Type: | Package |
Title: | Log Rotation and Conditional Backups |
Version: | 0.3.7 |
Maintainer: | Stefan Fleck <stefan.b.fleck@gmail.com> |
Description: | Conditionally rotate or back-up files based on their size or the date of the last backup; inspired by the 'Linux' utility 'logrotate'. |
License: | MIT + file LICENSE |
URL: | https://s-fleck.github.io/rotor/ |
BugReports: | https://github.com/s-fleck/rotor/issues |
Imports: | dint, R6, tools |
Suggests: | covr, crayon, data.table, digest, rmarkdown, testthat, uuid, ulid, zip |
Encoding: | UTF-8 |
RoxygenNote: | 7.2.1.9000 |
NeedsCompilation: | no |
Packaged: | 2022-10-11 12:35:11 UTC; stefan.fleck |
Author: | Stefan Fleck |
Repository: | CRAN |
Date/Publication: | 2022-10-11 13:10:02 UTC |
rotor: Log Rotation and Conditional Backups
Description
Conditionally rotate or back-up files based on their size or the date of the last backup; inspired by the 'Linux' utility 'logrotate'.
Author(s)
Maintainer: Stefan Fleck stefan.b.fleck@gmail.com (ORCID)
See Also
Useful links:
An R6 Class for managing backups (abstract base class)
Description
BackupQueue
is an abstract class not intended for direct usage, please refer to
BackupQueueIndex, BackupQueueDateTime, BackupQueueDate instead.
Details
This class is part of the R6 API of rotor which is
intended for developers that want to extend this package. For normal usage,
the simpler functional API is recommended (see rotate()
).
Super class
rotor::DirectoryQueue
-> BackupQueue
Public fields
dir
character
scalar. Directory in which to place the backups.n
integer
scalar. The number of backups that exist forBackupQueue$origin
Active bindings
dir
character
scalar. Directory in which to place the backups.n
integer
scalar. The number of backups that exist forBackupQueue$origin
file
character
scalar. The file to backup/rotate.compression
(Optional) compression to use
compression
argument ofrotate()
.max_backups
Maximum number/size/age of backups. See
max_backups
argument ofrotate()
has_backups
Returns
TRUE
if at least one backup ofBackupQueue$origin
exists All backups of self$origin
Methods
Public methods
Inherited methods
Method new()
Usage
BackupQueue$new( origin, dir = dirname(origin), max_backups = Inf, compression = FALSE, backup_dir = NULL )
Method prune()
Delete all backups except max_backups
. See prune_backups()
.
Usage
BackupQueue$prune(max_backups = self$max_backups)
Method prune_identical()
Delete all identical backups. Uses tools::md5sum()
to
compare the files.
Set the file to be backed up
Usage
BackupQueue$prune_identical()
Method print()
Usage
BackupQueue$print()
Method push_backup()
Usage
BackupQueue$push_backup(...)
Method set_origin()
Usage
BackupQueue$set_origin(x)
Arguments
x
a
character
scalar. Path to a file Set the file to be backed up
Method set_compression()
Usage
BackupQueue$set_compression(x)
Arguments
x
a
character
scalar. Path to a file
Method set_max_backups()
Usage
BackupQueue$set_max_backups(x)
Method set_file()
Usage
BackupQueue$set_file(x)
Method set_backup_dir()
Usage
BackupQueue$set_backup_dir(x)
See Also
Other R6 Classes:
BackupQueueDateTime
,
BackupQueueDate
,
BackupQueueIndex
,
Cache
,
DirectoryQueue
An R6 class for managing datestamped backups
Description
A BackupQueue for date-stamped backups, e.g. foo.log
, foo.2020-07-24.log
Details
This class is part of the R6 API of rotor which is
intended for developers that want to extend this package. For normal usage,
the simpler functional API is recommended (see rotate()
).
