mingw/html/lib/Storable.html

   1 <?xml version="1.0" ?>
   2 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
   3 <html xmlns="http://www.w3.org/1999/xhtml">
   4 <head>
   5 <title>Storable - persistence for Perl data structures</title>
   6 <meta http-equiv="content-type" content="text/html; charset=utf-8" />
   7 <link rev="made" href="mailto:" />
   8 </head>
   9
  10 <body style="background-color: white">
  11 <table border="0" width="100%" cellspacing="0" cellpadding="3">
  12 <tr><td class="block" style="background-color: #cccccc" valign="middle">
  13 <big><strong><span class="block">&nbsp;Storable - persistence for Perl data structures</span></strong></big>
  14 </td></tr>
  15 </table>
  16
  17 <p><a name="__index__"></a></p>
  18 <!-- INDEX BEGIN -->
  19
  20 <ul>
  21
  22         <li><a href="#name">NAME</a></li>
  23         <li><a href="#synopsis">SYNOPSIS</a></li>
  24         <li><a href="#description">DESCRIPTION</a></li>
  25         <li><a href="#memory_store">MEMORY STORE</a></li>
  26         <li><a href="#advisory_locking">ADVISORY LOCKING</a></li>
  27         <li><a href="#speed">SPEED</a></li>
  28         <li><a href="#canonical_representation">CANONICAL REPRESENTATION</a></li>
  29         <li><a href="#code_references">CODE REFERENCES</a></li>
  30         <li><a href="#forward_compatibility">FORWARD COMPATIBILITY</a></li>
  31         <li><a href="#error_reporting">ERROR REPORTING</a></li>
  32         <li><a href="#wizards_only">WIZARDS ONLY</a></li>
  33         <ul>
  34
  35                 <li><a href="#hooks">Hooks</a></li>
  36                 <li><a href="#predicates">Predicates</a></li>
  37                 <li><a href="#recursion">Recursion</a></li>
  38                 <li><a href="#deep_cloning">Deep Cloning</a></li>
  39         </ul>
  40
  41         <li><a href="#storable_magic">Storable magic</a></li>
  42         <li><a href="#examples">EXAMPLES</a></li>
  43         <li><a href="#warning">WARNING</a></li>
  44         <li><a href="#bugs">BUGS</a></li>
  45         <ul>
  46
  47                 <li><a href="#64_bit_data_in_perl_5_6_0_and_5_6_1">64 bit data in perl 5.6.0 and 5.6.1</a></li>
  48         </ul>
  49
  50         <li><a href="#credits">CREDITS</a></li>
  51         <li><a href="#author">AUTHOR</a></li>
  52         <li><a href="#see_also">SEE ALSO</a></li>
  53 </ul>
  54 <!-- INDEX END -->
  55
  56 <hr />
  57 <p>
  58 </p>
  59 <h1><a name="name">NAME</a></h1>
  60 <p>Storable - persistence for Perl data structures</p>
  61 <p>
  62 </p>
  63 <hr />
  64 <h1><a name="synopsis">SYNOPSIS</a></h1>
  65 <pre>
  66  use Storable;
  67  store \%table, 'file';
  68  $hashref = retrieve('file');</pre>
  69 <pre>
  70  use Storable qw(nstore store_fd nstore_fd freeze thaw dclone);</pre>
  71 <pre>
  72  # Network order
  73  nstore \%table, 'file';
  74  $hashref = retrieve('file');   # There is NO nretrieve()</pre>
  75 <pre>
  76  # Storing to and retrieving from an already opened file
  77  store_fd \@array, \*STDOUT;
  78  nstore_fd \%table, \*STDOUT;
  79  $aryref = fd_retrieve(\*SOCKET);
  80  $hashref = fd_retrieve(\*SOCKET);</pre>
  81 <pre>
  82  # Serializing to memory
  83  $serialized = freeze \%table;
  84  %table_clone = %{ thaw($serialized) };</pre>
  85 <pre>
  86  # Deep (recursive) cloning
  87  $cloneref = dclone($ref);</pre>
  88 <pre>
  89  # Advisory locking
  90  use Storable qw(lock_store lock_nstore lock_retrieve)
  91  lock_store \%table, 'file';
  92  lock_nstore \%table, 'file';
  93  $hashref = lock_retrieve('file');</pre>
  94 <p>
  95 </p>
  96 <hr />
  97 <h1><a name="description">DESCRIPTION</a></h1>
  98 <p>The Storable package brings persistence to your Perl data structures
  99 containing SCALAR, ARRAY, HASH or REF objects, i.e. anything that can be
 100 conveniently stored to disk and retrieved at a later time.</p>
 101 <p>It can be used in the regular procedural way by calling <code>store</code> with
 102 a reference to the object to be stored, along with the file name where
 103 the image should be written.</p>
 104 <p>The routine returns <a href="file://C|\msysgit\mingw\html/pod/perlfunc.html#item_undef"><code>undef</code></a> for I/O problems or other internal error,
 105 a true value otherwise. Serious errors are propagated as a <code>die</code> exception.</p>
 106 <p>To retrieve data stored to disk, use <code>retrieve</code> with a file name.
