1 <!-- $Id: book.xml,v 1.26 2006-04-27 16:38:13 adam Exp $ -->
3 <title>Metaproxy - User's Guide and Reference</title>
5 <firstname>Mike</firstname><surname>Taylor</surname>
8 <firstname>Adam</firstname><surname>Dickmeiss</surname>
12 <holder>Index Data ApS</holder>
16 Metaproxy is a universal router, proxy and encapsulated
17 metasearcher for information retrieval protocols. It accepts,
18 processes, interprets and redirects requests from IR clients using
19 standard protocols such as
20 <ulink url="&url.z39.50;">ANSI/NISO Z39.50</ulink>
21 (and in the future <ulink url="&url.sru;">SRU</ulink>
22 and <ulink url="&url.srw;">SRW</ulink>), as
23 well as functioning as a limited
24 <ulink url="&url.http;">HTTP</ulink> server.
25 Metaproxy is configured by an XML file which
26 specifies how the software should function in terms of routes that
27 the request packets can take through the proxy, each step on a
28 route being an instantiation of a filter. Filters come in many
29 types, one for each operation: accepting Z39.50 packets, logging,
30 query transformation, multiplexing, etc. Further filter-types can
31 be added as loadable modules to extend Metaproxy functionality,
35 The terms under which Metaproxy will be distributed have yet to be
36 established, but it will not necessarily be open source; so users
37 should not at this stage redistribute the code without explicit
38 written permission from the copyright holders, Index Data ApS.
43 <imagedata fileref="common/id.png" format="PNG"/>
46 <imagedata fileref="common/id.eps" format="EPS"/>
53 <chapter id="introduction">
54 <title>Introduction</title>
58 <ulink url="&url.metaproxy;">Metaproxy</ulink>
59 is a standalone program that acts as a universal router, proxy and
60 encapsulated metasearcher for information retrieval protocols such
61 as <ulink url="&url.z39.50;">Z39.50</ulink>, and in the future
62 <ulink url="&url.sru;">SRU</ulink> and <ulink url="&url.srw;">SRW</ulink>.
63 To clients, it acts as a server of these protocols: it can be searched,
64 records can be retrieved from it, etc.
65 To servers, it acts as a client: it searches in them,
66 retrieves records from them, etc. it satisfies its clients'
67 requests by transforming them, multiplexing them, forwarding them
68 on to zero or more servers, merging the results, transforming
69 them, and delivering them back to the client. In addition, it
70 acts as a simple <ulink url="&url.http;">HTTP</ulink> server; support
71 for further protocols can be added in a modular fashion, through the
72 creation of new filters.
77 Fish, bananas, cold pyjamas,
78 Mutton, beef and trout!
79 - attributed to Cole Porter.
82 Metaproxy is a more capable alternative to
83 <ulink url="&url.yazproxy;">YAZ Proxy</ulink>,
84 being more powerful, flexible, configurable and extensible. Among
85 its many advantages over the older, more pedestrian work are
86 support for multiplexing (encapsulated metasearching), routing by
87 database name, authentication and authorisation and serving local
88 files via HTTP. Equally significant, its modular architecture
89 facilitites the creation of pluggable modules implementing further
93 This manual will briefly describe Metaproxy's licensing situation
94 before giving an overview of its architecture, then discussing the
95 key concept of a filter in some depth and giving an overview of
96 the various filter types, then discussing the configuration file
97 format. After this come several optional chapters which may be
98 freely skipped: a detailed discussion of virtual databases and
99 multi-database searching, some notes on writing extensions
100 (additional filter types) and a high-level description of the
101 source code. Finally comes the reference guide, which contains
102 instructions for invoking the <command>metaproxy</command>
103 program, and detailed information on each type of filter,
110 <chapter id="licence">
111 <title>The Metaproxy Licence</title>
113 <emphasis role="strong">
114 No decision has yet been made on the terms under which
115 Metaproxy will be distributed.
117 It is possible that, unlike
118 other Index Data products, metaproxy may not be released under a
119 free-software licence such as the GNU GPL. Until a decision is
120 made and a public statement made, then, and unless it has been
121 delivered to you other specific terms, please treat Metaproxy as
122 though it were proprietary software.
123 The code should not be redistributed without explicit
124 written permission from the copyright holders, Index Data ApS.
128 <chapter id="installation">
129 <title>Installation</title>
131 Metaproxy depends on the following tools/libraries:
133 <varlistentry><term><ulink url="&url.yazplusplus;">YAZ++</ulink></term>
136 This is a C++ library based on <ulink url="&url.yaz;">YAZ</ulink>.
140 <varlistentry><term><ulink url="&url.libxslt;">Libxslt</ulink></term>
142 <para>This is an XSLT processor - based on
143 <ulink url="&url.libxml2;">Libxml2</ulink>. Both Libxml2 and
144 Libxslt must be installed with the development components
145 (header files, etc.) as well as the run-time libraries.
149 <varlistentry><term><ulink url="&url.boost;">Boost</ulink></term>
152 The popular C++ library. Initial versions of Metaproxy
153 was built with 1.33.0. Version 1.33.1 works too.
160 In order to compile Metaproxy a modern C++ compiler is
161 required. Boost, in particular, requires the C++ compiler
162 to facilitate the newest features. Refer to Boost
163 <ulink url="&url.boost.compilers.status;">Compiler Status</ulink>
164 for more information.