Super classes
rotor::DirectoryQueue
-> rotor::BackupQueue
-> rotor::BackupQueueDateTime
-> BackupQueueDate
Methods
Public methods
Inherited methods
rotor::DirectoryQueue$set_dir()
rotor::BackupQueue$print()
rotor::BackupQueue$prune_identical()
rotor::BackupQueue$push_backup()
rotor::BackupQueue$set_backup_dir()
rotor::BackupQueue$set_compression()
rotor::BackupQueue$set_file()
rotor::BackupQueue$set_origin()
rotor::BackupQueueDateTime$prune()
rotor::BackupQueueDateTime$push()
rotor::BackupQueueDateTime$set_cache_backups()
rotor::BackupQueueDateTime$set_max_backups()
rotor::BackupQueueDateTime$should_rotate()
rotor::BackupQueueDateTime$update_backups_cache()
Method new()
Usage
BackupQueueDate$new( origin, dir = dirname(origin), max_backups = Inf, compression = FALSE, fmt = "%Y-%m-%d", cache_backups = FALSE, backup_dir = NULL )
Method set_fmt()
Usage
BackupQueueDate$set_fmt(x)
See Also
Other R6 Classes:
BackupQueueDateTime
,
BackupQueueIndex
,
BackupQueue
,
Cache
,
DirectoryQueue
An R6 class for managing timestamped backups
Description
A BackupQueue for timestamped backups, e.g. foo.log
, foo.2020-07-24_10-54-30.log
Details
This class is part of the R6 API of rotor which is
intended for developers that want to extend this package. For normal usage,
the simpler functional API is recommended (see rotate()
).
Super classes
rotor::DirectoryQueue
-> rotor::BackupQueue
-> BackupQueueDateTime
Active bindings
fmt
See
format
argument ofrotate_date()
logical
scalar. IfTRUE
(the default) the list of backups is cached, ifFALSE
it is read from disk every time this appender triggers. Caching brings a significant speedup for checking whether to rotate or not based on theage
of the last backup, but is only safe if there are no other programs/functions interacting with the backups. This is only advantageous for high frequency file rotation (i.e. several times per second)POSIXct
scalar. Timestamp of the last rotation (the last backup)
Methods
Public methods
Inherited methods
Method new()
Usage
BackupQueueDateTime$new( origin, dir = dirname(origin), max_backups = Inf, compression = FALSE, fmt = "%Y-%m-%d--%H-%M-%S", cache_backups = FALSE, backup_dir = NULL )
Method push()
Create a new time-stamped backup (e.g. ‘logfile.2020-07-22_12-26-29.log’)
Usage
BackupQueueDateTime$push(overwrite = FALSE, now = Sys.time())
Arguments
overwrite
logical
scalar. Overwrite backups with the same filename (i.e timestamp)?now
POSIXct
scalar. Can be used as an override mechanism for the current system time if necessary.
Method prune()
Usage
BackupQueueDateTime$prune(max_backups = self$max_backups)
Method should_rotate()
Should a file of size
and age
be rotated? See size
and age
arguments of rotate_date()
. now
overrides the current system time,
'last_rotation“ overrides the date of the last rotation.
Usage
BackupQueueDateTime$should_rotate( size, age, now = Sys.time(), last_rotation = self$last_rotation %||% file.info(self$origin)$ctime, verbose = FALSE )
Returns
TRUE
or FALSE
Method update_backups_cache()
Force update of the backups cache (only if $cache_backups == TRUE
).
Usage
BackupQueueDateTime$update_backups_cache()
Method set_max_backups()
Usage
BackupQueueDateTime$set_max_backups(x)
Method set_fmt()
Usage
BackupQueueDateTime$set_fmt(x)
Method set_cache_backups()
Usage
BackupQueueDateTime$set_cache_backups(x)
See Also
Other R6 Classes:
BackupQueueDate
,
BackupQueueIndex
,
BackupQueue
,
Cache
,
DirectoryQueue
An R6 class for managing indexed backups
Description
A BackupQueue for indexed backups, e.g. foo.log
, foo.1.log
, foo.2.log
, ...
Details
This class is part of the R6 API of rotor which is
intended for developers that want to extend this package. For normal usage,
the simpler functional API is recommended (see rotate()
).
Super classes
rotor::DirectoryQueue
-> rotor::BackupQueue
-> BackupQueueIndex
Methods
Public methods
Inherited methods
Method push()
Create a new index-stamped backup (e.g. ‘logfile.1.log’)
Usage
BackupQueueIndex$push()
Method prune()
Usage
BackupQueueIndex$prune(max_backups = self$max_backups)
Method prune_identical()
Usage
BackupQueueIndex$prune_identical()
Method should_rotate()
Should a file of size
be rotated? See size
argument of rotate()
Usage
BackupQueueIndex$should_rotate(size, verbose = FALSE)
Returns
TRUE
or FALSE
Method pad_index()
Pad the indices in the filenames of indexed backups to the number of digits of the largest index. Usually does not have to be called manually.