 107 The objects stored into that file are recreated into memory for you,
 108 and a <em>reference</em> to the root object is returned. In case an I/O error
 109 occurs while reading, <a href="file://C|\msysgit\mingw\html/pod/perlfunc.html#item_undef"><code>undef</code></a> is returned instead. Other serious
 110 errors are propagated via <code>die</code>.</p>
 111 <p>Since storage is performed recursively, you might want to stuff references
 112 to objects that share a lot of common data into a single array or hash
 113 table, and then store that object. That way, when you retrieve back the
 114 whole thing, the objects will continue to share what they originally shared.</p>
 115 <p>At the cost of a slight header overhead, you may store to an already
 116 opened file descriptor using the <code>store_fd</code> routine, and retrieve
 117 from a file via <code>fd_retrieve</code>. Those names aren't imported by default,
 118 so you will have to do that explicitly if you need those routines.
 119 The file descriptor you supply must be already opened, for read
 120 if you're going to retrieve and for write if you wish to store.</p>
 121 <pre>
 122         store_fd(\%table, *STDOUT) || die &quot;can't store to stdout\n&quot;;
 123         $hashref = fd_retrieve(*STDIN);</pre>
 124 <p>You can also store data in network order to allow easy sharing across
 125 multiple platforms, or when storing on a socket known to be remotely
 126 connected. The routines to call have an initial <a href="file://C|\msysgit\mingw\html/pod/perlguts.html#item_n"><code>n</code></a> prefix for <em>network</em>,
 127 as in <code>nstore</code> and <code>nstore_fd</code>. At retrieval time, your data will be
 128 correctly restored so you don't have to know whether you're restoring
 129 from native or network ordered data.  Double values are stored stringified
 130 to ensure portability as well, at the slight risk of loosing some precision
 131 in the last decimals.</p>
 132 <p>When using <code>fd_retrieve</code>, objects are retrieved in sequence, one
 133 object (i.e. one recursive tree) per associated <code>store_fd</code>.</p>
 134 <p>If you're more from the object-oriented camp, you can inherit from
 135 Storable and directly store your objects by invoking <code>store</code> as
 136 a method. The fact that the root of the to-be-stored tree is a
 137 blessed reference (i.e. an object) is special-cased so that the
 138 retrieve does not provide a reference to that object but rather the
 139 blessed object reference itself. (Otherwise, you'd get a reference
 140 to that blessed object).</p>
 141 <p>
 142 </p>
 143 <hr />
 144 <h1><a name="memory_store">MEMORY STORE</a></h1>
 145 <p>The Storable engine can also store data into a Perl scalar instead, to
 146 later retrieve them. This is mainly used to freeze a complex structure in
 147 some safe compact memory place (where it can possibly be sent to another
 148 process via some IPC, since freezing the structure also serializes it in
 149 effect). Later on, and maybe somewhere else, you can thaw the Perl scalar
 150 out and recreate the original complex structure in memory.</p>
 151 <p>Surprisingly, the routines to be called are named <code>freeze</code> and <code>thaw</code>.
 152 If you wish to send out the frozen scalar to another machine, use
 153 <code>nfreeze</code> instead to get a portable image.</p>
 154 <p>Note that freezing an object structure and immediately thawing it
 155 actually achieves a deep cloning of that structure:</p>
 156 <pre>
 157     dclone(.) = thaw(freeze(.))</pre>
 158 <p>Storable provides you with a <code>dclone</code> interface which does not create
 159 that intermediary scalar but instead freezes the structure in some
 160 internal memory space and then immediately thaws it out.</p>
 161 <p>
 162 </p>
 163 <hr />
 164 <h1><a name="advisory_locking">ADVISORY LOCKING</a></h1>
 165 <p>The <code>lock_store</code> and <code>lock_nstore</code> routine are equivalent to
 166 <code>store</code> and <code>nstore</code>, except that they get an exclusive lock on
 167 the file before writing.  Likewise, <code>lock_retrieve</code> does the same
 168 as <code>retrieve</code>, but also gets a shared lock on the file before reading.</p>
 169 <p>As with any advisory locking scheme, the protection only works if you
 170 systematically use <code>lock_store</code> and <code>lock_retrieve</code>.  If one side of
 171 your application uses <code>store</code> whilst the other uses <code>lock_retrieve</code>,
 172 you will get no protection at all.</p>
 173 <p>The internal advisory locking is implemented using Perl's <code>flock()</code>
 174 routine.  If your system does not support any form of flock(), or if
 175 you share your files across NFS, you might wish to use other forms
 176 of locking by using modules such as LockFile::Simple which lock a
 177 file using a filesystem entry, instead of locking the file descriptor.</p>
 178 <p>
 179 </p>
 180 <hr />
 181 <h1><a name="speed">SPEED</a></h1>
 182 <p>The heart of Storable is written in C for decent speed. Extra low-level
 183 optimizations have been made when manipulating perl internals, to
 184 sacrifice encapsulation for the benefit of greater speed.</p>
 185 <p>
 186 </p>
 187 <hr />
 188 <h1><a name="canonical_representation">CANONICAL REPRESENTATION</a></h1>
 189 <p>Normally, Storable stores elements of hashes in the order they are
 190 stored internally by Perl, i.e. pseudo-randomly.  If you set
 191 <code>$Storable::canonical</code> to some <code>TRUE</code> value, Storable will store
 192 hashes with the elements sorted by their key.  This allows you to
 193 compare data structures by comparing their frozen representations (or