167 We have succesfully used Metaproxy with Boost using the compilers
168 <ulink url="&url.gcc;">GCC</ulink> version 4.0 and
169 <ulink url="&url.vstudio;">Microsoft Visual Studio</ulink> 2003/2005.
172 <section id="installation.unix">
173 <title>Installation on Unix (from Source)</title>
175 Here is a quick step-by-step guide on how to compile all the
176 tools that Metaproxy uses. Only few systems have none of the required
177 tools binary packages. If, for example, Libxml2/libxslt are already
178 installed as development packages use those (and omit compilation).
185 gunzip -c libxml2-version.tar.gz|tar xf -
193 gunzip -c libxslt-version.tar.gz|tar xf -
204 gunzip -c yaz-version.tar.gz|tar xf -
212 gunzip -c yazpp-version.tar.gz|tar xf -
223 gunzip -c boost-version.tar.gz|tar xf -
234 gunzip -c metaproxy-version.tar.gz|tar xf -
243 <section id="installation.debian">
244 <title>Installation on Debian</title>
249 (Of course, since Debian is a Unix system, the instructions in the
250 previous section can be used.)
254 <section id="installation.windows">
255 <title>Installation on Windows</title>
257 Compilation of Metaproxy can be done using
258 Microsoft <ulink url="&url.vstudio;">Visual Studio</ulink>.
259 We know Version 2003 works. We expect Version 2005 to
262 <section id="installation.windows.boost">
265 Get Boost from its <ulink url="&url.boost;">home page</ulink>.
266 You also need Boost Jam (an alternative to make).
267 That's also available from this
268 home page. The files download are called something like:
269 <literal>boost_1_33-1.exe</literal>
271 <literal>boost-jam-3.1.12-1-ntx86.zip</literal>.
272 Unpack Boost Jam first. Put <literal>bjam.exe</literal>
273 in your system path. Make a command prompt and ensure
274 it can be found automatically. If not check the PATH.
275 The Boost .exe is a self-extracting exe with
276 complete source for Boost. Compile that source with
277 Boost Jam (An alternative to Make).
278 The compilation takes a while.
279 By default, the Boost build process puts the resulting
280 libraries + header files in
281 <literal>\boost\lib</literal>, <literal>\boost\include</literal>.
284 For more informatation about installing Boost refer to the
285 <ulink url="&url.boost.getting.started;">getting started</ulink>
290 <section id="installation.windows.libxslt">
291 <title>Libxslt</title>
293 <ulink url="&url.libxslt;">Libxslt</ulink> can be downloaded
295 <ulink url="&url.libxml2.download.win32;">here</ulink>.
298 Libxslt has other dependencies, but thes can all be downloaded
299 from the same site. Get the following:
300 iconv, zlib, libxml2, libxslt.
304 <section id="installation.windows.yaz">
307 <ulink url="&url.yaz;">YAZ</ulink> can be downloaded
309 <ulink url="&url.yaz.download.win32;">here</ulink>.
313 <section id="installation.windows.yazplusplus">
316 Get <ulink url="&url.yazplusplus;">YAZ++</ulink> as well.
317 Version 1.0 or later is required. For now get it from
319 <ulink url="&url.snapshot.download;">Snapshot area</ulink>.
322 YAZ++ includes NMAKE makefiles, similar to those found in the
327 <section id="installation.windows.metaproxy">
328 <title>Metaproxy</title>
330 Metaproxy is shipped with NMAKE makfiles as well - similar
331 to those found in the YAZ++/YAZ packages. Adjust this Makefile
332 to point to the proper locations of Boost, Libxslt, Libxml2,
333 zlib, iconv, yaz and yazpp.
336 After succesful compilation you'll find
337 <literal>metaproxy.exe</literal> in the
338 <literal>bin</literal> directory.
345 <chapter id="architecture">
346 <title>The Metaproxy Architecture</title>
348 The Metaproxy architecture is based on three concepts:
349 the <emphasis>package</emphasis>,
350 the <emphasis>route</emphasis>
351 and the <emphasis>filter</emphasis>.
355 <term>Packages</term>
358 A package is request or response, encoded in some protocol,
359 issued by a client, making its way through Metaproxy, send to or
360 received from a server, or sent back to the client.
