1 <?xml version="1.0" standalone="no"?>
2 <!DOCTYPE book PUBLIC "-//OASIS//DTD DocBook XML V4.4//EN"
3 "http://www.oasis-open.org/docbook/xml/4.4/docbookx.dtd"
5 <!ENTITY copyright SYSTEM "copyright.xml">
6 <!ENTITY % local SYSTEM "local.ent">
8 <!ENTITY manref SYSTEM "manref.xml">
9 <!ENTITY gpl2 SYSTEM "gpl-2.0.xml">
10 <!ENTITY % idcommon SYSTEM "common/common.ent">
15 <title>Metaproxy - User's Guide and Reference</title>
18 <firstname>Adam</firstname><surname>Dickmeiss</surname>
21 <firstname>Marc</firstname><surname>Cromme</surname>
24 <firstname>Mike</firstname><surname>Taylor</surname>
27 <releaseinfo>&version;</releaseinfo>
29 <year>2005-2011</year>
30 <holder>Index Data</holder>
34 This manual is part of Metaproxy version &version;.
37 Metaproxy is a universal router, proxy and encapsulated
38 metasearcher for information retrieval protocols. It accepts,
39 processes, interprets and redirects requests from IR clients using
40 standard protocols such as the binary
41 <ulink url="&url.z39.50;">ANSI/NISO Z39.50</ulink>
42 and the information search and retrieval
43 web service <ulink url="&url.sru;">SRU</ulink>
44 as well as functioning as a limited
45 <ulink url="&url.http;">HTTP</ulink> server.
48 Metaproxy is configured by an XML file which
49 specifies how the software should function in terms of routes that
50 the request packets can take through the proxy, each step on a
51 route being an instantiation of a filter. Filters come in many
52 types, one for each operation: accepting Z39.50 packets, logging,
53 query transformation, multiplexing, etc. Further filter-types can
54 be added as loadable modules to extend Metaproxy functionality,
58 Metaproxy is covered by the GNU General Public License version 2.
63 <imagedata fileref="common/id.png" format="PNG"/>
66 <imagedata fileref="common/id.eps" format="EPS"/>
73 <chapter id="introduction">
74 <title>Introduction</title>
77 <ulink url="&url.metaproxy;">Metaproxy</ulink>
78 is a stand alone program that acts as a universal router, proxy and
79 encapsulated metasearcher for information retrieval protocols such
80 as <ulink url="&url.z39.50;">Z39.50</ulink> and
81 <ulink url="&url.sru;">SRU</ulink>.
82 To clients, it acts as a server of these protocols: it can be searched,
83 records can be retrieved from it, etc.
84 To servers, it acts as a client: it searches in them,
85 retrieves records from them, etc. it satisfies its clients'
86 requests by transforming them, multiplexing them, forwarding them
87 on to zero or more servers, merging the results, transforming
88 them, and delivering them back to the client. In addition, it
89 acts as a simple <ulink url="&url.http;">HTTP</ulink> server; support
90 for further protocols can be added in a modular fashion, through the
91 creation of new filters.
96 Fish, bananas, cold pyjamas,
97 Mutton, beef and trout!
98 - attributed to Cole Porter.
101 Metaproxy is a more capable alternative to
102 <ulink url="&url.yazproxy;">YAZ Proxy</ulink>,
103 being more powerful, flexible, configurable and extensible. Among
104 its many advantages over the older, more pedestrian work are
105 support for multiplexing (encapsulated metasearching), routing by
106 database name, authentication and authorization and serving local
107 files via HTTP. Equally significant, its modular architecture
108 facilitites the creation of pluggable modules implementing further
112 This manual will describe how to install Metaproxy
113 before giving an overview of its architecture, then discussing the
114 key concept of a filter in some depth and giving an overview of
115 the various filter types, then discussing the configuration file
116 format. After this come several optional chapters which may be
117 freely skipped: a detailed discussion of virtual databases and
118 multi-database searching, some notes on writing extensions
119 (additional filter types) and a high-level description of the
120 source code. Finally comes the reference guide, which contains
121 instructions for invoking the <command>metaproxy</command>
122 program, and detailed information on each type of filter,
127 <chapter id="installation">
128 <title>Installation</title>
130 Metaproxy depends on the following tools/libraries:
132 <varlistentry><term><ulink url="&url.yazplusplus;">YAZ++</ulink></term>
135 This is a C++ library based on <ulink url="&url.yaz;">YAZ</ulink>.
139 <varlistentry><term><ulink url="&url.libxslt;">Libxslt</ulink></term>
141 <para>This is an XSLT processor - based on
142 <ulink url="&url.libxml2;">Libxml2</ulink>. Both Libxml2 and
143 Libxslt must be installed with the development components
144 (header files, etc.) as well as the run-time libraries.
148 <varlistentry><term><ulink url="&url.boost;">Boost</ulink></term>
151 The popular C++ library. Initial versions of Metaproxy
152 was built with 1.32 but this is no longer supported.
153 Metaproxy is known to work with Boost version 1.33 through 1.38.
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 successfully built Metaproxy 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).
181 <section id="libxml2.fromsource">
182 <title>Libxml2/libxslt</title>
187 gunzip -c libxml2-version.tar.gz|tar xf -
195 gunzip -c libxslt-version.tar.gz|tar xf -
203 <section id="yaz.fromsource">
204 <title>YAZ/YAZ++</title>
206 gunzip -c yaz-version.tar.gz|tar xf -
214 gunzip -c yazpp-version.tar.gz|tar xf -
223 <title id="boost.fromsource">Boost</title>
225 Metaproxy needs components thread and test from
229 gunzip -c boost-version.tar.gz|tar xf -
231 ./configure --with-libraries=thread,test --with-toolset=gcc
237 However, under the hood bjam is used. You can invoke that with
240 ./bjam --toolset=gcc --with-thread --with-test stage
243 Replace <literal>stage</literal> with <literal>clean</literal> /
244 <literal>install</literal> to perform clean and install respectively.
