'), ('3000 tb'), ('1e6 mb'), ('99 pb'), ('45 eB'), ('5.1Zb'), + ('1.17 yb')) x(size); + size | pg_size_bytes +----------+--------------------------- + 1 | 1 + 123bYteS | 123 + 1kb | 1024 + 1mb | 1048576 + 1 Gb | 1073741824 + 1.5 gB | 1610612736 + 1tb | 1099511627776 + 3000 tb | 3298534883328000 + 1e6 mb | 1048576000000 + 99 pb | 111464090777419776 + 45 eB | 51881467707308113920 + 5.1Zb | 6021017265658797647463 + 1.17 yb | 1414443208949116134406226 +(13 rows) -- negative numbers are supported SELECT size, pg_size_bytes(size) FROM (VALUES ('-1'), ('-123bytes'), ('-1kb'), ('-1mb'), (' -1 Gb'), ('-1.5 gB '), - ('-1tb'), ('-3000 TB'), ('-10e-1 MB')) x(size); - size | pg_size_bytes ------------+------------------- - -1 | -1 - -123bytes | -123 - -1kb | -1024 - -1mb | -1048576 - -1 Gb | -1073741824 - -1.5 gB | -1610612736 - -1tb | -1099511627776 - -3000 TB | -3298534883328000 - -10e-1 MB | -1048576 -(9 rows) + ('-1tb'), ('-3000 TB'), ('-10e-1 MB'), ('-19e-4eb'), ('-18YB')) x(size); + size | pg_size_bytes +-----------+----------------------------- + -1 | -1 + -123bytes | -123 + -1kb | -1024 + -1mb | -1048576 + -1 Gb | -1073741824 + -1.5 gB | -1610612736 + -1tb | -1099511627776 + -3000 TB | -3298534883328000 + -10e-1 MB | -1048576 + -19e-4eb | -2190550858753009 + -18YB | -21760664753063325144711168 +(11 rows) -- different cases with allowed points SELECT size, pg_size_bytes(size) FROM @@ -128,29 +140,38 @@ SELECT size, pg_size_bytes(size) FROM -.0 gb | 0 (8 rows) +-- valid inputs outside bigint range (previous errors) +SELECT pg_size_bytes('9223372036854775807.9'); + pg_size_bytes +--------------------- + 9223372036854775808 +(1 row) + +SELECT pg_size_bytes('1e100'); + pg_size_bytes +------------------------------------------------------------------------------------------------------- + 10000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 +(1 row) + -- invalid inputs SELECT pg_size_bytes('1 AB'); ERROR: invalid size: "1 AB" DETAIL: Invalid size unit: "AB". -HINT: Valid units are "bytes", "kB", "MB", "GB", and "TB". +HINT: Valid units are "bytes", "kB", "MB", "GB", "TB", "PB", "EB", "ZB", and "YB". SELECT pg_size_bytes('1 AB A'); ERROR: invalid size: "1 AB A" DETAIL: Invalid size unit: "AB A". -HINT: Valid units are "bytes", "kB", "MB", "GB", and "TB". +HINT: Valid units are "bytes", "kB", "MB", "GB", "TB", "PB", "EB", "ZB", and "YB". SELECT pg_size_bytes('1 AB A '); ERROR: invalid size: "1 AB A " DETAIL: Invalid size unit: "AB A". -HINT: Valid units are "bytes", "kB", "MB", "GB", and "TB". -SELECT pg_size_bytes('9223372036854775807.9'); -ERROR: bigint out of range -SELECT pg_size_bytes('1e100'); -ERROR: bigint out of range +HINT: Valid units are "bytes", "kB", "MB", "GB", "TB", "PB", "EB", "ZB", and "YB". SELECT pg_size_bytes('1e1000000000000000000'); ERROR: value overflows numeric format SELECT pg_size_bytes('1 byte'); -- the singular "byte" is not supported ERROR: invalid size: "1 byte" DETAIL: Invalid size unit: "byte". -HINT: Valid units are "bytes", "kB", "MB", "GB", and "TB". +HINT: Valid units are "bytes", "kB", "MB", "GB", "TB", "PB", "EB", "ZB", and "YB". SELECT pg_size_bytes(''); ERROR: invalid size: "" SELECT pg_size_bytes('kb'); @@ -168,6 +189,6 @@ ERROR: invalid size: ".+912" SELECT pg_size_bytes('+912+ kB'); ERROR: invalid size: "+912+ kB" DETAIL: Invalid size unit: "+ kB". -HINT: Valid units are "bytes", "kB", "MB", "GB", and "TB". +HINT: Valid units are "bytes", "kB", "MB", "GB", "TB", "PB", "EB", "ZB", and "YB". SELECT pg_size_bytes('++123 kB'); ERROR: invalid size: "++123 kB" diff --git a/src/test/regress/sql/dbsize.sql b/src/test/regress/sql/dbsize.sql index a97dacbf7b..2a4abcfc65 100644 --- a/src/test/regress/sql/dbsize.sql +++ b/src/test/regress/sql/dbsize.sql @@ -30,29 +30,33 @@ SELECT size, pg_size_pretty(size), pg_size_pretty(-1 * size) FROM SELECT size, pg_size_bytes(size) FROM (VALUES ('1'), ('123bytes'), ('1kB'), ('1MB'), (' 1 GB'), ('1.5 GB '), - ('1TB'), ('3000 TB'), ('1e6 MB')) x(size); + ('1TB'), ('3000 TB'), ('1e6 MB'), ('99 PB'), ('45 EB'), ('5.1ZB'), + ('1.17 YB')) x(size); -- case-insensitive units are supported SELECT size, pg_size_bytes(size) FROM (VALUES ('1'), ('123bYteS'), ('1kb'), ('1mb'), (' 1 Gb'), ('1.5 gB '), - ('1tb'), ('3000 tb'), ('1e6 mb')) x(size); + ('1tb'), ('3000 tb'), ('1e6 mb'), ('99 pb'), ('45 eB'), ('5.1Zb'), + ('1.17 yb')) x(size); -- negative numbers are supported SELECT size, pg_size_bytes(size) FROM (VALUES ('-1'), ('-123bytes'), ('-1kb'), ('-1mb'), (' -1 Gb'), ('-1.5 gB '), - ('-1tb'), ('-3000 TB'), ('-10e-1 MB')) x(size); + ('-1tb'), ('-3000 TB'), ('-10e-1 MB'), ('-19e-4eb'), ('-18YB')) x(size); -- different cases with allowed points SELECT size, pg_size_bytes(size) FROM (VALUES ('-1.'), ('-1.kb'), ('-1. kb'), ('-0. gb'), ('-.1'), ('-.1kb'), ('-.1 kb'), ('-.0 gb')) x(size); +-- valid inputs outside bigint range (previous errors) +SELECT pg_size_bytes('9223372036854775807.9'); +SELECT pg_size_bytes('1e100'); + -- invalid inputs SELECT pg_size_bytes('1 AB'); SELECT pg_size_bytes('1 AB A'); SELECT pg_size_bytes('1 AB A '); -SELECT pg_size_bytes('9223372036854775807.9'); -SELECT pg_size_bytes('1e100'); SELECT pg_size_bytes('1e1000000000000000000'); SELECT pg_size_bytes('1 byte'); -- the singular "byte" is not supported SELECT pg_size_bytes(''); -- 2.30.1 (Apple Git-130)