363 The core of a package is the protocol unit - for example, a
364 Z39.50 Init Request or Search Response, or an SRU searchRetrieve
365 URL or Explain Response. In addition to this core, a package
366 also carries some extra information added and used by Metaproxy
370 In general, packages are doctored as they pass through
371 Metaproxy. For example, when the proxy performs authentication
372 and authorisation on a Z39.50 Init request, it removes the
373 authentication credentials from the package so that they are not
374 passed onto the back-end server; and when search-response
375 packages are obtained from multiple servers, they are merged
376 into a single unified package that makes its way back to the
385 Packages make their way through routes, which can be thought of
386 as programs that operate on the package data-type. Each
387 incoming package initially makes its way through a default
388 route, but may be switched to a different route based on various
389 considerations. Routes are made up of sequences of filters (see
398 Filters provide the individual instructions within a route, and
399 effect the necessary transformations on packages. A particular
400 configuration of Metaproxy is essentially a set of filters,
401 described by configuration details and arranged in order in one
402 or more routes. There are many kinds of filter - about a dozen
403 at the time of writing with more appearing all the time - each
404 performing a specific function and configured by different
408 The word ``filter'' is sometimes used rather loosely, in two
409 different ways: it may be used to mean a particular
410 <emphasis>type</emphasis> of filter, as when we speak of ``the
411 auth_simplefilter'' or ``the multi filter''; or it may be used
412 to be a specific <emphasis>instance</emphasis> of a filter
413 within a Metaproxy configuration. For example, a single
414 configuration will often contain multiple instances of the
415 <literal>z3950_client</literal> filter. In
416 operational terms, of these is a separate filter. In practice,
417 context always make it clear which sense of the word ``filter''
421 Extensibility of Metaproxy is primarily through the creation of
422 plugins that provide new filters. The filter API is small and
423 conceptually simple, but there are many details to master. See
425 <link linkend="extensions">extensions</link>.
431 Since packages are created and handled by the system itself, and
432 routes are conceptually simple, most of the remainder of this
433 document concentrates on filters. After a brief overview of the
434 filter types follows, along with some thoughts on possible future
441 <chapter id="filters">
442 <title>Filters</title>
446 <title>Introductory notes</title>
448 It's useful to think of Metaproxy as an interpreter providing a small
449 number of primitives and operations, but operating on a very
450 complex data type, namely the ``package''.
453 A package represents a Z39.50 or SRU/W request (whether for Init,
454 Search, Scan, etc.) together with information about where it came
455 from. Packages are created by front-end filters such as
456 <literal>frontend_net</literal> (see below), which reads them from
457 the network; other front-end filters are possible. They then pass
458 along a route consisting of a sequence of filters, each of which
459 transforms the package and may also have side-effects such as
460 generating logging. Eventually, the route will yield a response,
461 which is sent back to the origin.
464 There are many kinds of filter: some that are defined statically
465 as part of Metaproxy, and others may be provided by third parties
466 and dynamically loaded. They all conform to the same simple API
467 of essentially two methods: <function>configure()</function> is
468 called at startup time, and is passed a DOM tree representing that
469 part of the configuration file that pertains to this filter
470 instance: it is expected to walk that tree extracting relevant
471 information; and <function>process()</function> is called every
472 time the filter has to processes a package.
475 While all filters provide the same API, there are different modes
476 of functionality. Some filters are sources: they create
478 (<literal>frontend_net</literal>);
479 others are sinks: they consume packages and return a result
480 (<literal>z3950_client</literal>,
481 <literal>backend_test</literal>,
482 <literal>http_file</literal>);
483 the others are true filters, that read, process and pass on the
484 packages they are fed
485 (<literal>auth_simple</literal>,
486 <literal>log</literal>,
487 <literal>multi</literal>,
488 <literal>query_rewrite</literal>,
489 <literal>session_shared</literal>,
490 <literal>template</literal>,
491 <literal>virt_db</literal>).
496 <section id="overview.filter.types">
497 <title>Overview of filter types</title>
499 We now briefly consider each of the types of filter supported by
500 the core Metaproxy binary. This overview is intended to give a
501 flavour of the available functionality; more detailed information
502 about each type of filter is included below in
503 <link linkend="filterref"
504 >the reference guide to Metaproxy filters</link>.
507 The filters are here named by the string that is used as the
508 <literal>type</literal> attribute of a
509 <literal><filter></literal> element in the configuration
510 file to request them, with the name of the class that implements
511 them in parentheses. (The classname is not needed for normal
512 configuration and use of Metaproxy; it is useful only to
516 The filters are here listed in alphabetical order:
520 <title><literal>auth_simple</literal>
521 (mp::filter::AuthSimple)</title>
523 Simple authentication and authorisation. The configuration
524 specifies the name of a file that is the user register, which
525 lists <varname>username</varname>:<varname>password</varname>
526 pairs, one per line, colon separated. When a session begins, it
527 is rejected unless username and passsword are supplied, and match
528 a pair in the register. The configuration file may also specific
529 the name of another file that is the target register: this lists
530 lists <varname>username</varname>:<varname>dbname</varname>,<varname>dbname</varname>...
531 sets, one per line, with multiple database names separated by
532 commas. When a search is processed, it is rejected unless the
533 database to be searched is one of those listed as available to
539 <title><literal>backend_test</literal>
540 (mp::filter::Backend_test)</title>
542 A sink that provides dummy responses in the manner of the
543 <literal>yaz-ztest</literal> Z39.50 server. This is useful only
544 for testing. Seriously, you don't need this. Pretend you didn't
545 even read this section.
550 <title><literal>frontend_net</literal>
551 (mp::filter::FrontendNet)</title>
553 A source that accepts Z39.50 connections from a port
554 specified in the configuration, reads protocol units, and
555 feeds them into the next filter in the route. When the result is
556 revceived, it is returned to the original origin.
561 <title><literal>http_file</literal>
562 (mp::filter::HttpFile)</title>
564 A sink that returns the contents of files from the local
565 filesystem in response to HTTP requests. (Yes, Virginia, this
566 does mean that Metaproxy is also a Web-server in its spare time. So
567 far it does not contain either an email-reader or a Lisp
568 interpreter, but that day is surely coming.)
