123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276 |
- <!doctype HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
- <html>
- <!--
- (C) Copyright 2002-4 Robert Ramey - http://www.rrsd.com .
- Use, modification and distribution is subject to the Boost Software
- License, Version 1.0. (See accompanying file LICENSE_1_0.txt or copy at
- http://www.boost.org/LICENSE_1_0.txt)
- -->
- <head>
- <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
- <link rel="stylesheet" type="text/css" href="../../../boost.css">
- <link rel="stylesheet" type="text/css" href="style.css">
- <title>Serialization - Archive Exceptions</title>
- </head>
- <body link="#0000ff" vlink="#800080">
- <table border="0" cellpadding="7" cellspacing="0" width="100%" summary="header">
- <tr>
- <td valign="top" width="300">
- <h3><a href="../../../index.htm"><img height="86" width="277" alt="C++ Boost" src="../../../boost.png" border="0"></a></h3>
- </td>
- <td valign="top">
- <h1 align="center">Serialization</h1>
- <h2 align="center">Archive Exceptions</h2>
- </td>
- </tr>
- </table>
- <hr>
- <dl class="page-index">
- <dt><a href="#unregistered_class"><code style="white-space: normal">unregistered_class</code></a>
- <dt><a href="#invalid_signature"><code style="white-space: normal">invalid_signature</code></a>
- <dt><a href="#unsupported_version"><code style="white-space: normal">unsupported_version</code></a>
- <dt><a href="#unsupported_class_version"><code style="white-space: normal">unsupported_class_version</code></a>
- <dt><a href="#pointer_conflict"><code style="white-space: normal">pointer_conflict</code></a>
- <dt><a href="#incompatible_native_format"><code style="white-space: normal">incompatible_native_format</code></a>
- <dt><a href="#array_size_too_short"><code style="white-space: normal">array_size_too_short</code></a>
- <dt><a href="#input_stream_error"><code style="white-space: normal">input_stream_error</code></a>
- <dt><a href="#output_stream_error"><code style="white-space: normal">output_stream_error</code></a>
- <dt><a href="#invalid_class_name"><code style="white-space: normal">invalid_class_name</code></a>
- <dt><a href="#unregistered_class"><code style="white-space: normal">unregistered_class</code></a>
- <dt><a href="#multiple_code_instantiation"><code style="white-space: normal">multiple_code_instantiation</code></a>
- <dt><a href="#xml_archive_parsing_error"><code style="white-space: normal">xml_archive_parsing_error</code></a>
- <dt><a href="#xml_archive_tag_mismatch"><code style="white-space: normal">xml_archive_tag_mismatch</code></a>
- <dt><a href="#xml_archive_tag_name_error"><code style="white-space: normal">xml_archive_tag_name_error</code></a>
- </dl>
- Archive operators can throw a <code style="white-space: normal">boost::archive_exception</code>
- object which can be caught by an application program. These exceptions are defined
- in the files <a target="archive_exception_hpp" href="../../../boost/archive/archive_exception.hpp">
- archive_exception.hpp</a>
- and <a target="basic_xml_archive_hpp" href="../../../boost/archive/basic_xml_archive.hpp">
- basic_xml_archive.hpp</a>.
- <pre><code>
- namespace boost {
- namespace archive {
- class archive_exception : public std::exception
- {
- public:
- typedef enum {
- unregistered_class, // attempt to serialize a pointer of
- // an unregistered class
- invalid_signature, // first line of archive does not contain
- // expected string
- unsupported_version, // archive created with library version subsequent
- // to this one
- pointer_conflict // an attempt has been made to directly serialize
- // an object after having already serialized the same
- // object through a pointer. Were this permitted,
- // the archive load would result in the creation
- // of an extraneous object.
- incompatible_native_format, // attempt to read native binary format
- // on incompatible platform
- array_size_too_short, // array being loaded doesn't fit in array allocated
- input_stream_error // error on stream input
- invalid_class_name, // class name greater than the maximum permitted.
- // most likely a corrupted archive or an attempt
- // to insert virus via buffer overrun method.
- unregistered_cast, // base - derived relationship not registered with
- // void_cast_register
- unsupported_class_version, // type saved with a version # greater than the
- // one used by the program. This indicates that the proggram
- // needs to be rebuilt.