Usage
BackupQueueIndex$pad_index()
Method increment_index()
Increment die Indices of all backups by n
Usually does
not have to be called manually.
Usage
BackupQueueIndex$increment_index(n = 1)
Arguments
n
integer
> 0
See Also
Other R6 Classes:
BackupQueueDateTime
,
BackupQueueDate
,
BackupQueue
,
Cache
,
DirectoryQueue
An R6 class for managing a persistent file-based cache
Description
Cache
provides an R6 API for managing an on-disk key-value
store for R objects. The objects are serialized to a single folder as
.rds files and the key of the object equals the name of the file.
Cache
supports automatic removal of old files if the cache folder exceeds a
predetermined number of files, total size, or if the individual files exceed
a certain age.
Details
This class is part of the R6 API of rotor which is
intended for developers that want to extend this package. For normal usage,
the simpler functional API is recommended (see rotate()
).
Super class
rotor::DirectoryQueue
-> Cache
Public fields
dir
a
character
scalar. path of the directory in which to store the cache filesn
integer
scalar: number of files in the cachemax_files
see the
compress
argument ofbase::saveRDS()
. Note: this differs from the$compress
argument ofrotate()
.max_files
integer
scalar: maximum number of files to keep in the cache
Active bindings
dir
a
character
scalar. path of the directory in which to store the cache filesn
integer
scalar: number of files in the cachemax_files
see the
compress
argument ofbase::saveRDS()
. Note: this differs from the$compress
argument ofrotate()
.max_files
integer
scalar: maximum number of files to keep in the cachemax_size
scalar
integer
,character
orInf
. Delete cached files (starting with the oldest) until the total size of the cache is belowmax_size
.Integers
are interpreted as bytes. You can passcharacter
vectors that contain a file size suffix like1k
(kilobytes),3M
(megabytes),4G
(gigabytes),5T
(terabytes). Instead of these short forms you can also be explicit and use the IEC suffixesKiB
,MiB
,GiB
,TiB
. In Both cases1
kilobyte is1024
bytes, 1megabyte
is1024
kilobytes, etc... .max_age
-
a
Date
scalar: Remove all backups before this datea
character
scalar representing a Date in ISO format (e.g."2019-12-31"
)a
character
scalar representing an Interval in the form"<number> <interval>"
(seerotate()
)
hashfun
NULL
or afunction
to generate a unique hash from the object to be cached (see example). The hash must be a text string that is a valid filename on the target system. If$hashfun
isNULL
, a storage key must be supplied manually incache$push()
. If a new object is added with the same key as an existing object, the existing object will be overwritten without warning. All cached files
Methods
Public methods
Inherited methods
Method new()
Usage
Cache$new( dir = dirname(file), max_files = Inf, max_size = Inf, max_age = Inf, compression = TRUE, hashfun = digest::digest, create_dir = TRUE )
Arguments
create_dir
logical
scalar. IfTRUE
dir
is created if it does not exist.
Examples
td <- file.path(tempdir(), "cache-test") # When using a real hash function as hashfun, identical objects will only # be added to the cache once cache_hash <- Cache$new(td, hashfun = digest::digest) cache_hash$push(iris) cache_hash$push(iris) cache_hash$files cache_hash$purge() # To override this behaviour use a generator for unique ids, such as uuid if (requireNamespace("uuid")){ cache_uid <- Cache$new(td, hashfun = function(x) uuid::UUIDgenerate()) cache_uid$push(iris) cache_uid$push(iris) cache_uid$files cache_uid$purge() } unlink(td, recursive = TRUE)
Method push()
push a new object to the cache
Usage
Cache$push(x, key = self$hashfun(x))
Arguments
x
any R object
key
a
character
scalar. Key under which to store the cached object. Must be a valid filename. Defaults to being generated by$hashfun()
but may also be supplied manually.
Returns
a character
scalar: the key of the newly added object
Method read()
read a cached file
Usage
Cache$read(key)
Arguments
key
character
scalar. key of the cached file to read.