247 Add <literal>--prefix=DIR</literal> to install Boost in other
248 prefix than <literal>/usr/local</literal>.
251 <section id="metaproxy.fromsource">
252 <title>Metaproxy</title>
254 gunzip -c metaproxy-version.tar.gz|tar xf -
262 You may have to tell configure where Boost is installed by supplying
263 options <literal>--with-boost</literal> and <literal>--with-boost-toolset</literal>.
264 The former sets the PREFIX for Boost (same as --prefix for Boost above).
265 The latter the compiler toolset (eg. gcc34).
268 Pass <literal>--help</literal> to configure to get a list of
274 <section id="installation.debian">
275 <title>Installation on Debian GNU/Linux</title>
277 All dependencies for Metaproxy are available as
278 <ulink url="&url.debian;">Debian</ulink>
279 packages for the sarge (stable in 2005) and etch (testing in 2005)
283 The procedures for Debian based systems, such as
284 <ulink url="&url.ubuntu;">Ubuntu</ulink> is probably similar
287 There is currently no official Debian package for YAZ++.
288 And the official Debian package for YAZ is probably too old.
289 But Index Data bulds "new" versions of those for Debian (i386 only).
292 Update the <filename>/etc/apt/sources.list</filename>
293 to include the Index Data repository.
294 See YAZ' <ulink url="&url.yaz.download.debian;">Download Debian</ulink>
295 for more information.
298 apt-get install libxslt1-dev
299 apt-get install libyazpp2-dev
300 apt-get install libboost-dev
301 apt-get install libboost-thread-dev
302 apt-get install libboost-test-dev
305 With these packages installed, the usual configure + make
306 procedure can be used for Metaproxy as outlined in
307 <xref linkend="installation.unix"/>.
311 <section id="installation.rpm">
312 <title>Installation on RPM based Linux Systems</title>
314 All external dependencies for Metaproxy are available as
315 RPM packages, either from your distribution site, or from the
316 <ulink url="http://fr.rpmfind.net/">RPMfind</ulink> site.
319 For example, an installation of the requires Boost C++ development
320 libraries on RedHat Fedora C4 and C5 can be done like this:
322 wget ftp://fr.rpmfind.net/wlinux/fedora/core/updates/testing/4/SRPMS/boost-1.33.0-3.fc4.src.rpm
323 sudo rpmbuild --buildroot src/ --rebuild -p fc4/boost-1.33.0-3.fc4.src.rpm
324 sudo rpm -U /usr/src/redhat/RPMS/i386/boost-*rpm
328 The <ulink url="&url.yaz;">YAZ</ulink> library is needed to
329 compile &metaproxy;, see there
330 for more information on available RPM packages.
333 There is currently no official RPM package for YAZ++.
334 See the <ulink url="&url.yazplusplus;">YAZ++</ulink> pages
335 for more information on a Unix tarball install.
338 With these packages installed, the usual configure + make
339 procedure can be used for Metaproxy as outlined in
340 <xref linkend="installation.unix"/>.
344 <section id="installation.windows">
345 <title>Installation on Windows</title>
347 Metaproxy can be compiled with Microsoft
348 <ulink url="&url.vstudio;">Visual Studio</ulink>.
349 Versions 2003 (C 7.1), 2005 (C 8.0) and 2008 (C 9.0) is known to work.
351 <section id="installation.windows.boost">
354 For Windows, it's easiest to get the precompiled Boost
355 package from <ulink url="&url.boost.windows.download;">here</ulink>.
356 Several versions of the Boost libraries may be selected when
357 installing Boost for windows. Please choose at least the
358 <emphasis>multithreaded</emphasis> (non-DLL) version because
359 the Metaproxy makefile uses that.
362 For more information about installing Boost refer to the
363 <ulink url="&url.boost.getting.started;">getting started</ulink>
368 <section id="installation.windows.libxslt">
369 <title>Libxslt</title>
371 <ulink url="&url.libxslt;">Libxslt</ulink> can be downloaded
373 <ulink url="&url.libxml2.download.win32;">here</ulink>.
376 Libxslt has other dependencies, but these can all be downloaded
377 from the same site. Get the following package:
378 iconv, zlib, libxml2, libxslt.
382 <section id="installation.windows.yaz">
385 <ulink url="&url.yaz;">YAZ</ulink> can be downloaded
387 <ulink url="&url.yaz.download.win32;">here</ulink>.
391 <section id="installation.windows.yazplusplus">
394 Get <ulink url="&url.yazplusplus;">YAZ++</ulink> as well.
395 Version 1.1.0 or later is required.
398 YAZ++ includes NMAKE makefiles, similar to those found in the
403 <section id="installation.windows.metaproxy">
404 <title>Metaproxy</title>
406 Metaproxy is shipped with NMAKE makefiles as well - similar
407 to those found in the YAZ++/YAZ packages. Adjust this Makefile
408 to point to the proper locations of Boost, Libxslt, Libxml2,
409 zlib, iconv, yaz and yazpp.
413 <varlistentry><term><literal>DEBUG</literal></term>
415 If set to 1, the software is
416 compiled with debugging libraries (code generation is
417 multi-threaded debug DLL).
418 If set to 0, the software is compiled with release libraries
419 (code generation is multi-threaded DLL).
424 <term><literal>BOOST</literal></term>
427 Boost install location
433 <term><literal>BOOST_VERSION</literal></term>
436 Boost version (replace . with _).
442 <term><literal>BOOST_TOOLSET</literal></term>
451 <term><literal>LIBXSLT_DIR</literal>,
452 <literal>LIBXML2_DIR</literal> ..</term>
455 Specify the locations of Libxslt, libiconv, libxml2 and
464 After successful compilation you'll find
465 <literal>metaproxy.exe</literal> in the
466 <literal>bin</literal> directory.
474 <chapter id="yazproxy-comparison">
475 <title>YAZ Proxy Comparison</title>
477 The table below lists facilities either supported by either
478 <ulink url="&url.yazproxy;">YAZ Proxy</ulink> or Metaproxy.