 194 even the compressed frozen representations), which can be useful for
 195 creating lookup tables for complicated queries.</p>
 196 <p>Canonical order does not imply network order; those are two orthogonal
 197 settings.</p>
 198 <p>
 199 </p>
 200 <hr />
 201 <h1><a name="code_references">CODE REFERENCES</a></h1>
 202 <p>Since Storable version 2.05, CODE references may be serialized with
 203 the help of <a href="file://C|\msysgit\mingw\html/lib/B/Deparse.html">the B::Deparse manpage</a>. To enable this feature, set
 204 <code>$Storable::Deparse</code> to a true value. To enable deserializazion,
 205 <code>$Storable::Eval</code> should be set to a true value. Be aware that
 206 deserialization is done through <a href="file://C|\msysgit\mingw\html/pod/perlfunc.html#item_eval"><code>eval</code></a>, which is dangerous if the
 207 Storable file contains malicious data. You can set <code>$Storable::Eval</code>
 208 to a subroutine reference which would be used instead of <a href="file://C|\msysgit\mingw\html/pod/perlfunc.html#item_eval"><code>eval</code></a>. See
 209 below for an example using a <a href="file://C|\msysgit\mingw\html/lib/Safe.html">the Safe manpage</a> compartment for deserialization
 210 of CODE references.</p>
 211 <p>If <code>$Storable::Deparse</code> and/or <code>$Storable::Eval</code> are set to false
 212 values, then the value of <code>$Storable::forgive_me</code> (see below) is
 213 respected while serializing and deserializing.</p>
 214 <p>
 215 </p>
 216 <hr />
 217 <h1><a name="forward_compatibility">FORWARD COMPATIBILITY</a></h1>
 218 <p>This release of Storable can be used on a newer version of Perl to
 219 serialize data which is not supported by earlier Perls.  By default,
 220 Storable will attempt to do the right thing, by <code>croak()</code>ing if it
 221 encounters data that it cannot deserialize.  However, the defaults
 222 can be changed as follows:</p>
 223 <dl>
 224 <dt><strong><a name="item_utf8_data">utf8 data</a></strong>
 225
 226 <dd>
 227 <p>Perl 5.6 added support for Unicode characters with code points &gt; 255,
 228 and Perl 5.8 has full support for Unicode characters in hash keys.
 229 Perl internally encodes strings with these characters using utf8, and
 230 Storable serializes them as utf8.  By default, if an older version of
 231 Perl encounters a utf8 value it cannot represent, it will <code>croak()</code>.
 232 To change this behaviour so that Storable deserializes utf8 encoded
 233 values as the string of bytes (effectively dropping the <em>is_utf8</em> flag)
 234 set <code>$Storable::drop_utf8</code> to some <code>TRUE</code> value.  This is a form of
 235 data loss, because with <code>$drop_utf8</code> true, it becomes impossible to tell
 236 whether the original data was the Unicode string, or a series of bytes
 237 that happen to be valid utf8.</p>
 238 </dd>
 239 </li>
 240 <dt><strong><a name="item_restricted_hashes">restricted hashes</a></strong>
 241
 242 <dd>
 243 <p>Perl 5.8 adds support for restricted hashes, which have keys
 244 restricted to a given set, and can have values locked to be read only.
 245 By default, when Storable encounters a restricted hash on a perl
 246 that doesn't support them, it will deserialize it as a normal hash,
 247 silently discarding any placeholder keys and leaving the keys and
 248 all values unlocked.  To make Storable <code>croak()</code> instead, set
 249 <code>$Storable::downgrade_restricted</code> to a <code>FALSE</code> value.  To restore
 250 the default set it back to some <code>TRUE</code> value.</p>
 251 </dd>
 252 </li>
 253 <dt><strong><a name="item_files_from_future_versions_of_storable">files from future versions of Storable</a></strong>
 254
 255 <dd>
 256 <p>Earlier versions of Storable would immediately croak if they encountered
 257 a file with a higher internal version number than the reading Storable
 258 knew about.  Internal version numbers are increased each time new data
 259 types (such as restricted hashes) are added to the vocabulary of the file
 260 format.  This meant that a newer Storable module had no way of writing a
 261 file readable by an older Storable, even if the writer didn't store newer
 262 data types.</p>
 263 </dd>
 264 <dd>
 265 <p>This version of Storable will defer croaking until it encounters a data
 266 type in the file that it does not recognize.  This means that it will
 267 continue to read files generated by newer Storable modules which are careful
 268 in what they write out, making it easier to upgrade Storable modules in a
 269 mixed environment.</p>
 270 </dd>
 271 <dd>
 272 <p>The old behaviour of immediate croaking can be re-instated by setting
 273 <code>$Storable::accept_future_minor</code> to some <code>FALSE</code> value.</p>
 274 </dd>
 275 </li>
 276 </dl>
 277 <p>All these variables have no effect on a newer Perl which supports the
 278 relevant feature.</p>
 279 <p>
 280 </p>
 281 <hr />
 282 <h1><a name="error_reporting">ERROR REPORTING</a></h1>
 283 <p>Storable uses the ``exception'' paradigm, in that it does not try to workaround
 284 failures: if something bad happens, an exception is generated from the
 285 caller's perspective (see <a href="file://C|\msysgit\mingw\html/lib/Carp.html">the Carp manpage</a> and <code>croak()</code>).  Use eval {} to trap
 286 those exceptions.</p>
 287 <p>When Storable croaks, it tries to report the error via the <code>logcroak()</code>
 288 routine from the <code>Log::Agent</code> package, if it is available.</p>
 289 <p>Normal errors are reported by having <code>store()</code> or <code>retrieve()</code> return <a href="file://C|\msysgit\mingw\html/pod/perlfunc.html#item_undef"><code>undef</code></a>.
