1Cache(3) User Contributed Perl Documentation Cache(3)
2
6 Tie::Cache - LRU Cache in Memory
7
9 use Tie::Cache;
10 tie %cache, 'Tie::Cache', 100, { Debug => 1 };
11 tie %cache2, 'Tie::Cache', { MaxCount => 100, MaxBytes => 50000 };
12 tie %cache3, 'Tie::Cache', 100, { Debug => 1 , WriteSync => 0};
13 # Options ##################################################################
15 #
16 # Debug => 0 - DEFAULT, no debugging output
17 # 1 - prints cache statistics upon destroying
18 # 2 - prints detailed debugging info
19 #
20 # MaxCount => Maximum entries in cache.
21 #
22 # MaxBytes => Maximum bytes taken in memory for cache based on approximate
23 # size of total cache structure in memory
24 #
25 # There is approximately 240 bytes used per key/value pair in the cache for
26 # the cache data structures, so a cache of 5000 entries would take
27 # at approximately 1.2M plus the size of the data being cached.
28 #
29 # MaxSize => Maximum size of each cache entry. Larger entries are not cached.
30 # This helps prevent much of the cache being flushed when
31 # you set an exceptionally large entry. Defaults to MaxBytes/10
32 #
33 # WriteSync => 1 - DEFAULT, write() when data is dirtied for
34 # TRUE CACHE (see below)
35 # 0 - write() dirty data as late as possible, when leaving
36 # cache, or when cache is being DESTROY'd
37 #
38 ############################################################################
39 # cache supports normal tied hash functions
41 $cache{1} = 2; # STORE
42 print "$cache{1}\n"; # FETCH
43 # FIRSTKEY, NEXTKEY
45 while(($k, $v) = each %cache) { print "$k: $v\n"; }
46 delete $cache{1}; # DELETE
48 %cache = (); # CLEAR
49
51 This module implements a least recently used (LRU) cache in memory
52 through a tie interface. Any time data is stored in the tied hash,
53 that key/value pair has an entry time associated with it, and as the
54 cache fills up, those members of the cache that are the oldest are
55 removed to make room for new entries.
56 So, the cache only "remembers" the last written entries, up to the size
58 of the cache. This can be especially useful if you access great
59 amounts of data, but only access a minority of the data a majority of
60 the time.
61 The implementation is a hash, for quick lookups, overlaying a doubly
63 linked list for quick insertion and deletion. On a WinNT PII 300,
64 writes to the hash were done at a rate 3100 per second, and reads from
65 the hash at 6300 per second. Work has been done to optimize refreshing
66 cache entries that are frequently read from, code like $cache{entry},
67 which moves the entry to the end of the linked list internally.
68
70 Tie::Cache installs easily using the make or nmake commands as shown
71 below. Otherwise, just copy Cache.pm to $PERLLIB/site/Tie
72 > perl Makefile.PL
74 > make
75 > make test
76 > make install
77 * use nmake for win32
79 ** you can also just copy Cache.pm to $perllib/Tie
80
82 There is another simpler LRU cache implementation in CPAN,
83 Tie::Cache::LRU, which has the same basic size limiting functionality,
84 and for this functionality, the exact same interface.
85 Through healthy competition, Michael G Schwern got Tie::Cache::LRU
87 mostly faster than Tie::Cache on reads & writes:
88 Cache Size 5000 Tie::Cache 0.17 Tie::Cache::LRU 20110205.00
90 10000 Writes 0.63 CPU sec 0.47 CPU sec
91 40000 Reads 0.79 CPU sec 0.71 CPU sec
92 10000 Deletes 0.23 CPU sec 0.26 CPU sec
93 Unless you are using TRUE CACHE or MaxBytes functionality, using
95 Tie::Cache::LRU could be an easy replacement for Tie::Cache.
96 OTOH one nice thing about this module is its lack of external module
98 dependencies!
99
101 To use class as a true cache, which acts as the sole interface for some
102 data set, subclass the real cache off Tie::Cache, with @ISA = qw(
103 'Tie::Cache' ) notation. Then override the read() method for behavior
104 when there is a cache miss, and the write() method for behavior when
105 the cache's data changes.
106 When WriteSync is 1 or TRUE (DEFAULT), write() is called immediately
108 when data in the cache is modified. If set to 0, data that has been
109 modified in the cache gets written out when the entries are deleted or
110 during the DESTROY phase of the cache object, usually at the end of a
111 script.
112 To have the dirty data write() periodically while WriteSync is set to
114 0, there is a flush() cache API call that will flush the dirty writes
115 in this way. Just call the flush() API like:
116 my $write_flush_count = tied(%cache)->flush();
118 The flush() API was added in the .17 release thanks to Rob Bloodgood.
120
122 use Tie::Cache;
123 # personalize the Tie::Cache object, by inheriting from it
125 package My::Cache;
126 @ISA = qw(Tie::Cache);
127 # override the read() and write() member functions
129 # these tell the cache what to do with a cache miss or flush
130 sub read {
131 my($self, $key) = @_;
132 print "cache miss for $key, read() data\n";
133 rand() * $key;
134 }
135 sub write {
136 my($self, $key, $value) = @_;
137 print "flushing [$key, $value] from cache, write() data\n";
138 }
139 my $cache_size = $ARGV[0] || 2;
141 my $num_to_cache = $ARGV[1] || 4;
142 my $Debug = $ARGV[2] || 1;
143 tie %cache, 'My::Cache', $cache_size, {Debug => $Debug};
145 # load the cache with new data, each through its contents,
147 # and then reload in reverse order.
148 for(1..$num_to_cache) { print "read data $_: $cache{$_}\n" }
149 while(my($k, $v) = each %cache) { print "each data $k: $v\n"; }
150 for(my $i=$num_to_cache; $i>0; $i--) { print "read data $i: $cache{$i}\n"; }
151 # flush writes now, trivial use since will happen in DESTROY() anyway
153 tied(%cache)->flush();
154 # clear cache in 2 ways, write will flush out to disk
156 %cache = ();
157 undef %cache;
158
160 Many thanks to all those who helped me make this module a reality,
161 including:
162 :) Tom Hukins who provided me insight and motivation for
164 finishing this module.
165 :) Jamie McCarthy, for trying to make Tie::Cache be all
166 that it can be.
167 :) Rob Fugina who knows how to "TRULY CACHE".
168 :) Rob Bloodgood, for the TRUE CACHE flush() API
169
171 Please send any questions or comments to Joshua Chamas at
172 chamas@alumni.stanford.org
173
175 Copyright (c) 1999-2012 Joshua Chamas, Chamas Enterprises Inc.
176 Sponsored by development on NodeWorks http://nodeworks.com and Web
177 Test.org http://web-test.org
178 All rights reserved. This program is free software; you can
180 redistribute it and/or modify it under the same terms as Perl itself.
181perl v5.30.0 2019-07-26 Cache(3)