573 <title><literal>log</literal>
574 (mp::filter::Log)</title>
576 Writes logging information to standard output, and passes on
577 the package unchanged.
582 <title><literal>multi</literal>
583 (mp::filter::Multi)</title>
585 Performs multi-database searching.
587 <link linkend="multidb">the extended discussion</link>
588 of virtual databases and multi-database searching below.
593 <title><literal>query_rewrite</literal>
594 (mp::filter::QueryRewrite)</title>
596 Rewrites Z39.50 Type-1 and Type-101 (``RPN'') queries by a
597 three-step process: the query is transliterated from Z39.50
598 packet structures into an XML representation; that XML
599 representation is transformed by an XSLT stylesheet; and the
600 resulting XML is transliterated back into the Z39.50 packet
606 <title><literal>session_shared</literal>
607 (mp::filter::SessionShared)</title>
609 When this is finished, it will implement global sharing of
610 result sets (i.e. between threads and therefore between
611 clients), yielding performance improvements especially when
612 incoming requests are from a stateless environment such as a
613 web-server, in which the client process representing a session
614 might be any one of many. However:
618 This filter is not yet completed.
624 <title><literal>template</literal>
625 (mp::filter::Template)</title>
627 Does nothing at all, merely passing the packet on. (Maybe it
628 should be called <literal>nop</literal> or
629 <literal>passthrough</literal>?) This exists not to be used, but
630 to be copied - to become the skeleton of new filters as they are
631 written. As with <literal>backend_test</literal>, this is not
632 intended for civilians.
637 <title><literal>virt_db</literal>
638 (mp::filter::Virt_db)</title>
640 Performs virtual database selection: based on the name of the
641 database in the search request, a server is selected, and its
642 address added to the request in a <literal>VAL_PROXY</literal>
643 otherInfo packet. It will subsequently be used by a
644 <literal>z3950_client</literal> filter.
646 <link linkend="multidb">the extended discussion</link>
647 of virtual databases and multi-database searching below.
652 <title><literal>z3950_client</literal>
653 (mp::filter::Z3950Client)</title>
655 Performs Z39.50 searching and retrieval by proxying the
656 packages that are passed to it. Init requests are sent to the
657 address specified in the <literal>VAL_PROXY</literal> otherInfo
658 attached to the request: this may have been specified by client,
659 or generated by a <literal>virt_db</literal> filter earlier in
660 the route. Subsequent requests are sent to the same address,
661 which is remembered at Init time in a Session object.
667 <section id="future.directions">
668 <title>Future directions</title>
670 Some other filters that do not yet exist, but which would be
671 useful, are briefly described. These may be added in future
672 releases (or may be created by third parties, as loadable
678 <term><literal>frontend_cli</literal> (source)</term>
681 Command-line interface for generating requests.
686 <term><literal>frontend_sru</literal> (source)</term>
689 Receive SRU (and perhaps SRW) requests.
694 <term><literal>sru2z3950</literal> (filter)</term>
697 Translate SRU requests into Z39.50 requests.
702 <term><literal>sru_client</literal> (sink)</term>
705 SRU searching and retrieval.
710 <term><literal>srw_client</literal> (sink)</term>
713 SRW searching and retrieval.
718 <term><literal>opensearch_client</literal> (sink)</term>
721 A9 OpenSearch searching and retrieval.
731 <chapter id="configuration">
732 <title>Configuration: the Metaproxy configuration file format</title>
736 <title>Introductory notes</title>
738 If Metaproxy is an interpreter providing operations on packages, then
739 its configuration file can be thought of as a program for that
740 interpreter. Configuration is by means of a single file, the name
741 of which is supplied as the sole command-line argument to the
742 <command>metaproxy</command> program. (See
743 <link linkend="progref">the reference guide</link>
744 below for more information on invoking Metaproxy.)
747 The configuration files are written in XML. (But that's just an
748 implementation detail - they could just as well have been written
749 in YAML or Lisp-like S-expressions, or in a custom syntax.)
752 Since XML has been chosen, an XML schema,
753 <filename>config.xsd</filename>, is provided for validating
754 configuration files. This file is supplied in the
755 <filename>etc</filename> directory of the Metaproxy distribution. It
756 can be used by (among other tools) the <command>xmllint</command>
757 program supplied as part of the <literal>libxml2</literal>
761 xmllint --noout --schema etc/config.xsd my-config-file.xml
764 (A recent version of <literal>libxml2</literal> is required, as
765 support for XML Schemas is a relatively recent addition.)
769 <section id="overview.xml.structure">
770 <title>Overview of XML structure</title>
772 All elements and attributes are in the namespace
773 <ulink url="http://indexdata.dk/yp2/config/1"/>.
774 This is most easily achieved by setting the default namespace on
775 the top-level element, as here:
778 <yp2 xmlns="http://indexdata.dk/yp2/config/1">
781 The top-level element is <yp2>. This contains a
782 <start> element, a <filters> element and a
783 <routes> element, in that order. <filters> is
784 optional; the other two are mandatory. All three are
788 The <start> element is empty, but carries a
789 <literal>route</literal> attribute, whose value is the name of
790 route at which to start running - analogous to the name of the
791 start production in a formal grammar.