- multiple_code_instantiation, // code for implementing serialization for some
- // type has been instantiated in more than one module.
- output_stream_error // error on stream output
- } exception_code;
- exception_code code;
- archive_exception(exception_code c) : code(c) {}
- virtual const char *what( ) const throw();
- };
- class xml_archive_exception : public virtual archive_exception
- {
- public:
- typedef enum {
- xml_archive_parsing_error, // archive doesn't contain expected data
- xml_archive_tag_mismatch, // start/end tag in archive doesn't match program
- xml_archive_tag_name_error // tag name contains invalid characters
- } exception_code;
- xml_archive_exception(exception_code c){}
- virtual const char *what( ) const throw();
- };
- } // archive
- } // boost
- </code></pre>
- <p>
- <h3><a name="unregistered_class"><code style="white-space: normal">unregistered_class</code></a></h3>
- An attempt has been made to serialize a polymorphic class through a pointer
- without either registering it or associating it with an export key. This can also occur
- when using a new archive whose class name has not been added to the system with the
- <code style="white-space: normal">BOOST_ARCHIVE_CUSTOM_ARCHIVE_TYPES</code> macro.
- <h3><a name="invalid_signature"><code style="white-space: normal">invalid_signature</code></a></h3>
- Archives are initiated with a known string. If this string is not found when
- the archive is opened, It is presumed that this file is not a valid archive and this
- exception is thrown.
- <h3><a name="unsupported_version"><code style="white-space: normal">unsupported_version</code></a></h3>
- This system records the current library version number to all archives created. Note that this is in
- no way related to version number of classes used by application programs. This refers
- to the version of the serialization system used to create the archive. Future versions
- of this serialization system will be able to identify archives created under a previous
- (i.e. this) system and alter the loading procedure accordingly. Hence, future enhancements
- to this serialization system should not obsolete any existing archive files. It is only
- necessary to increment this version number when the newer system creates archives
- incompatible in format with the current one.
- <p>Should it ever occur that an older program attempts to read newer archives whose
- format has changed, this exception is thrown.
- <h3><a name="unsupported_class_version"><code style="white-space: normal">unsupported_class_version</code></a></h3>
- An attempt has been made to load a class whose version has been incremented since the
- program was written. Suppose that a class has been assigned version number 3 and the program
- has been built and sent to third parties. Now suppose that the definition of that class
- has been altered, the version number has been incremented to 4 and new archives have been
- built. If one attempts to load these new archives with the original program, this
- exception will be thrown.
- <h3><a name="pointer_conflict"><code style="white-space: normal">pointer_conflict</code></a></h3>
- To understand what this exception means consider the following scenario
- <pre><code>
- template<class Archive>
- void T::save(Archive &ar) const
- {
- const A * aptr = &a;
- ar << aptr; // save an instance of object of class A through a pointer
- ...
- ar << a; // save an instance of an object of class A
- assert(aptr == &a); // this must be true
- }
- template<class Archive>
- void T::load(Archive &ar)
- {
- A * aptr;
- ar >> aptr; // create and initialize a new instance of class A
- ...
- ar >> a; // restore state of on object of class A
- assert(aptr == &a); // this won't be true
- }
- </pre></code>
- An object is saved first through a pointer then directly. Upon loading back
- in the same sequence, we first create an new object and load in its data. Then
- we load the data into another existing object. Where we started with one
- object during save, we have two objects after restore. In a more realistic
- situation, it could be very difficult to find this error. Fortunately,
- these situations can be detected when the archive is created. When
- this occurs, this exception is thrown.
- <h3><a name = "incompatible_native_format"><code style="white-space: normal">incompatible_native_format</code></a></h3>
- The library currently supports char text, wide char text and native binary
- archive files. At the beginning of every archive, a signature is written indicating
- the type of archive. This exception is thrown when an attempt is made to read
- an archive written in a different format.
- <h3><a name="array_size_too_short"><code style="white-space: normal">array_size_too_short</code></a></h3>
- An attempt has been made to read an array that is larger than the array size.
- This should only occur when the size of an array in code is reduced after an
- archive has already been created.