 290 Such errors are usually I/O errors (or truncated stream errors at retrieval).</p>
 291 <p>
 292 </p>
 293 <hr />
 294 <h1><a name="wizards_only">WIZARDS ONLY</a></h1>
 295 <p>
 296 </p>
 297 <h2><a name="hooks">Hooks</a></h2>
 298 <p>Any class may define hooks that will be called during the serialization
 299 and deserialization process on objects that are instances of that class.
 300 Those hooks can redefine the way serialization is performed (and therefore,
 301 how the symmetrical deserialization should be conducted).</p>
 302 <p>Since we said earlier:</p>
 303 <pre>
 304     dclone(.) = thaw(freeze(.))</pre>
 305 <p>everything we say about hooks should also hold for deep cloning. However,
 306 hooks get to know whether the operation is a mere serialization, or a cloning.</p>
 307 <p>Therefore, when serializing hooks are involved,</p>
 308 <pre>
 309     dclone(.) &lt;&gt; thaw(freeze(.))</pre>
 310 <p>Well, you could keep them in sync, but there's no guarantee it will always
 311 hold on classes somebody else wrote.  Besides, there is little to gain in
 312 doing so: a serializing hook could keep only one attribute of an object,
 313 which is probably not what should happen during a deep cloning of that
 314 same object.</p>
 315 <p>Here is the hooking interface:</p>
 316 <dl>
 317 <dt><strong><a name="item_storable_freeze_obj_2c_cloning"><code>STORABLE_freeze</code> <em>obj</em>, <em>cloning</em></a></strong>
 318
 319 <dd>
 320 <p>The serializing hook, called on the object during serialization.  It can be
 321 inherited, or defined in the class itself, like any other method.</p>
 322 </dd>
 323 <dd>
 324 <p>Arguments: <em>obj</em> is the object to serialize, <em>cloning</em> is a flag indicating
 325 whether we're in a <code>dclone()</code> or a regular serialization via <code>store()</code> or freeze().</p>
 326 </dd>
 327 <dd>
 328 <p>Returned value: A LIST <code>($serialized, $ref1, $ref2, ...)</code> where $serialized
 329 is the serialized form to be used, and the optional $ref1, $ref2, etc... are
 330 extra references that you wish to let the Storable engine serialize.</p>
 331 </dd>
 332 <dd>
 333 <p>At deserialization time, you will be given back the same LIST, but all the
 334 extra references will be pointing into the deserialized structure.</p>
 335 </dd>
 336 <dd>
 337 <p>The <strong>first time</strong> the hook is hit in a serialization flow, you may have it
 338 return an empty list.  That will signal the Storable engine to further
 339 discard that hook for this class and to therefore revert to the default
 340 serialization of the underlying Perl data.  The hook will again be normally
 341 processed in the next serialization.</p>
 342 </dd>
 343 <dd>
 344 <p>Unless you know better, serializing hook should always say:</p>
 345 </dd>
 346 <dd>
 347 <pre>
 348     sub STORABLE_freeze {
 349         my ($self, $cloning) = @_;
 350         return if $cloning;         # Regular default serialization
 351         ....
 352     }</pre>
 353 </dd>
 354 <dd>
 355 <p>in order to keep reasonable <code>dclone()</code> semantics.</p>
 356 </dd>
 357 </li>
 358 <dt><strong><a name="item_storable_thaw_obj_2c_cloning_2c_serialized_2c__2e_"><code>STORABLE_thaw</code> <em>obj</em>, <em>cloning</em>, <em>serialized</em>, ...</a></strong>
 359
 360 <dd>
 361 <p>The deserializing hook called on the object during deserialization.