Method remove()
remove a single file from the cache
Usage
Cache$remove(key)
Arguments
key
character
scalar. key of the cached file to remove
Method pop()
Read and remove a single file from the cache
Usage
Cache$pop(key)
Arguments
key
character
scalar. key of the cached file to read/remove
Method prune()
Prune the cache
Delete cached objects that match certain criteria. max_files
and
max_size
deletes the oldest cached objects first; however, this is
dependent on accuracy of the file modification timestamps on your system.
For example, ext3 only supports second-accuracy, and some windows
version only support timestamps at a resolution of two seconds.
If two files have the same timestamp, they are deleted in the lexical
sort order of their key. This means that by using a function that
generates lexically sortable keys as hashfun
(such as
ulid::generate()
) you can enforce the correct deletion order. There
is no such workaround if you use a real hash function.
Usage
Cache$prune( max_files = self$max_files, max_size = self$max_size, max_age = self$max_age, now = Sys.time() )
Arguments
max_files, max_size, max_age
see section Active Bindings.
now
a
POSIXct
datetime scalar. The current time (for max_age)
Method purge()
purge the cache (remove all cached files)
Usage
Cache$purge()
Method destroy()
purge the cache (remove all cached files)
Usage
Cache$destroy()
Method print()
Usage
Cache$print()
Method set_max_files()
Usage
Cache$set_max_files(x)
Method set_max_age()
Usage
Cache$set_max_age(x)
Method set_max_size()
Usage
Cache$set_max_size(x)
Method set_compression()
Usage
Cache$set_compression(x)
Method set_hashfun()
Usage
Cache$set_hashfun(x)
See Also
Other R6 Classes:
BackupQueueDateTime
,
BackupQueueDate
,
BackupQueueIndex
,
BackupQueue
,
DirectoryQueue
Examples
## ------------------------------------------------
## Method `Cache$new`
## ------------------------------------------------
td <- file.path(tempdir(), "cache-test")
# When using a real hash function as hashfun, identical objects will only
# be added to the cache once
cache_hash <- Cache$new(td, hashfun = digest::digest)
cache_hash$push(iris)
cache_hash$push(iris)
cache_hash$files
cache_hash$purge()
# To override this behaviour use a generator for unique ids, such as uuid
if (requireNamespace("uuid")){
cache_uid <- Cache$new(td, hashfun = function(x) uuid::UUIDgenerate())
cache_uid$push(iris)
cache_uid$push(iris)
cache_uid$files
cache_uid$purge()
}
unlink(td, recursive = TRUE)
An R6 class for managing persistent file-based queues (abstract base class)
Description
Abstract class from which all other classes in rotor inherit their basic fields and methods.
Details
This class is part of the R6 API of rotor which is
intended for developers that want to extend this package. For normal usage,
the simpler functional API is recommended (see rotate()
).
Active bindings
dir
a
character
scalar. path of the directory in which to store the cache files
Methods
Public methods
Method new()
Usage
DirectoryQueue$new(...)
Method push()
Usage
DirectoryQueue$push(x, ...)
Method prune()
Usage
DirectoryQueue$prune(x, ...)
Method set_dir()
Usage
DirectoryQueue$set_dir(x, create = TRUE)
See Also
Other R6 Classes:
BackupQueueDateTime
,
BackupQueueDate
,
BackupQueueIndex
,
BackupQueue
,
Cache
Discover existing backups
Description
These function return information on the backups of a file (if any exist)
Usage
backup_info(file, dir = dirname(file))
list_backups(file, dir = dirname(file))
n_backups(file, dir = dirname(file))
newest_backup(file, dir = dirname(file))
oldest_backup(file, dir = dirname(file))
Arguments
file |
|
dir |
|
Value
backup_info()
returns a data.frame
similar to the output of
file.info()
list_backups()
returns the paths to all backups of file
n_backups()
returns the number of backups of file
as an integer
scalar
newest_backup()
and oldest_backup()
return the paths to the
newest or oldest backup of file
(or an empty character
vector if none exist)
Intervals
In rotor, an interval is a character string in the form
"<number> <interval>"
. The following intervals are possible:
"day(s)"
, "week(s)"
, "month(s)"
, "quarter(s)"
, "year(s)"
.
The plural "s"
is optional (so "2 weeks"
and "2 week"
are equivalent).