794 If present, <filters> contains zero or more <filter>
795 elements. Each filter carries a <literal>type</literal> attribute
796 which specifies what kind of filter is being defined
797 (<literal>frontend_net</literal>, <literal>log</literal>, etc.)
798 and contain various elements that provide suitable configuration
799 for a filter of its type. The filter-specific elements are
801 <link linkend="filterref">the reference guide below</link>.
802 Filters defined in this part of the file must carry an
803 <literal>id</literal> attribute so that they can be referenced
807 <routes> contains one or more <route> elements, each
808 of which must carry an <literal>id</literal> element. One of the
809 routes must have the ID value that was specified as the start
810 route in the <start> element's <literal>route</literal>
811 attribute. Each route contains zero or more <filter>
812 elements. These are of two types. They may be empty, but carry a
813 <literal>refid</literal> attribute whose value is the same as the
814 <literal>id</literal> of a filter previously defined in the
815 <filters> section. Alternatively, a route within a filter
816 may omit the <literal>refid</literal> attribute, but contain
817 configuration elements similar to those used for filters defined
818 in the <filters> section. (In other words, each filter in a
819 route may be included either by reference or by physical
825 <section id="example.configuration">
826 <title>An example configuration</title>
828 The following is a small, but complete, Metaproxy configuration
829 file (included in the distribution as
830 <literal>metaproxy/etc/config0.xml</literal>).
831 This file defines a very simple configuration that simply proxies
832 to whatever back-end server the client requests, but logs each
833 request and response. This can be useful for debugging complex
834 client-server dialogues.
836 <screen><![CDATA[<?xml version="1.0"?>
837 <yp2 xmlns="http://indexdata.dk/yp2/config/1">
838 <start route="start"/>
840 <filter id="frontend" type="frontend_net">
843 <filter id="backend" type="z3950_client">
848 <filter refid="frontend"/>
850 <filter refid="backend"/>
856 It works by defining a single route, called
857 <literal>start</literal>, which consists of a sequence of three
858 filters. The first and last of these are included by reference:
859 their <literal><filter></literal> elements have
860 <literal>refid</literal> attributes that refer to filters defined
861 within the prior <literal><filters></literal> section. The
862 middle filter is included inline in the route.
865 The three filters in the route are as follows: first, a
866 <literal>frontend_net</literal> filter accepts Z39.50 requests
867 from any host on port 9000; then these requests are passed through
868 a <literal>log</literal> filter that emits a message for each
869 request; they are then fed into a <literal>z3950_client</literal>
870 filter, which forwards the requests to the client-specified
871 back-end Z39.509 server. When the response arrives, it is handed
872 back to the <literal>log</literal> filter, which emits another
873 message; and then to the front-end filter, which returns the
874 response to the client.
881 <chapter id="multidb">
882 <title>Virtual databases and multi-database searching</title>
886 <title>Introductory notes</title>
888 Two of Metaproxy's filters are concerned with multiple-database
889 operations. Of these, <literal>virt_db</literal> can work alone
890 to control the routing of searches to one of a number of servers,
891 while <literal>multi</literal> can work together with
892 <literal>virt_db</literal> to perform multi-database searching, merging
893 the results into a unified result-set - ``metasearch in a box''.
896 The interaction between
897 these two filters is necessarily complex: it reflects the real,
898 irreducible complexity of multi-database searching in a protocol such
899 as Z39.50 that separates initialisation from searching, and in
900 which the database to be searched is not known at initialisation
904 It's possible to use these filters without understanding the
905 details of their functioning and the interaction between them; the
906 next two sections of this chapter are ``HOWTO'' guides for doing
907 just that. However, debugging complex configurations will require
908 a deeper understanding, which the last two sections of this
909 chapters attempt to provide.
914 <section id="multidb.virt_db">
915 <title>Virtual databases with the <literal>virt_db</literal> filter</title>
917 Working alone, the purpose of the
918 <literal>virt_db</literal>
919 filter is to route search requests to one of a selection of
920 back-end databases. In this way, a single Z39.50 endpoint
921 (running Metaproxy) can provide access to several different
922 underlying services, including those that would otherwise be
923 inaccessible due to firewalls. In many useful configurations, the
924 back-end databases are local to the Metaproxy installation, but
925 the software does not enforce this, and any valid Z39.50 servers
926 may be used as back-ends.
929 For example, a <literal>virt_db</literal>
930 filter could be set up so that searches in the virtual database
931 ``lc'' are forwarded to the Library of Congress bibliographic
932 catalogue server, and searches in the virtual database ``marc''
933 are forwarded to the toy database of MARC records that Index Data
934 hosts for testing purposes. A <literal>virt_db</literal>
935 configuration to make this switch would look like this:
937 <screen><![CDATA[<filter type="virt_db">
939 <database>lc</database>
940 <target>z3950.loc.gov:7090/voyager</target>
943 <database>marc</database>
944 <target>indexdata.dk/marc</target>
946 </filter>]]></screen>
948 As well as being useful in it own right, this filter also provides
949 the foundation for multi-database searching.