480 <table id="yazproxy-comparison-table">
481 <title>Metaproxy / YAZ Proxy comparison</title>
485 <entry>Facility</entry>
486 <entry>Metaproxy</entry>
487 <entry>YAZ Proxy</entry>
492 <entry>Z39.50 server</entry>
493 <entry>Using filter <literal>frontend_net</literal></entry>
494 <entry>Supported</entry>
497 <entry>SRU server</entry>
498 <entry>Supported with filter <literal>sru_z3950</literal></entry>
499 <entry>Supported</entry>
502 <entry>Z39.50 client</entry>
503 <entry>Supported with filter <literal>z3950_client</literal></entry>
504 <entry>Supported</entry>
507 <entry>SRU client</entry>
508 <entry>Unsupported</entry>
509 <entry>Unsupported</entry>
512 <entry>Connection reuse</entry>
513 <entry>Supported with filter <literal>session_shared</literal></entry>
514 <entry>Supported</entry>
517 <entry>Connection share</entry>
518 <entry>Supported with filter <literal>session_shared</literal></entry>
519 <entry>Unsupported</entry>
522 <entry>Result set reuse</entry>
523 <entry>Supported with filter <literal>session_shared</literal></entry>
524 <entry>Within one Z39.50 session / HTTP keep-alive</entry>
527 <entry>Record cache</entry>
528 <entry>Supported by filter <literal>session_shared</literal></entry>
529 <entry>Supported for last result set within one Z39.50/HTTP-keep alive session</entry>
532 <entry>Z39.50 Virtual database, i.e. select any Z39.50 target for database</entry>
533 <entry>Supported with filter <literal>virt_db</literal></entry>
534 <entry>Unsupported</entry>
537 <entry>SRU Virtual database, i.e. select any Z39.50 target for path</entry>
538 <entry>Supported with filter <literal>virt_db</literal>,
539 <literal>sru_z3950</literal></entry>
540 <entry>Supported</entry>
543 <entry>Multi target search</entry>
544 <entry>Supported with filter <literal>multi</literal> (round-robin)</entry>
545 <entry>Unsupported</entry>
548 <entry>Retrieval and search limits</entry>
549 <entry>Supported using filter <literal>limit</literal></entry>
550 <entry>Supported</entry>
553 <entry>Bandwidth limits</entry>
554 <entry>Supported using filter <literal>limit</literal></entry>
555 <entry>Supported</entry>
558 <entry>Connect limits</entry>
559 <entry>Supported by filter <literal>frontend_net</literal> (connect-max)</entry>
560 <entry>Supported</entry>
563 <entry>Retrieval sanity check and conversions</entry>
564 <entry>Supported using filter <literal>record_transform</literal></entry>
565 <entry>Supported</entry>
568 <entry>Query check</entry>
570 Supported by <literal>query_rewrite</literal> which may be check
571 a query and throw diagnostics (errors)
573 <entry>Supported</entry>
576 <entry>Query rewrite</entry>
577 <entry>Supported with <literal>query_rewrite</literal></entry>
578 <entry>Unsupported</entry>
581 <entry>Session invalidate for -1 hits</entry>
582 <entry>Unsupported</entry>
583 <entry>Supported</entry>
586 <entry>Architecture</entry>
587 <entry>Multi-threaded + select for networked modules such as
588 <literal>frontend_net</literal>)</entry>
589 <entry>Single-threaded using select</entry>
593 <entry>Extensability</entry>
594 <entry>Most functionality implemented as loadable modules</entry>
595 <entry>Unsupported and experimental</entry>
599 <entry><ulink url="&url.usemarcon;">USEMARCON</ulink></entry>
600 <entry>Unsupported</entry>
601 <entry>Supported</entry>
605 <entry>Portability</entry>
607 Requires YAZ, YAZ++ and modern C++ compiler supporting
608 <ulink url="&url.boost;">Boost</ulink>.
611 Requires YAZ and YAZ++.
612 STL is not required so pretty much any C++ compiler out there should work.
621 <chapter id="architecture">
622 <title>The Metaproxy Architecture</title>
624 The Metaproxy architecture is based on three concepts:
625 the <emphasis>package</emphasis>,
626 the <emphasis>route</emphasis>
627 and the <emphasis>filter</emphasis>.
631 <term>Packages</term>
634 A package is request or response, encoded in some protocol,
635 issued by a client, making its way through Metaproxy, send to or
636 received from a server, or sent back to the client.
639 The core of a package is the protocol unit - for example, a
640 Z39.50 Init Request or Search Response, or an SRU searchRetrieve
641 URL or Explain Response. In addition to this core, a package
642 also carries some extra information added and used by Metaproxy
646 In general, packages are doctored as they pass through
647 Metaproxy. For example, when the proxy performs authentication
648 and authorization on a Z39.50 Init request, it removes the
649 authentication credentials from the package so that they are not
650 passed onto the back-end server; and when search-response
651 packages are obtained from multiple servers, they are merged
652 into a single unified package that makes its way back to the
661 Packages make their way through routes, which can be thought of
662 as programs that operate on the package data-type. Each
663 incoming package initially makes its way through a default
664 route, but may be switched to a different route based on various
665 considerations. Routes are made up of sequences of filters (see
674 Filters provide the individual instructions within a route, and
675 effect the necessary transformations on packages. A particular
676 configuration of Metaproxy is essentially a set of filters,
677 described by configuration details and arranged in order in one
678 or more routes. There are many kinds of filter - about a dozen
679 at the time of writing with more appearing all the time - each
680 performing a specific function and configured by different
684 The word ``filter'' is sometimes used rather loosely, in two
685 different ways: it may be used to mean a particular
686 <emphasis>type</emphasis> of filter, as when we speak of ``the
687 auth_simple filter'' or ``the multi filter''; or it may be used
688 to be a specific <emphasis>instance</emphasis> of a filter
689 within a Metaproxy configuration. For example, a single
690 configuration will often contain multiple instances of the
691 <literal>z3950_client</literal> filter. In
692 operational terms, of these is a separate filter. In practice,
693 context always make it clear which sense of the word ``filter''
697 Extensibility of Metaproxy is primarily through the creation of
698 plugins that provide new filters. The filter API is small and
699 conceptually simple, but there are many details to master. See
701 <link linkend="filters">Filters</link>.