 362 But wait: if we're deserializing, there's no object yet... right?</p>
 363 </dd>
 364 <dd>
 365 <p>Wrong: the Storable engine creates an empty one for you.  If you know Eiffel,
 366 you can view <code>STORABLE_thaw</code> as an alternate creation routine.</p>
 367 </dd>
 368 <dd>
 369 <p>This means the hook can be inherited like any other method, and that
 370 <em>obj</em> is your blessed reference for this particular instance.</p>
 371 </dd>
 372 <dd>
 373 <p>The other arguments should look familiar if you know <code>STORABLE_freeze</code>:
 374 <em>cloning</em> is true when we're part of a deep clone operation, <em>serialized</em>
 375 is the serialized string you returned to the engine in <code>STORABLE_freeze</code>,
 376 and there may be an optional list of references, in the same order you gave
 377 them at serialization time, pointing to the deserialized objects (which
 378 have been processed courtesy of the Storable engine).</p>
 379 </dd>
 380 <dd>
 381 <p>When the Storable engine does not find any <code>STORABLE_thaw</code> hook routine,
 382 it tries to load the class by requiring the package dynamically (using
 383 the blessed package name), and then re-attempts the lookup.  If at that
 384 time the hook cannot be located, the engine croaks.  Note that this mechanism
 385 will fail if you define several classes in the same file, but <a href="file://C|\msysgit\mingw\html/pod/perlmod.html">the perlmod manpage</a>
 386 warned you.</p>
 387 </dd>
 388 <dd>
 389 <p>It is up to you to use this information to populate <em>obj</em> the way you want.</p>
 390 </dd>
 391 <dd>
 392 <p>Returned value: none.</p>
 393 </dd>
 394 </li>
 395 <dt><strong><a name="item_storable_attach_class_2c_cloning_2c_serialized"><code>STORABLE_attach</code> <em>class</em>, <em>cloning</em>, <em>serialized</em></a></strong>
 396
 397 <dd>
 398 <p>While <code>STORABLE_freeze</code> and <code>STORABLE_thaw</code> are useful for classes where
 399 each instance is independant, this mechanism has difficulty (or is
 400 incompatible) with objects that exist as common process-level or
 401 system-level resources, such as singleton objects, database pools, caches
 402 or memoized objects.</p>
 403 </dd>
 404 <dd>
 405 <p>The alternative <code>STORABLE_attach</code> method provides a solution for these
 406 shared objects. Instead of <code>STORABLE_freeze</code> --&GT; <code>STORABLE_thaw</code>,
 407 you implement <code>STORABLE_freeze</code> --&GT; <code>STORABLE_attach</code> instead.</p>
 408 </dd>
 409 <dd>
 410 <p>Arguments: <em>class</em> is the class we are attaching to, <em>cloning</em> is a flag
 411 indicating whether we're in a <code>dclone()</code> or a regular de-serialization via
 412 thaw(), and <em>serialized</em> is the stored string for the resource object.</p>
 413 </dd>
 414 <dd>
 415 <p>Because these resource objects are considered to be owned by the entire
 416 process/system, and not the ``property'' of whatever is being serialized,
 417 no references underneath the object should be included in the serialized
 418 string. Thus, in any class that implements <code>STORABLE_attach</code>, the
 419 <code>STORABLE_freeze</code> method cannot return any references, and <code>Storable</code>
 420 will throw an error if <code>STORABLE_freeze</code> tries to return references.</p>
 421 </dd>
 422 <dd>
 423 <p>All information required to ``attach'' back to the shared resource object
 424 <strong>must</strong> be contained <strong>only</strong> in the <code>STORABLE_freeze</code> return string.
 425 Otherwise, <code>STORABLE_freeze</code> behaves as normal for <code>STORABLE_attach</code>
 426 classes.</p>
 427 </dd>
 428 <dd>
 429 <p>Because <code>STORABLE_attach</code> is passed the class (rather than an object),
 430 it also returns the object directly, rather than modifying the passed
 431 object.</p>
 432 </dd>
 433 <dd>
 434 <p>Returned value: object of type <code>class</code></p>
 435 </dd>
 436 </li>
 437 </dl>
 438 <p>
 439 </p>
 440 <h2><a name="predicates">Predicates</a></h2>
 441 <p>Predicates are not exportable.  They must be called by explicitly prefixing
 442 them with the Storable package name.</p>
 443 <dl>
 444 <dt><strong><a name="item_storable_3a_3alast_op_in_netorder"><code>Storable::last_op_in_netorder</code></a></strong>
 445
 446 <dd>
 447 <p>The <code>Storable::last_op_in_netorder()</code> predicate will tell you whether
 448 network order was used in the last store or retrieve operation.  If you
 449 don't know how to use this, just forget about it.</p>
 450 </dd>
 451 </li>
 452 <dt><strong><a name="item_storable_3a_3ais_storing"><code>Storable::is_storing</code></a></strong>
 453
 454 <dd>
 455 <p>Returns true if within a store operation (via STORABLE_freeze hook).</p>
 456 </dd>
 457 </li>
 458 <dt><strong><a name="item_storable_3a_3ais_retrieving"><code>Storable::is_retrieving</code></a></strong>
 459
 460 <dd>
 461 <p>Returns true if within a retrieve operation (via STORABLE_thaw hook).</p>
 462 </dd>
 463 </li>
 464 </dl>
 465 <p>
 466 </p>
 467 <h2><a name="recursion">Recursion</a></h2>
 468 <p>With hooks comes the ability to recurse back to the Storable engine.