954 <section id="multidb.multi">
955 <title>Multi-database search with the <literal>multi</literal> filter</title>
957 To arrange for Metaproxy to broadcast searches to multiple back-end
958 servers, the configuration needs to include two components: a
959 <literal>virt_db</literal>
960 filter that specifies multiple
961 <literal><target></literal>
962 elements, and a subsequent
963 <literal>multi</literal>
964 filter. Here, for example, is a complete configuration that
965 broadcasts searches to both the Library of Congress catalogue and
966 Index Data's tiny testing database of MARC records:
968 <screen><![CDATA[<?xml version="1.0"?>
969 <yp2 xmlns="http://indexdata.dk/yp2/config/1">
970 <start route="start"/>
973 <filter type="frontend_net">
974 <threads>10</threads>
977 <filter type="virt_db">
979 <database>lc</database>
980 <target>z3950.loc.gov:7090/voyager</target>
983 <database>marc</database>
984 <target>indexdata.dk/marc</target>
987 <database>all</database>
988 <target>z3950.loc.gov:7090/voyager</target>
989 <target>indexdata.dk/marc</target>
992 <filter type="multi"/>
993 <filter type="z3950_client">
994 <timeout>30</timeout>
1001 <literal>virt_db</literal>
1002 filter that specifies multiple
1003 <literal><target></literal>
1004 elements but without a subsequent
1005 <literal>multi</literal>
1006 filter yields surprising and undesirable results, as will be
1007 described below. Don't do that.)
1010 Metaproxy can be invoked with this configuration as follows:
1012 <screen>../src/metaproxy --config config-simple-multi.xml</screen>
1014 And thereafter, Z39.50 clients can connect to the running server
1015 (on port 9000, as specified in the configuration) and search in
1016 any of the databases
1017 <literal>lc</literal> (the Library of Congress catalogue),
1018 <literal>marc</literal> (Index Data's test database of MARC records)
1020 <literal>all</literal> (both of these). As an example, a session
1021 using the YAZ command-line client <literal>yaz-client</literal> is
1022 here included (edited for brevity and clarity):
1024 <screen><![CDATA[$ yaz-client @:9000
1028 Search was a success.
1029 Number of hits: 10000, setno 1
1033 Search was a success.
1034 Number of hits: 10, setno 3
1038 Search was a success.
1039 Number of hits: 10010, setno 4
1042 [marc]Record type: USmarc
1045 005 00000000000000.0
1046 008 910710c19910701nju 00010 eng
1050 100 10 $a Jack Collins
1051 245 10 $a How to program a computer
1057 [VOYAGER]Record type: USmarc
1059 005 20041229102447.0
1060 008 030910s2004 caua 000 0 eng
1061 035 $a (DLC) 2003112666
1062 906 $a 7 $b cbc $c orignew $d 4 $e epcn $f 20 $g y-gencatlg
1063 925 0 $a acquire $b 1 shelf copy $x policy default
1064 955 $a pc10 2003-09-10 $a pv12 2004-06-23 to SSCD; $h sj05 2004-11-30 $e sj05 2004-11-30 to Shelf.
1067 040 $a DLC $c DLC $d DLC
1068 050 00 $a MLCM 2004/03312 (G)
1069 245 10 $a 007, everything or nothing : $b Prima's official strategy guide / $c created by Kaizen Media Group.
1070 246 3 $a Double-O-seven, everything or nothing
1071 246 30 $a Prima's official strategy guide
1072 260 $a Roseville, CA : $b Prima Games, $c c2004.
1073 300 $a 161 p. : $b col. ill. ; $c 28 cm.
1074 500 $a "Platforms: Nintendo GameCube, Macintosh, PC, PlayStation 2 computer entertainment system, Xbox"--P. [4] of cover.
1075 650 0 $a Video games.
1076 710 2 $a Kaizen Media Group.
1077 856 42 $3 Publisher description $u http://www.loc.gov/catdir/description/random052/2003112666.html
1082 As can be seen, the first record in the result set is from the
1083 Index Data test database, and the second from the Library of
1084 Congress database. The result-set continues alternating records
1085 round-robin style until the point where one of the databases'
1086 records are exhausted.
1089 This example uses only two back-end databases; more may be used.
1090 There is no limitation imposed on the number of databases that may
1091 be metasearched in this way: issues of resource usage and
1092 administrative complexity dictate the practical limits.
1097 <section id="multidb.what">
1098 <title>What's going on?</title>
1100 <title>Lark's vomit</title>
1102 This section goes into a level of technical detail that is
1103 probably not necessary in order to configure and use Metaproxy.
1104 It is provided only for those who like to know how things work.
1105 You should feel free to skip on to the next section if this one
1106 doesn't seem like fun.
1110 Hold on tight - this may get a little hairy.
1113 In the general course of things, a Z39.50 Init request may carry
1114 with it an otherInfo packet of type <literal>VAL_PROXY</literal>,
1115 whose value indicates the address of a Z39.50 server to which the
1116 ultimate connection is to be made. (This otherInfo packet is
1117 supported by YAZ-based Z39.50 clients and servers, but has not yet
1118 been ratified by the Maintenance Agency and so is not widely used
1119 in non-Index Data software. We're working on it.)
1120 The <literal>VAL_PROXY</literal> packet functions
1121 analogously to the absoluteURI-style Request-URI used with the GET
1122 method when a web browser asks a proxy to forward its request: see
1124 <ulink url="http://www.w3.org/Protocols/rfc2616/rfc2616-sec5.html#sec5.1.2"
1125 >Request-URI</ulink>
1127 <ulink url="http://www.w3.org/Protocols/rfc2616/rfc2616.html"
1128 >the HTTP 1.1 specification</ulink>.