707 Since packages are created and handled by the system itself, and
708 routes are conceptually simple, most of the remainder of this
709 document concentrates on filters. After a brief overview of the
710 filter types follows, along with some thoughts on possible future
717 <chapter id="filters">
718 <title>Filters</title>
721 <section id="filters-introductory-notes">
722 <title>Introductory notes</title>
724 It's useful to think of Metaproxy as an interpreter providing a small
725 number of primitives and operations, but operating on a very
726 complex data type, namely the ``package''.
729 A package represents a Z39.50 or SRU/W request (whether for Init,
730 Search, Scan, etc.) together with information about where it came
731 from. Packages are created by front-end filters such as
732 <literal>frontend_net</literal> (see below), which reads them from
733 the network; other front-end filters are possible. They then pass
734 along a route consisting of a sequence of filters, each of which
735 transforms the package and may also have side-effects such as
736 generating logging. Eventually, the route will yield a response,
737 which is sent back to the origin.
740 There are many kinds of filter: some that are defined statically
741 as part of Metaproxy, and others may be provided by third parties
742 and dynamically loaded. They all conform to the same simple API
743 of essentially two methods: <function>configure()</function> is
744 called at startup time, and is passed an XML DOM tree representing that
745 part of the configuration file that pertains to this filter
746 instance: it is expected to walk that tree extracting relevant
747 information; and <function>process()</function> is called every
748 time the filter has to processes a package.
751 While all filters provide the same API, there are different modes
752 of functionality. Some filters are sources: they create
754 (<literal>frontend_net</literal>);
755 others are sinks: they consume packages and return a result
756 (<literal>backend_test</literal>,
757 <literal>bounce</literal>,
758 <literal>http_file</literal>,
759 <literal>z3950_client</literal>);
760 the others are true filters, that read, process and pass on the
761 packages they are fed
762 (<literal>auth_simple</literal>,
763 <literal>log</literal>,
764 <literal>multi</literal>,
765 <literal>query_rewrite</literal>,
766 <literal>record_transform</literal>,
767 <literal>session_shared</literal>,
768 <literal>sru_z3950</literal>,
769 <literal>template</literal>,
770 <literal>virt_db</literal>).
775 <section id="overview.filter.types">
776 <title>Overview of filter types</title>
778 We now briefly consider each of the types of filter supported by
779 the core Metaproxy binary. This overview is intended to give a
780 flavor of the available functionality; more detailed information
781 about each type of filter is included below in
782 <xref linkend="reference"/>.
785 The filters are here named by the string that is used as the
786 <literal>type</literal> attribute of a
787 <literal><filter></literal> element in the configuration
788 file to request them, with the name of the class that implements
789 them in parentheses. (The classname is not needed for normal
790 configuration and use of Metaproxy; it is useful only to
794 The filters are here listed in alphabetical order:
801 New virt_db-alike that does inteligent peer choice, explain merging,
802 adds FD&N to explain. Keeps init responses (like "virt_db Classic"),
803 makes routing choices based on local explain knowledge. Ref IDDI
806 Filter to convert Explain Classic to ZeeRex.
808 CQL2PQF (which needs augmented ZeeRex) - MARC for Talis.
812 Figure out what additional information we need in:
813 ZeeRex (check against D3.1)
814 Init request (e.g. loop detection)
815 Query package (e.g. number of hops)
816 Query response (e.g. record source)
820 <section id="auth_simple">
821 <title><literal>auth_simple</literal>
822 (mp::filter::AuthSimple)</title>
824 Simple authentication and authorization. The configuration
825 specifies the name of a file that is the user register, which
826 lists <varname>username</varname>:<varname>password</varname>
827 pairs, one per line, colon separated. When a session begins, it
828 is rejected unless username and passsword are supplied, and match
829 a pair in the register. The configuration file may also specific
830 the name of another file that is the target register: this lists
831 lists <varname>username</varname>:<varname>dbname</varname>,<varname>dbname</varname>...
832 sets, one per line, with multiple database names separated by
833 commas. When a search is processed, it is rejected unless the
834 database to be searched is one of those listed as available to
839 <section id="backend_test">
840 <title><literal>backend_test</literal>
841 (mp::filter::Backend_test)</title>
843 A partial sink that provides dummy responses in the manner of the
844 <literal>yaz-ztest</literal> Z39.50 server. This is useful only
845 for testing. Seriously, you don't need this. Pretend you didn't
846 even read this section.
850 <section id="bounce">
851 <title><literal>bounce</literal>
852 (mp::filter::Bounce)</title>
854 A sink that swallows <emphasis>all packages</emphasis>,
855 and returns them almost unprocessed.
856 It never sends any package of any type further down the row, but
857 sets Z39.50 packages to Z_Close, and HTTP_Request packages to
858 HTTP_Response err code 400 packages, and adds a suitable bounce
860 The bounce filter is usually added at end of each filter chain route
861 to prevent infinite hanging of for example HTTP
862 requests packages when only the Z39.50 client partial sink
863 filter is found in the
868 <section id="cql_rpn">
869 <title><literal>cql_rpn</literal>
870 (mp::filter::CQLtoRPN)</title>
872 A query language transforming filter which catches Z39.50
873 <literal>searchRequest</literal>
874 packages containing <literal>CQL</literal> queries, transforms
875 those to <literal>RPN</literal> queries,
876 and sends the <literal>searchRequests</literal> on to the next
877 filters. It is among other things useful in a SRU context.
881 <section id="frontend_net">
882 <title><literal>frontend_net</literal>
883 (mp::filter::FrontendNet)</title>
885 A source that accepts Z39.50 connections from a port
886 specified in the configuration, reads protocol units, and
887 feeds them into the next filter in the route. When the result is
888 received, it is returned to the original origin.