 469 Indeed, hooks are regular Perl code, and Storable is convenient when
 470 it comes to serializing and deserializing things, so why not use it
 471 to handle the serialization string?</p>
 472 <p>There are a few things you need to know, however:</p>
 473 <ul>
 474 <li>
 475 <p>You can create endless loops if the things you serialize via <code>freeze()</code>
 476 (for instance) point back to the object we're trying to serialize in
 477 the hook.</p>
 478 </li>
 479 <li>
 480 <p>Shared references among objects will not stay shared: if we're serializing
 481 the list of object [A, C] where both object A and C refer to the SAME object
 482 B, and if there is a serializing hook in A that says freeze(B), then when
 483 deserializing, we'll get [A', C'] where A' refers to B', but C' refers to D,
 484 a deep clone of B'.  The topology was not preserved.</p>
 485 </li>
 486 </ul>
 487 <p>That's why <code>STORABLE_freeze</code> lets you provide a list of references
 488 to serialize.  The engine guarantees that those will be serialized in the
 489 same context as the other objects, and therefore that shared objects will
 490 stay shared.</p>
 491 <p>In the above [A, C] example, the <code>STORABLE_freeze</code> hook could return:</p>
 492 <pre>
 493         (&quot;something&quot;, $self-&gt;{B})</pre>
 494 <p>and the B part would be serialized by the engine.  In <code>STORABLE_thaw</code>, you
 495 would get back the reference to the B' object, deserialized for you.</p>
 496 <p>Therefore, recursion should normally be avoided, but is nonetheless supported.</p>
 497 <p>
 498 </p>
 499 <h2><a name="deep_cloning">Deep Cloning</a></h2>
 500 <p>There is a Clone module available on CPAN which implements deep cloning
 501 natively, i.e. without freezing to memory and thawing the result.  It is
 502 aimed to replace Storable's <code>dclone()</code> some day.  However, it does not currently
 503 support Storable hooks to redefine the way deep cloning is performed.</p>
 504 <p>
 505 </p>
 506 <hr />
 507 <h1><a name="storable_magic">Storable magic</a></h1>
 508 <p>Yes, there's a lot of that :-) But more precisely, in UNIX systems
 509 there's a utility called <code>file</code>, which recognizes data files based on
 510 their contents (usually their first few bytes).  For this to work,
 511 a certain file called <em>magic</em> needs to taught about the <em>signature</em>
 512 of the data.  Where that configuration file lives depends on the UNIX
 513 flavour; often it's something like <em>/usr/share/misc/magic</em> or
 514 <em>/etc/magic</em>.  Your system administrator needs to do the updating of
 515 the <em>magic</em> file.  The necessary signature information is output to
 516 STDOUT by invoking Storable::show_file_magic().  Note that the GNU
 517 implementation of the <code>file</code> utility, version 3.38 or later,
 518 is expected to contain support for recognising Storable files
 519 out-of-the-box, in addition to other kinds of Perl files.</p>
 520 <p>
 521 </p>
 522 <hr />
 523 <h1><a name="examples">EXAMPLES</a></h1>
 524 <p>Here are some code samples showing a possible usage of Storable:</p>
 525 <pre>
 526         use Storable qw(store retrieve freeze thaw dclone);</pre>
 527 <pre>
 528         %color = ('Blue' =&gt; 0.1, 'Red' =&gt; 0.8, 'Black' =&gt; 0, 'White' =&gt; 1);</pre>
 529 <pre>
 530         store(\%color, 'mycolors') or die &quot;Can't store %a in mycolors!\n&quot;;</pre>
 531 <pre>
 532         $colref = retrieve('mycolors');
 533         die &quot;Unable to retrieve from mycolors!\n&quot; unless defined $colref;
 534         printf &quot;Blue is still %lf\n&quot;, $colref-&gt;{'Blue'};</pre>
 535 <pre>
 536         $colref2 = dclone(\%color);</pre>
 537 <pre>
 538         $str = freeze(\%color);
 539         printf &quot;Serialization of %%color is %d bytes long.\n&quot;, length($str);
 540         $colref3 = thaw($str);</pre>
 541 <p>which prints (on my machine):</p>
 542 <pre>
 543         Blue is still 0.100000
 544         Serialization of %color is 102 bytes long.</pre>
 545 <p>Serialization of CODE references and deserialization in a safe
 546 compartment:</p>
 547 <pre>
 548         use Storable qw(freeze thaw);
 549         use Safe;
 550         use strict;
 551         my $safe = new Safe;
 552         # because of opcodes used in &quot;use strict&quot;:
 553         $safe-&gt;permit(qw(:default require));
 554         local $Storable::Deparse = 1;
 555         local $Storable::Eval = sub { $safe-&gt;reval($_[0]) };
 556         my $serialized = freeze(sub { 42 });
 557         my $code = thaw($serialized);
 558         $code-&gt;() == 42;</pre>
 559 <p>
 560 </p>
 561 <hr />
 562 <h1><a name="warning">WARNING</a></h1>
 563 <p>If you're using references as keys within your hash tables, you're bound
 564 to be disappointed when retrieving your data. Indeed, Perl stringifies