Please be aware that weeks are
ISOweeks
and start on Monday (not Sunday as in some countries).
Interval strings can be used as arguments when backing up or rotating files, or for pruning backup queues (i.e. limiting the number of backups of a single) file.
When rotating/backing up "1 months"
means "make a new backup if the last
backup is from the preceding month". E.g if the last backup of myfile
is from 2019-02-01
then backup_time(myfile, age = "1 month")
will only
create a backup if the current date is at least 2019-03-01
.
When pruning/limiting backup queues, "1 year"
means "keep at least most
one year worth of backups". So if you call
backup_time(myfile, max_backups = "1 year")
on 2019-03-01
, it will create
a backup and then remove all backups of myfile
before 2019-01-01
.
See Also
Examples
# setup example files
tf <- tempfile("test", fileext = ".rds")
saveRDS(cars, tf)
backup(tf)
backup(tf)
backup_info(tf)
list_backups(tf)
n_backups(tf)
newest_backup(tf)
oldest_backup(tf)
# cleanup
prune_backups(tf, 0)
n_backups(tf)
file.remove(tf)
Rotate or backup files
Description
Functions starting with backup
create backups of a file
, while functions
starting with rotate
do the same but also replace the original file
with an empty one (this is useful for log rotation)
Note:: rotate()
and co will not work reliable on filenames that contain
dots but have no file extension (e.g. my.holiday.picture.jpg
is OK but
my.holiday.picture
is not)
prune_backups()
physically deletes all backups of a file
based on max_backups
prune_backups()
physically deletes all backups of a file
based on max_backups
Usage
rotate(
file,
size = 1,
max_backups = Inf,
compression = FALSE,
dir = dirname(file),
create_file = TRUE,
dry_run = FALSE,
verbose = dry_run
)
backup(
file,
size = 0,
max_backups = Inf,
compression = FALSE,
dir = dirname(file),
dry_run = FALSE,
verbose = dry_run
)
prune_backups(
file,
max_backups,
dir = dirname(file),
dry_run = FALSE,
verbose = dry_run
)
prune_identical_backups(
file,
dir = dirname(file),
dry_run = FALSE,
verbose = dry_run
)
rotate_date(
file,
age = 1,
size = 1,
max_backups = Inf,
compression = FALSE,
format = "%Y-%m-%d",
dir = dirname(file),
overwrite = FALSE,
create_file = TRUE,
now = Sys.Date(),
dry_run = FALSE,
verbose = dry_run
)
backup_date(
file,
age = 1,
size = 1,
max_backups = Inf,
compression = FALSE,
format = "%Y-%m-%d",
dir = dirname(file),
overwrite = FALSE,
now = Sys.Date(),
dry_run = FALSE,
verbose = dry_run
)
rotate_time(
file,
age = -1,
size = 1,
max_backups = Inf,
compression = FALSE,
format = "%Y-%m-%d--%H-%M-%S",
dir = dirname(file),
overwrite = FALSE,
create_file = TRUE,
now = Sys.time(),
dry_run = FALSE,
verbose = dry_run
)
backup_time(
file,
age = -1,
size = 1,
max_backups = Inf,
compression = FALSE,
format = "%Y-%m-%d--%H-%M-%S",
dir = dirname(file),
overwrite = FALSE,
now = Sys.time(),
dry_run = FALSE,
verbose = dry_run
)
Arguments
file |
|
size |
scalar (if |
max_backups |
maximum number of backups to keep
In addition for timestamped backups the following value are supported:
|
compression |
Whether or not backups should be compressed |
dir |
|
create_file |
|
dry_run |
|
verbose |
|
age |
minimum age after which to backup/rotate a file; can be
(if |
format |
a scalar
|
overwrite |
|
now |
The current |
Value
file
as a character
scalar (invisibly)
Side Effects
backup()
, backup_date()
, and backup_time()
may create files (if the
specified conditions are met). They may also delete backups, based on
max_backup
.
rotate()
, rotate_date()
and rotate_time()
do the same, but in
addition delete the input file
, or replace it with an empty file if
create_file == TRUE
(the default).
prune_backups()
may delete files, depending on max_backups
.
prune_backups()
may delete files, depending on max_backups
.