892 <section id="http_file">
893 <title><literal>http_file</literal>
894 (mp::filter::HttpFile)</title>
896 A partial sink which swallows only
897 <literal>HTTP_Request</literal> packages, and
898 returns the contents of files from the local
899 filesystem in response to HTTP requests.
900 It lets Z39.50 packages and all other forthcoming package types
903 does mean that Metaproxy is also a Web-server in its spare time. So
904 far it does not contain either an email-reader or a Lisp
905 interpreter, but that day is surely coming.)
909 <section id="load_balance">
910 <title><literal>load_balance</literal>
911 (mp::filter::LoadBalance)</title>
913 Performs load balancing for incoming Z39.50 init requests.
914 It is used together with the <literal>virt_db</literal> filter,
915 but unlike the <literal>multi</literal> filter it does send an
916 entire session to only one of the virtual backends. The
917 <literal>load_balance</literal> filter is assuming that
918 all backend targets have equal content, and chooses the backend
919 with least load cost for a new session.
922 This filter is experimental and yet not mature for heavy load
930 <title><literal>log</literal>
931 (mp::filter::Log)</title>
933 Writes logging information to standard output, and passes on
934 the package unchanged. A log file name can be specified, as well
935 as multiple different logging formats.
940 <title><literal>multi</literal>
941 (mp::filter::Multi)</title>
943 Performs multi-database searching.
945 <link linkend="multidb">the extended discussion</link>
946 of virtual databases and multi-database searching below.
950 <section id="query_rewrite">
951 <title><literal>query_rewrite</literal>
952 (mp::filter::QueryRewrite)</title>
954 Rewrites Z39.50 <literal>Type-1</literal>
955 and <literal>Type-101</literal> (``<literal>RPN</literal>'')
957 three-step process: the query is transliterated from Z39.50
958 packet structures into an XML representation; that XML
959 representation is transformed by an XSLT stylesheet; and the
960 resulting XML is transliterated back into the Z39.50 packet
966 <section id="record_transform">
967 <title><literal>record_transform</literal>
968 (mp::filter::RecordTransform)</title>
970 This filter acts only on Z3950 present requests, and let all
971 other types of packages and requests pass untouched. It's use is
972 twofold: blocking Z3950 present requests, which the backend
973 server does not understand and can not honor, and transforming
974 the present syntax and elementset name according to the rules
975 specified, to fetch only existing record formats, and transform
976 them on the fly to requested record syntaxes.
980 <section id="session_shared">
981 <title><literal>session_shared</literal>
982 (mp::filter::SessionShared)</title>
984 This filter implements global sharing of
985 result sets (i.e. between threads and therefore between
986 clients), yielding performance improvements by clever resource
991 <section id="sru_z3950">
992 <title><literal>sru_z3950</literal>
993 (mp::filter::SRUtoZ3950)</title>
995 This filter transforms valid
996 SRU GET/POST/SOAP searchRetrieve requests to Z3950 init, search,
997 and present requests, and wraps the
998 received hit counts and XML records into suitable SRU response
1000 The <literal>sru_z3950</literal> filter processes also SRU
1001 GET/POST/SOAP explain requests, returning
1002 either the absolute minimum required by the standard, or a full
1003 pre-defined ZeeReX explain record.
1005 <ulink url="&url.zeerex.explain;">ZeeReX Explain</ulink>
1006 standard pages and the
1007 <ulink url="&url.sru.explain;">SRU Explain</ulink> pages
1008 for more information on the correct explain syntax.
1009 SRU scan requests are not supported yet.
1013 <section id="template">
1014 <title><literal>template</literal>
1015 (mp::filter::Template)</title>
1017 Does nothing at all, merely passing the packet on. (Maybe it
1018 should be called <literal>nop</literal> or
1019 <literal>passthrough</literal>?) This exists not to be used, but
1020 to be copied - to become the skeleton of new filters as they are
1021 written. As with <literal>backend_test</literal>, this is not
1022 intended for civilians.
1026 <section id="virt_db">
1027 <title><literal>virt_db</literal>
1028 (mp::filter::VirtualDB)</title>
1030 Performs virtual database selection: based on the name of the
1031 database in the search request, a server is selected, and its
1032 address added to the request in a <literal>VAL_PROXY</literal>
1033 otherInfo packet. It will subsequently be used by a
1034 <literal>z3950_client</literal> filter.
1036 <link linkend="multidb">the extended discussion</link>
1037 of virtual databases and multi-database searching below.
1041 <section id="z3950_client">
1042 <title><literal>z3950_client</literal>
1043 (mp::filter::Z3950Client)</title>
1045 A partial sink which swallows only Z39.50 packages.
1046 It performs Z39.50 searching and retrieval by proxying the
1047 packages that are passed to it. Init requests are sent to the
1048 address specified in the <literal>VAL_PROXY</literal> otherInfo
1049 attached to the request: this may have been specified by client,
1050 or generated by a <literal>virt_db</literal> filter earlier in
1051 the route. Subsequent requests are sent to the same address,
1052 which is remembered at Init time in a Session object.
1053 HTTP_Request packages and all other forthcoming package types
1054 are passed untouched.
1059 <section id="zeerex_explain">
1060 <title><literal>zeerex_explain</literal>
1061 (mp::filter::ZeerexExplain)</title>
1063 This filter acts as a sink for
1064 Z39.50 explain requests, returning a static ZeeReX
1065 Explain XML record from the config section. All other packages
1068 <ulink url="&url.zeerex.explain;">ZeeReX Explain</ulink>
1070 for more information on the correct explain syntax.