 565 references used as hash table keys. If you later wish to access the
 566 items via another reference stringification (i.e. using the same
 567 reference that was used for the key originally to record the value into
 568 the hash table), it will work because both references stringify to the
 569 same string.</p>
 570 <p>It won't work across a sequence of <code>store</code> and <code>retrieve</code> operations,
 571 however, because the addresses in the retrieved objects, which are
 572 part of the stringified references, will probably differ from the
 573 original addresses. The topology of your structure is preserved,
 574 but not hidden semantics like those.</p>
 575 <p>On platforms where it matters, be sure to call <a href="file://C|\msysgit\mingw\html/pod/perlfunc.html#item_binmode"><code>binmode()</code></a> on the
 576 descriptors that you pass to Storable functions.</p>
 577 <p>Storing data canonically that contains large hashes can be
 578 significantly slower than storing the same data normally, as
 579 temporary arrays to hold the keys for each hash have to be allocated,
 580 populated, sorted and freed.  Some tests have shown a halving of the
 581 speed of storing -- the exact penalty will depend on the complexity of
 582 your data.  There is no slowdown on retrieval.</p>
 583 <p>
 584 </p>
 585 <hr />
 586 <h1><a name="bugs">BUGS</a></h1>
 587 <p>You can't store GLOB, FORMLINE, etc.... If you can define semantics
 588 for those operations, feel free to enhance Storable so that it can
 589 deal with them.</p>
 590 <p>The store functions will <code>croak</code> if they run into such references
 591 unless you set <code>$Storable::forgive_me</code> to some <code>TRUE</code> value. In that
 592 case, the fatal message is turned in a warning and some
 593 meaningless string is stored instead.</p>
 594 <p>Setting <code>$Storable::canonical</code> may not yield frozen strings that
 595 compare equal due to possible stringification of numbers. When the
 596 string version of a scalar exists, it is the form stored; therefore,
 597 if you happen to use your numbers as strings between two freezing
 598 operations on the same data structures, you will get different
 599 results.</p>
 600 <p>When storing doubles in network order, their value is stored as text.
 601 However, you should also not expect non-numeric floating-point values
 602 such as infinity and ``not a number'' to pass successfully through a
 603 <code>nstore()/retrieve()</code> pair.</p>
 604 <p>As Storable neither knows nor cares about character sets (although it
 605 does know that characters may be more than eight bits wide), any difference
 606 in the interpretation of character codes between a host and a target
 607 system is your problem.  In particular, if host and target use different
 608 code points to represent the characters used in the text representation
 609 of floating-point numbers, you will not be able be able to exchange
 610 floating-point data, even with nstore().</p>
 611 <p><code>Storable::drop_utf8</code> is a blunt tool.  There is no facility either to
 612 return <strong>all</strong> strings as utf8 sequences, or to attempt to convert utf8
 613 data back to 8 bit and <code>croak()</code> if the conversion fails.</p>
 614 <p>Prior to Storable 2.01, no distinction was made between signed and
 615 unsigned integers on storing.  By default Storable prefers to store a
 616 scalars string representation (if it has one) so this would only cause
 617 problems when storing large unsigned integers that had never been coverted
 618 to string or floating point.  In other words values that had been generated
 619 by integer operations such as logic ops and then not used in any string or
 620 arithmetic context before storing.</p>
 621 <p>
 622 </p>
 623 <h2><a name="64_bit_data_in_perl_5_6_0_and_5_6_1">64 bit data in perl 5.6.0 and 5.6.1</a></h2>
 624 <p>This section only applies to you if you have existing data written out
 625 by Storable 2.02 or earlier on perl 5.6.0 or 5.6.1 on Unix or Linux which
 626 has been configured with 64 bit integer support (not the default)
 627 If you got a precompiled perl, rather than running Configure to build
 628 your own perl from source, then it almost certainly does not affect you,
 629 and you can stop reading now (unless you're curious). If you're using perl
 630 on Windows it does not affect you.</p>
 631 <p>Storable writes a file header which contains the sizes of various C
 632 language types for the C compiler that built Storable (when not writing in
 633 network order), and will refuse to load files written by a Storable not
 634 on the same (or compatible) architecture.  This check and a check on
 635 machine byteorder is needed because the size of various fields in the file
 636 are given by the sizes of the C language types, and so files written on
 637 different architectures are incompatible.  This is done for increased speed.