1131 The role of the <literal>virt_db</literal> filter is to rewrite
1132 this otherInfo packet dependent on the virtual database that the
1133 client wants to search.
1136 When Metaproxy receives a Z39.50 Init request from a client, it
1137 doesn't immediately forward that request to the back-end server.
1138 Why not? Because it doesn't know <emphasis>which</emphasis>
1139 back-end server to forward it to until the client sends a search
1140 request that specifies the database that it wants to search in.
1141 Instead, it just treasures the Init request up in its heart; and,
1142 later, the first time the client does a search on one of the
1143 specified virtual databases, a connection is forged to the
1144 appropriate server and the Init request is forwarded to it. If,
1145 later in the session, the same client searches in a different
1146 virtual database, then a connection is forged to the server that
1147 hosts it, and the same cached Init request is forwarded there,
1151 All of this clever Init-delaying is done by the
1152 <literal>frontend_net</literal> filter. The
1153 <literal>virt_db</literal> filter knows nothing about it; in
1154 fact, because the Init request that is received from the client
1155 doesn't get forwarded until a Search request is received, the
1156 <literal>virt_db</literal> filter (and the
1157 <literal>z3950_client</literal> filter behind it) doesn't even get
1158 invoked at Init time. The <emphasis>only</emphasis> thing that a
1159 <literal>virt_db</literal> filter ever does is rewrite the
1160 <literal>VAL_PROXY</literal> otherInfo in the requests that pass
1164 ### Describe the use of multiple <literal>VAL_PROXY</literal>
1165 otherInfos, added by <literal>virt_db</literal> and used by
1166 <literal>multi</literal>.
1171 <section id="multidb.picture">
1172 <title>A picture is worth a thousand words (but only five hundred on 64-bit architectures)</title>
1176 <imagedata fileref="multi.pdf" format="PDF" scale="50"/>
1179 <imagedata fileref="multi.png" format="PNG"/>
1182 <!-- Fall back if none of the images can be used -->
1184 [Here there should be a diagram showing the progress of
1185 packages through the filters during a simple virtual-database
1186 search and a multi-database search, but is seems that your
1187 toolchain has not been able to include the diagram in this
1188 document. This is because of LaTeX suckage. Time to move to
1189 OpenOffice. Yes, really.]
1192 <!-- ### This used to work with an older version of DocBook
1194 <para>Caption: progress of packages through filters.</para>
1197 </inlinemediaobject>
1204 <chapter id="extensions">
1205 <title>Writing extensions for Metaproxy</title>
1206 <para>### To be written</para>
1212 <chapter id="classes">
1213 <title>Classes in the Metaproxy source code</title>
1217 <title>Introductory notes</title>
1219 <emphasis>Stop! Do not read this!</emphasis>
1220 You won't enjoy it at all. You should just skip ahead to
1221 <link linkend="refguide">the reference guide</link>,
1223 <!-- The remainder of this paragraph is lifted verbatim from
1224 Douglas Adams' _Hitch Hiker's Guide to the Galaxy_, chapter 8 -->
1225 you things you really need to know, like the fact that the
1226 fabulously beautiful planet Bethselamin is now so worried about
1227 the cumulative erosion by ten billion visiting tourists a year
1228 that any net imbalance between the amount you eat and the amount
1229 you excrete whilst on the planet is surgically removed from your
1230 bodyweight when you leave: so every time you go to the lavatory it
1231 is vitally important to get a receipt.
1234 This chapter contains documentation of the Metaproxy source code, and is
1235 of interest only to maintainers and developers. If you need to
1236 change Metaproxy's behaviour or write a new filter, then you will most
1237 likely find this chapter helpful. Otherwise it's a waste of your
1238 good time. Seriously: go and watch a film or something.
1239 <citetitle>This is Spinal Tap</citetitle> is particularly good.
1242 Still here? OK, let's continue.
1245 In general, classes seem to be named big-endianly, so that
1246 <literal>FactoryFilter</literal> is not a filter that filters
1247 factories, but a factory that produces filters; and
1248 <literal>FactoryStatic</literal> is a factory for the statically
1249 registered filters (as opposed to those that are dynamically
1254 <section id="individual.classes">
1255 <title>Individual classes</title>
1257 The classes making up the Metaproxy application are here listed by
1258 class-name, with the names of the source files that define them in
1263 <title><literal>mp::FactoryFilter</literal>
1264 (<filename>factory_filter.cpp</filename>)</title>
1266 A factory class that exists primarily to provide the
1267 <literal>create()</literal> method, which takes the name of a
1268 filter class as its argument and returns a new filter of that
1269 type. To enable this, the factory must first be populated by
1270 calling <literal>add_creator()</literal> for static filters (this
1271 is done by the <literal>FactoryStatic</literal> class, see below)
1272 and <literal>add_creator_dyn()</literal> for filters loaded
1278 <title><literal>mp::FactoryStatic</literal>
1279 (<filename>factory_static.cpp</filename>)</title>
1281 A subclass of <literal>FactoryFilter</literal> which is
1282 responsible for registering all the statically defined filter
1283 types. It does this by knowing about all those filters'
1284 structures, which are listed in its constructor. Merely
1285 instantiating this class registers all the static classes. It is
1286 for the benefit of this class that <literal>struct
1287 metaproxy_1_filter_struct</literal> exists, and that all the filter
1288 classes provide a static object of that type.