1074 This filter is not yet completed.
1083 <section id="future.directions">
1084 <title>Future directions</title>
1086 Some other filters that do not yet exist, but which would be
1087 useful, are briefly described. These may be added in future
1088 releases (or may be created by third parties, as loadable
1094 <term><literal>frontend_cli</literal> (source)</term>
1097 Command-line interface for generating requests.
1102 <term><literal>sru_client</literal> (sink)</term>
1105 SRU/GET and SRU/SOAP searching and retrieval.
1110 <term><literal>opensearch_client</literal> (sink)</term>
1113 A9 OpenSearch searching and retrieval.
1123 <chapter id="configuration">
1124 <title>Configuration: the Metaproxy configuration file format</title>
1127 <section id="configuration-introductory-notes">
1128 <title>Introductory notes</title>
1130 If Metaproxy is an interpreter providing operations on packages, then
1131 its configuration file can be thought of as a program for that
1132 interpreter. Configuration is by means of a single XML file, the name
1133 of which is supplied as the sole command-line argument to the
1134 <command>metaproxy</command> program. (See
1135 <xref linkend="reference"/> below for more information on invoking
1140 <section id="overview.xml.structure">
1141 <title>Overview of the config file XML structure</title>
1143 All elements and attributes are in the namespace
1144 <ulink url="http://indexdata.com/metaproxy"/>.
1145 This is most easily achieved by setting the default namespace on
1146 the top-level element, as here:
1149 <metaproxy xmlns="http://indexdata.com/metaproxy" version="1.0">
1152 The top-level element is <metaproxy>. This contains
1153 a <dlpath> element,
1154 a <start> element,
1155 a <filters> element and
1156 a <routes> element, in that order. <dlpath> and
1157 <filters> are optional; the other two are mandatory.
1158 All four are non-repeatable.
1161 The <dlpath;> element contains a text element which
1162 specifies the location of filter modules. This is only needed
1163 if Metaproxy must load 3rd party filters (most filters with Metaproxy
1164 are built into the Metaproxy application).
1167 The <start> element is empty, but carries a
1168 <literal>route</literal> attribute, whose value is the name of
1169 route at which to start running - analogous to the name of the
1170 start production in a formal grammar.
1173 If present, <filters> contains zero or more <filter>
1174 elements. Each filter carries a <literal>type</literal> attribute
1175 which specifies what kind of filter is being defined
1176 (<literal>frontend_net</literal>, <literal>log</literal>, etc.)
1177 and contain various elements that provide suitable configuration
1178 for a filter of its type. The filter-specific elements are
1180 <xref linkend="reference"/>.
1181 Filters defined in this part of the file must carry an
1182 <literal>id</literal> attribute so that they can be referenced
1186 <routes> contains one or more <route> elements, each
1187 of which must carry an <literal>id</literal> element. One of the
1188 routes must have the ID value that was specified as the start
1189 route in the <start> element's <literal>route</literal>
1190 attribute. Each route contains zero or more <filter>
1191 elements. These are of two types. They may be empty, but carry a
1192 <literal>refid</literal> attribute whose value is the same as the
1193 <literal>id</literal> of a filter previously defined in the
1194 <filters> section. Alternatively, a route within a filter
1195 may omit the <literal>refid</literal> attribute, but contain
1196 configuration elements similar to those used for filters defined
1197 in the <filters> section. (In other words, each filter in a
1198 route may be included either by reference or by physical
1204 <section id="example.configuration">
1205 <title>An example configuration</title>
1207 The following is a small, but complete, Metaproxy configuration
1208 file (included in the distribution as
1209 <literal>metaproxy/etc/config1.xml</literal>).
1210 This file defines a very simple configuration that simply proxies
1211 to whatever back-end server the client requests, but logs each
1212 request and response. This can be useful for debugging complex
1213 client-server dialogues.
1215 <screen><![CDATA[<?xml version="1.0"?>
1216 <metaproxy xmlns="http://indexdata.com/metaproxy" version="1.0">
1217 <dlpath>/usr/lib/metaproxy/modules</dlpath>
1218 <start route="start"/>
1220 <filter id="frontend" type="frontend_net">
1223 <filter id="backend" type="z3950_client">
1228 <filter refid="frontend"/>
1229 <filter type="log"/>
1230 <filter refid="backend"/>
1231 <filter type="bounce"/>
1237 It works by defining a single route, called
1238 <literal>start</literal>, which consists of a sequence of four
1239 filters. The first and last of these are included by reference:
1240 their <literal><filter></literal> elements have
1241 <literal>refid</literal> attributes that refer to filters defined
1242 within the prior <literal><filters></literal> section. The
1243 middle filter is included inline in the route.
1246 The four filters in the route are as follows: first, a
1247 <literal>frontend_net</literal> filter accepts Z39.50 requests
1248 from any host on port 9000; then these requests are passed through
1249 a <literal>log</literal> filter that emits a message for each
1250 request; they are then fed into a <literal>z3950_client</literal>
1251 filter, which forwards all Z39.50 requests to the client-specified
1252 back-end Z39.509 server. Those Z39.50 packages are returned by the
1253 <literal>z3950_client</literal> filter, with the response data
1254 filled by the external Z39.50 server targeted.
1255 All non-Z39.50 packages are passed through to the
1256 <literal>bounce</literal> filter, which definitely bounces
1257 everything, including fish, bananas, cold pyjamas,
1258 mutton, beef and trout packages.
1259 When the response arrives, it is handed
1260 back to the <literal>log</literal> filter, which emits another
1261 message; and then to the <literal>frontend_net</literal> filter,
1262 which returns the response to the client.
1266 <section id="config-file-modularity">
1267 <title>Config file modularity</title>
1269 Metaproxy XML configuration snippets can be reused by other
1270 filters using the <literal>XInclude</literal> standard, as seen in
1271 the <literal>/etc/config-sru-to-z3950.xml</literal> example SRU
1274 <filter id="sru" type="sru_z3950">
1275 <database name="Default">
1276 <xi:include xmlns:xi="http://www.w3.org/2001/XInclude"
1277 href="explain.xml"/>
1284 <section id="config-file-syntax-check">
1285 <title>Config file syntax checking</title>
1287 The distribution contains RelaxNG Compact and XML syntax checking
1288 files, as well as XML Schema files. These are found in the
1291 xml/schema/metaproxy.rnc
1292 xml/schema/metaproxy.rng
1293 xml/schema/metaproxy.xsd
1295 and can be used to verify or debug the XML structure of
1296 configuration files. For example, using the utility
1297 <filename>xmllint</filename>, syntax checking is done like this:
1299 xmllint --noout --schema xml/schema/metaproxy.xsd etc/config-local.xml
1300 xmllint --noout --relaxng xml/schema/metaproxy.rng etc/config-local.xml
1302 (A recent version of <literal>libxml2</literal> is required, as
1303 support for XML Schemas is a relatively recent addition.)