 638 (When writing in network order, all fields are written out as standard
 639 lengths, which allows full interworking, but takes longer to read and write)</p>
 640 <p>Perl 5.6.x introduced the ability to optional configure the perl interpreter
 641 to use C's <code>long long</code> type to allow scalars to store 64 bit integers on 32
 642 bit systems.  However, due to the way the Perl configuration system
 643 generated the C configuration files on non-Windows platforms, and the way
 644 Storable generates its header, nothing in the Storable file header reflected
 645 whether the perl writing was using 32 or 64 bit integers, despite the fact
 646 that Storable was storing some data differently in the file.  Hence Storable
 647 running on perl with 64 bit integers will read the header from a file
 648 written by a 32 bit perl, not realise that the data is actually in a subtly
 649 incompatible format, and then go horribly wrong (possibly crashing) if it
 650 encountered a stored integer.  This is a design failure.</p>
 651 <p>Storable has now been changed to write out and read in a file header with
 652 information about the size of integers.  It's impossible to detect whether
 653 an old file being read in was written with 32 or 64 bit integers (they have
 654 the same header) so it's impossible to automatically switch to a correct
 655 backwards compatibility mode.  Hence this Storable defaults to the new,
 656 correct behaviour.</p>
 657 <p>What this means is that if you have data written by Storable 1.x running
 658 on perl 5.6.0 or 5.6.1 configured with 64 bit integers on Unix or Linux
 659 then by default this Storable will refuse to read it, giving the error
 660 <em>Byte order is not compatible</em>.  If you have such data then you you
 661 should set <code>$Storable::interwork_56_64bit</code> to a true value to make this
 662 Storable read and write files with the old header.  You should also
 663 migrate your data, or any older perl you are communicating with, to this
 664 current version of Storable.</p>
 665 <p>If you don't have data written with specific configuration of perl described
 666 above, then you do not and should not do anything.  Don't set the flag -
 667 not only will Storable on an identically configured perl refuse to load them,
 668 but Storable a differently configured perl will load them believing them
 669 to be correct for it, and then may well fail or crash part way through
 670 reading them.</p>
 671 <p>
 672 </p>
 673 <hr />
 674 <h1><a name="credits">CREDITS</a></h1>
 675 <p>Thank you to (in chronological order):</p>
 676 <pre>
 677         Jarkko Hietaniemi &lt;jhi@iki.fi&gt;
 678         Ulrich Pfeifer &lt;pfeifer@charly.informatik.uni-dortmund.de&gt;
 679         Benjamin A. Holzman &lt;bah@ecnvantage.com&gt;
 680         Andrew Ford &lt;A.Ford@ford-mason.co.uk&gt;
 681         Gisle Aas &lt;gisle@aas.no&gt;
 682         Jeff Gresham &lt;gresham_jeffrey@jpmorgan.com&gt;
 683         Murray Nesbitt &lt;murray@activestate.com&gt;
 684         Marc Lehmann &lt;pcg@opengroup.org&gt;
 685         Justin Banks &lt;justinb@wamnet.com&gt;
 686         Jarkko Hietaniemi &lt;jhi@iki.fi&gt; (AGAIN, as perl 5.7.0 Pumpkin!)
 687         Salvador Ortiz Garcia &lt;sog@msg.com.mx&gt;
 688         Dominic Dunlop &lt;domo@computer.org&gt;
 689         Erik Haugan &lt;erik@solbors.no&gt;</pre>
 690 <p>for their bug reports, suggestions and contributions.</p>
 691 <p>Benjamin Holzman contributed the tied variable support, Andrew Ford
 692 contributed the canonical order for hashes, and Gisle Aas fixed
 693 a few misunderstandings of mine regarding the perl internals,
 694 and optimized the emission of ``tags'' in the output streams by
 695 simply counting the objects instead of tagging them (leading to
 696 a binary incompatibility for the Storable image starting at version
 697 0.6--older images are, of course, still properly understood).
 698 Murray Nesbitt made Storable thread-safe.  Marc Lehmann added overloading
 699 and references to tied items support.</p>
 700 <p>
 701 </p>
 702 <hr />
 703 <h1><a name="author">AUTHOR</a></h1>
 704 <p>Storable was written by Raphael Manfredi <em>&lt;<a href="mailto:Raphael_Manfredi@pobox.com">Raphael_Manfredi@pobox.com</a>&gt;</em>
 705 Maintenance is now done by the perl5-porters <em>&lt;<a href="mailto:perl5-porters@perl.org">perl5-porters@perl.org</a>&gt;</em></p>
 706 <p>Please e-mail us with problems, bug fixes, comments and complaints,
 707 although if you have complements you should send them to Raphael.
 708 Please don't e-mail Raphael with problems, as he no longer works on
 709 Storable, and your message will be delayed while he forwards it to us.</p>
 710 <p>
 711 </p>
 712 <hr />
 713 <h1><a name="see_also">SEE ALSO</a></h1>
 714 <p><em>Clone</em>.</p>
 715 <table border="0" width="100%" cellspacing="0" cellpadding="3">
 716 <tr><td class="block" style="background-color: #cccccc" valign="middle">
 717 <big><strong><span class="block">&nbsp;Storable - persistence for Perl data structures</span></strong></big>
 718 </td></tr>
 719 </table>
 720
 721 </body>
 722
 723 </html>