Intervals
In rotor, an interval is a character string in the form
"<number> <interval>"
. The following intervals are possible:
"day(s)"
, "week(s)"
, "month(s)"
, "quarter(s)"
, "year(s)"
.
The plural "s"
is optional (so "2 weeks"
and "2 week"
are equivalent).
Please be aware that weeks are
ISOweeks
and start on Monday (not Sunday as in some countries).
Interval strings can be used as arguments when backing up or rotating files, or for pruning backup queues (i.e. limiting the number of backups of a single) file.
When rotating/backing up "1 months"
means "make a new backup if the last
backup is from the preceding month". E.g if the last backup of myfile
is from 2019-02-01
then backup_time(myfile, age = "1 month")
will only
create a backup if the current date is at least 2019-03-01
.
When pruning/limiting backup queues, "1 year"
means "keep at least most
one year worth of backups". So if you call
backup_time(myfile, max_backups = "1 year")
on 2019-03-01
, it will create
a backup and then remove all backups of myfile
before 2019-01-01
.
See Also
Examples
# setup example file
tf <- tempfile("test", fileext = ".rds")
saveRDS(cars, tf)
# create two backups of `tf``
backup(tf)
backup(tf)
list_backups(tf) # find all backups of a file
# If `size` is set, a backup is only created if the target file is at least
# that big. This is more useful for log rotation than for backups.
backup(tf, size = "100 mb") # no backup becuase `tf` is to small
list_backups(tf)
# If `dry_run` is TRUE, backup() only shows what would happen without
# actually creating or deleting files
backup(tf, size = "0.1kb", dry_run = TRUE)
# rotate() is the same as backup(), but replaces `tf`` with an empty file
rotate(tf)
list_backups(tf)
file.size(tf)
file.size(list_backups(tf))
# prune_backups() can remove old backups
prune_backups(tf, 1) # keep only one backup
list_backups(tf)
# rotate/backup_date() adds a date instead of an index
# you should not mix index backups and timestamp backups
# so we clean up first
prune_backups(tf, 0)
saveRDS(cars, tf)
# backup_date() adds the date instead of an index to the filename
backup_date(tf)
# `age` sets the minimum age of the last backup before creating a new one.
# the example below creates no new backup since it's less than a week
# since the last.
backup_date(tf, age = "1 week")
# `now` overrides the current date.
backup_date(tf, age = "1 year", now = "2999-12-31")
list_backups(tf)
# backup_time() creates backups with a full timestamp
backup_time(tf)
# It's okay to mix backup_date() and backup_time()
list_backups(tf)
# cleanup
prune_backups(tf, 0)
file.remove(tf)
Serialize R objects to disk (with backup)
Description
The rotate_rds*()
functions are wrappers around base::saveRDS() that
create a backup of the destination file (if it exists) instead of just
overwriting it.
Usage
rotate_rds(
object,
file = "",
ascii = FALSE,
version = NULL,
compress = TRUE,
refhook = NULL,
...,
on_change_only = FALSE
)
rotate_rds_date(
object,
file = "",
ascii = FALSE,
version = NULL,
compress = TRUE,
refhook = NULL,
...,
age = -1L,
on_change_only = FALSE
)
rotate_rds_time(
object,
file = "",
ascii = FALSE,
version = NULL,
compress = TRUE,
refhook = NULL,
...,
age = -1L,
on_change_only = FALSE
)
Arguments
object |
R object to serialize. |
file |
a connection or the name of the file where the R object is saved to or read from. |
ascii |
a logical. If |
version |
the workspace format version to use. |
compress |
a logical specifying whether saving to a named file is
to use |
refhook |
a hook function for handling reference objects. |
... |
Arguments passed on to
|
on_change_only |
|
age |
minimum age after which to backup/rotate a file; can be
(if |
Value
the path to file
(invisibly)
Note
The default value for age
is different for rotate_rds_date()
(-1
)
than for rotate_date()
(1
) to make it a bit safer. This means if you
execute rotate_date()
twice on the same file on a given day it will
silently not rotate the file, while rotate_rds_date()
will throw an
error.
Examples
dest <- tempfile()
rotate_rds(iris, dest)
rotate_rds(iris, dest)
rotate_rds(iris, dest)
list_backups(dest)
# cleanup
unlink(list_backups(dest))
unlink(dest)