1306 You can of course use any other RelaxNG or XML Schema compliant tool
1314 <chapter id="multidb">
1315 <title>Virtual databases and multi-database searching</title>
1318 <section id="multidb-introductory-notes">
1319 <title>Introductory notes</title>
1321 Two of Metaproxy's filters are concerned with multiple-database
1322 operations. Of these, <literal>virt_db</literal> can work alone
1323 to control the routing of searches to one of a number of servers,
1324 while <literal>multi</literal> can work together with
1325 <literal>virt_db</literal> to perform multi-database searching, merging
1326 the results into a unified result-set - ``metasearch in a box''.
1329 The interaction between
1330 these two filters is necessarily complex: it reflects the real,
1331 irreducible complexity of multi-database searching in a protocol such
1332 as Z39.50 that separates initialization from searching, and in
1333 which the database to be searched is not known at initialization
1337 It's possible to use these filters without understanding the
1338 details of their functioning and the interaction between them; the
1339 next two sections of this chapter are ``HOW-TO'' guides for doing
1340 just that. However, debugging complex configurations will require
1341 a deeper understanding, which the last two sections of this
1342 chapters attempt to provide.
1347 <section id="multidb.virt_db">
1348 <title>Virtual databases with the <literal>virt_db</literal> filter</title>
1350 Working alone, the purpose of the
1351 <literal>virt_db</literal>
1352 filter is to route search requests to one of a selection of
1353 back-end databases. In this way, a single Z39.50 endpoint
1354 (running Metaproxy) can provide access to several different
1355 underlying services, including those that would otherwise be
1356 inaccessible due to firewalls. In many useful configurations, the
1357 back-end databases are local to the Metaproxy installation, but
1358 the software does not enforce this, and any valid Z39.50 servers
1359 may be used as back-ends.
1362 For example, a <literal>virt_db</literal>
1363 filter could be set up so that searches in the virtual database
1364 ``lc'' are forwarded to the Library of Congress bibliographic
1365 catalogue server, and searches in the virtual database ``marc''
1366 are forwarded to the toy database of MARC records that Index Data
1367 hosts for testing purposes. A <literal>virt_db</literal>
1368 configuration to make this switch would look like this:
1370 <screen><![CDATA[<filter type="virt_db">
1372 <database>lc</database>
1373 <target>z3950.loc.gov:7090/voyager</target>
1376 <database>marc</database>
1377 <target>indexdata.com/marc</target>
1379 </filter>]]></screen>
1381 As well as being useful in it own right, this filter also provides
1382 the foundation for multi-database searching.
1387 <section id="multidb.multi">
1388 <title>Multi-database search with the <literal>multi</literal> filter</title>
1390 To arrange for Metaproxy to broadcast searches to multiple back-end
1391 servers, the configuration needs to include two components: a
1392 <literal>virt_db</literal>
1393 filter that specifies multiple
1394 <literal><target></literal>
1395 elements, and a subsequent
1396 <literal>multi</literal>
1397 filter. Here, for example, is a complete configuration that
1398 broadcasts searches to both the Library of Congress catalogue and
1399 Index Data's tiny testing database of MARC records:
1401 <screen><![CDATA[<?xml version="1.0"?>
1402 <metaproxy xmlns="http://indexdata.com/metaproxy" version="1.0">
1403 <start route="start"/>
1406 <filter type="frontend_net">
1407 <threads>10</threads>
1410 <filter type="virt_db">
1412 <database>lc</database>
1413 <target>z3950.loc.gov:7090/voyager</target>
1416 <database>marc</database>
1417 <target>indexdata.com/marc</target>
1420 <database>all</database>
1421 <target>z3950.loc.gov:7090/voyager</target>
1422 <target>indexdata.com/marc</target>
1425 <filter type="multi"/>
1426 <filter type="z3950_client">
1427 <timeout>30</timeout>
1429 <filter type="bounce"/>
1432 </metaproxy>]]></screen>
1435 <literal>virt_db</literal>
1436 filter that specifies multiple
1437 <literal><target></literal>
1438 elements but without a subsequent
1439 <literal>multi</literal>
1440 filter yields surprising and undesirable results, as will be
1441 described below. Don't do that.)
1444 Metaproxy can be invoked with this configuration as follows:
1446 <screen>../src/metaproxy --config config-simple-multi.xml</screen>
1448 And thereafter, Z39.50 clients can connect to the running server
1449 (on port 9000, as specified in the configuration) and search in
1450 any of the databases
1451 <literal>lc</literal> (the Library of Congress catalogue),
1452 <literal>marc</literal> (Index Data's test database of MARC records)
1454 <literal>all</literal> (both of these). As an example, a session
1455 using the YAZ command-line client <literal>yaz-client</literal> is
1456 here included (edited for brevity and clarity):
1458 <screen><![CDATA[$ yaz-client @:9000
1462 Search was a success.
1463 Number of hits: 10000, setno 1
1467 Search was a success.
1468 Number of hits: 10, setno 3
1472 Search was a success.
1473 Number of hits: 10010, setno 4
1476 [marc]Record type: USmarc
1479 005 00000000000000.0
1480 008 910710c19910701nju 00010 eng
1484 100 10 $a Jack Collins
1485 245 10 $a How to program a computer
1491 [VOYAGER]Record type: USmarc
1493 005 20041229102447.0
1494 008 030910s2004 caua 000 0 eng
1495 035 $a (DLC) 2003112666
1496 906 $a 7 $b cbc $c orignew $d 4 $e epcn $f 20 $g y-gencatlg
1497 925 0 $a acquire $b 1 shelf copy $x policy default
1498 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.