1293 <title><literal>mp::filter::Base</literal>
1294 (<filename>filter.cpp</filename>)</title>
1296 The virtual base class of all filters. The filter API is, on the
1297 surface at least, extremely simple: two methods.
1298 <literal>configure()</literal> is passed a DOM tree representing
1299 that part of the configuration file that pertains to this filter
1300 instance, and is expected to walk that tree extracting relevant
1301 information. And <literal>process()</literal> processes a
1302 package (see below). That surface simplicitly is a bit
1303 misleading, as <literal>process()</literal> needs to know a lot
1304 about the <literal>Package</literal> class in order to do
1310 <title><literal>mp::filter::AuthSimple</literal>,
1311 <literal>Backend_test</literal>, etc.
1312 (<filename>filter_auth_simple.cpp</filename>,
1313 <filename>filter_backend_test.cpp</filename>, etc.)</title>
1315 Individual filters. Each of these is implemented by a header and
1316 a source file, named <filename>filter_*.hpp</filename> and
1317 <filename>filter_*.cpp</filename> respectively. All the header
1318 files should be pretty much identical, in that they declare the
1319 class, including a private <literal>Rep</literal> class and a
1320 member pointer to it, and the two public methods. The only extra
1321 information in any filter header is additional private types and
1322 members (which should really all be in the <literal>Rep</literal>
1323 anyway) and private methods (which should also remain known only
1324 to the source file, but C++'s brain-damaged design requires this
1325 dirty laundry to be exhibited in public. Thanks, Bjarne!)
1328 The source file for each filter needs to supply:
1333 A definition of the private <literal>Rep</literal> class.
1338 Some boilerplate constructors and destructors.
1343 A <literal>configure()</literal> method that uses the
1344 appropriate XML fragment.
1349 Most important, the <literal>process()</literal> method that
1350 does all the actual work.
1357 <title><literal>mp::Package</literal>
1358 (<filename>package.cpp</filename>)</title>
1360 Represents a package on its way through the series of filters
1361 that make up a route. This is essentially a Z39.50 or SRU APDU
1362 together with information about where it came from, which is
1363 modified as it passes through the various filters.
1368 <title><literal>mp::Pipe</literal>
1369 (<filename>pipe.cpp</filename>)</title>
1371 This class provides a compatibility layer so that we have an IPC
1372 mechanism that works the same under Unix and Windows. It's not
1373 particularly exciting.
1378 <title><literal>mp::RouterChain</literal>
1379 (<filename>router_chain.cpp</filename>)</title>
1386 <title><literal>mp::RouterFleXML</literal>
1387 (<filename>router_flexml.cpp</filename>)</title>
1394 <title><literal>mp::Session</literal>
1395 (<filename>session.cpp</filename>)</title>
1402 <title><literal>mp::ThreadPoolSocketObserver</literal>
1403 (<filename>thread_pool_observer.cpp</filename>)</title>
1410 <title><literal>mp::util</literal>
1411 (<filename>util.cpp</filename>)</title>
1413 A namespace of various small utility functions and classes,
1414 collected together for convenience. Most importantly, includes
1415 the <literal>mp::util::odr</literal> class, a wrapper for YAZ's
1421 <title><literal>mp::xml</literal>
1422 (<filename>xmlutil.cpp</filename>)</title>
1424 A namespace of various XML utility functions and classes,
1425 collected together for convenience.
1431 <section id="other.source.files">
1432 <title>Other Source Files</title>
1434 In addition to the Metaproxy source files that define the classes
1435 described above, there are a few additional files which are
1436 briefly described here:
1440 <term><literal>metaproxy_prog.cpp</literal></term>
1443 The main function of the <command>metaproxy</command> program.
1448 <term><literal>ex_router_flexml.cpp</literal></term>
1451 Identical to <literal>metaproxy_prog.cpp</literal>: it's not clear why.
1456 <term><literal>test_*.cpp</literal></term>
1459 Unit-tests for various modules.
1465 ### Still to be described:
1466 <literal>ex_filter_frontend_net.cpp</literal>,
1467 <literal>filter_dl.cpp</literal>,
1468 <literal>plainfile.cpp</literal>,
1469 <literal>tstdl.cpp</literal>.
1476 <chapter id="refguide">
1477 <title>Reference guide</title>
1479 The material in this chapter is drawn directly from the individual
1480 manual entries. In particular, the Metaproxy invocation section is
1481 available using <command>man metaproxy</command>, and the section
1482 on each individual filter is available using the name of the filter
1483 as the argument to the <command>man</command> command.
1487 <section id="progref">
1488 <title>Metaproxy invocation</title>
1493 <section id="filterref">
1494 <title>Reference guide to Metaproxy filters</title>
1501 <!-- Keep this comment at the end of the file
1506 sgml-minimize-attributes:nil
1507 sgml-always-quote-attributes:t
1510 sgml-parent-document: "main.xml"
1511 sgml-local-catalogs: nil
1512 sgml-namecase-general:t