- <h3>
- <a name="input_stream_error"><code style="white-space: normal">input_stream_error</code></a>
- <br>
- <a name="output_stream_error"><code style="white-space: normal">output_stream_error</code></a>
- </h3>
- An error has occured during stream input or ouput. Aside from the common
- situations such as a corrupted or truncated input file, there are
- several less obvious ones that sometimes occur.
- <p>
- This includes
- an attempt to read past the end of the file. Text files need a terminating
- new line character at the end of the file which is appended when the
- archive destructor is invoked. Be sure that an output archive on a stream
- is destroyed before opening an input archive on that same stream. That is,
- rather than using something like:
- <pre><code>
- std::stringstream ss;
- std::vector<V> v;
- boost::archive::text_oarchive oa(ss);
- oa << v;
- boost::archive::text_iarchive ia(ss);
- ia >> v;
- </code></pre>
- use
- <pre><code>
- std::stringstream ss;
- std::vector<V> v;
- {
- boost::archive::text_oarchive oa(ss);
- oa << v;
- }
- {
- boost::archive::text_iarchive ia(ss);
- ia >> v;
- }
- </code></pre>
- <p>
- Another one is the passing of uninitialized data. In general, the behavior
- of the serialization library when passed uninitialized data is undefined.
- If it can be detected, it will invoke an assertion in debug builds.
- Otherwise, depending on the type of archive, it may pass through without
- incident or it may result in an archive with unexpected data in it.
- This, in turn, can result in the throwing of this exception.
- <h3><a name="invalid_class_name"><code style="white-space: normal">invalid_class_name</code></a></h3>
- Class name length greater than the maximum permitted. Most likely cause is a corrupted
- archive or an attempt to insert a virus via the buffer overrun method.
- <h3><a name="unregistered_cast"><code style="white-space: normal">unregistered_cast</code></a></h3>
- In order to support casting between pointers of base and derived classes
- at runtime, a collection of legitimate conversions is maintained by the system.
- Normally this collection is maintained without any explicit action
- on the part of the user of the library. However, there are special cases
- where this might have to be done explicitly and could be overlooked. This
- is described in <a href="serialization.html#runtimecasting">Runtime Casting</a>.
- This exception is thrown if an attempt is made to convert between two pointers
- whose relationship has not been registered,
- <h3><a name="multiple_code_instantiation"><code style="white-space: normal">multiple_code_instantiation</code></a></h3>
- This exception is thrown when it is detected that the serialization of the same type
- has been instantiated more than once. This might occur when
- serialization code is instantiated in both the mainline and one or more DLLS.
- <h3><a name="xml_archive_parsing_error"><code style="white-space: normal">xml_archive_parsing_error</code></a></h3>
- The XML generated by the serialization process is intimately coupled to the
- C++ class structure, relationships between objects and the serialization
- specifications. If these become out of sync in any way, the XML may not map
- to the loading serialization and this exception might be thrown. This might
- occur for one of the following reasons:
- <ul>
- <li>The archive has been edited outside the serialization system. This might
- be possible if only the data is changed and the XML attributes and nesting
- structure are left unaltered. But any other editing is likely to render the
- archive unreadable by the serialization library.
- <li>The serialization has been altered and an archive generated by the old
- code is being read. That is, versioning has not been properly employed to
- properly deserialize previously created archives.
- </ul>
- <h3><a name="xml_archive_tag_mismatch"><code style="white-space: normal">xml_archive_tag_mismatch</code></a></h3>
- This exception will be thrown if the start or end tag of an XML element doesn't match
- the name specified for the object in the program.
- <h3><a name="xml_archive_tag_name_error"><code style="white-space: normal">xml_archive_tag_name_error</code></a></h3>
- This exception will be thrown if the tag name contains invalid characters. Valid characters
- for an XML tag are: upper and lower case letters, digits, and the following punctuation: .(period),
- _(underscore), :(colon), and -(hyphen).
- <hr>
- <p><i>© Copyright <a href="http://www.rrsd.com">Robert Ramey</a> 2002-2004.
- Distributed under the Boost Software License, Version 1.0. (See
- accompanying file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- </i></p>
- </body>
- </html>
|