1501 040 $a DLC $c DLC $d DLC
1502 050 00 $a MLCM 2004/03312 (G)
1503 245 10 $a 007, everything or nothing : $b Prima's official strategy guide / $c created by Kaizen Media Group.
1504 246 3 $a Double-O-seven, everything or nothing
1505 246 30 $a Prima's official strategy guide
1506 260 $a Roseville, CA : $b Prima Games, $c c2004.
1507 300 $a 161 p. : $b col. ill. ; $c 28 cm.
1508 500 $a "Platforms: Nintendo GameCube, Macintosh, PC, PlayStation 2 computer entertainment system, Xbox"--P. [4] of cover.
1509 650 0 $a Video games.
1510 710 2 $a Kaizen Media Group.
1511 856 42 $3 Publisher description $u http://www.loc.gov/catdir/description/random052/2003112666.html
1516 As can be seen, the first record in the result set is from the
1517 Index Data test database, and the second from the Library of
1518 Congress database. The result-set continues alternating records
1519 round-robin style until the point where one of the databases'
1520 records are exhausted.
1523 This example uses only two back-end databases; more may be used.
1524 There is no limitation imposed on the number of databases that may
1525 be metasearched in this way: issues of resource usage and
1526 administrative complexity dictate the practical limits.
1529 What happens when one of the databases doesn't respond? By default,
1530 the entire multi-database search fails, and the appropriate
1531 diagnostic is returned to the client. This is usually appropriate
1532 during development, when technicians need maximum information, but
1533 can be inconvenient in deployment, when users typically don't want
1534 to be bothered with problems of this kind and prefer just to get
1535 the records from the databases that are available. To obtain this
1536 latter behavior add an empty
1537 <literal><hideunavailable></literal>
1539 <literal>multi</literal> filter:
1541 <screen><![CDATA[ <filter type="multi">
1543 </filter>]]></screen>
1545 Under this regime, an error is reported to the client only if
1546 <emphasis>all</emphasis> the databases in a multi-database search
1552 <section id="multidb.what">
1553 <title>What's going on?</title>
1555 <title>Lark's vomit</title>
1557 This section goes into a level of technical detail that is
1558 probably not necessary in order to configure and use Metaproxy.
1559 It is provided only for those who like to know how things work.
1560 You should feel free to skip on to the next section if this one
1561 doesn't seem like fun.
1565 Hold on tight - this may get a little hairy.
1568 In the general course of things, a Z39.50 Init request may carry
1569 with it an otherInfo packet of type <literal>VAL_PROXY</literal>,
1570 whose value indicates the address of a Z39.50 server to which the
1571 ultimate connection is to be made. (This otherInfo packet is
1572 supported by YAZ-based Z39.50 clients and servers, but has not yet
1573 been ratified by the Maintenance Agency and so is not widely used
1574 in non-Index Data software. We're working on it.)
1575 The <literal>VAL_PROXY</literal> packet functions
1576 analogously to the absoluteURI-style Request-URI used with the GET
1577 method when a web browser asks a proxy to forward its request: see
1579 <ulink url="http://www.w3.org/Protocols/rfc2616/rfc2616-sec5.html#sec5.1.2"
1580 >Request-URI</ulink>
1582 <ulink url="http://www.w3.org/Protocols/rfc2616/rfc2616.html"
1583 >the HTTP 1.1 specification</ulink>.
1586 Within Metaproxy, Search requests that are part of the same
1587 session as an Init request that carries a
1588 <literal>VAL_PROXY</literal> otherInfo are also annotated with the
1589 same information. The role of the <literal>virt_db</literal>
1590 filter is to rewrite this otherInfo packet dependent on the
1591 virtual database that the client wants to search.
1594 When Metaproxy receives a Z39.50 Init request from a client, it
1595 doesn't immediately forward that request to the back-end server.
1596 Why not? Because it doesn't know <emphasis>which</emphasis>
1597 back-end server to forward it to until the client sends a Search
1598 request that specifies the database that it wants to search in.
1599 Instead, it just treasures the Init request up in its heart; and,
1600 later, the first time the client does a search on one of the
1601 specified virtual databases, a connection is forged to the
1602 appropriate server and the Init request is forwarded to it. If,
1603 later in the session, the same client searches in a different
1604 virtual database, then a connection is forged to the server that
1605 hosts it, and the same cached Init request is forwarded there,
1609 All of this clever Init-delaying is done by the
1610 <literal>frontend_net</literal> filter. The
1611 <literal>virt_db</literal> filter knows nothing about it; in
1612 fact, because the Init request that is received from the client
1613 doesn't get forwarded until a Search request is received, the
1614 <literal>virt_db</literal> filter (and the
1615 <literal>z3950_client</literal> filter behind it) doesn't even get
1616 invoked at Init time. The <emphasis>only</emphasis> thing that a
1617 <literal>virt_db</literal> filter ever does is rewrite the
1618 <literal>VAL_PROXY</literal> otherInfo in the requests that pass
1622 It is possible for a <literal>virt_db</literal> filter to contain
1624 <literal><target></literal>
1625 elements. What does this mean? Only that the filter will add
1626 multiple <literal>VAL_PROXY</literal> otherInfo packets to the
1627 Search requests that pass through it. That's because the virtual
1628 DB filter is dumb, and does exactly what it's told - no more, no
1630 If a Search request with multiple <literal>VAL_PROXY</literal>
1631 otherInfo packets reaches a <literal>z3950_client</literal>
1632 filter, this is an error. That filter doesn't know how to deal
1633 with multiple targets, so it will either just pick one and search
1634 in it, or (better) fail with an error message.
1637 The <literal>multi</literal> filter comes to the rescue! This is
1638 the only filter that knows how to deal with multiple
1639 <literal>VAL_PROXY</literal> otherInfo packets, and it does so by
1640 making multiple copies of the entire Search request: one for each
1641 <literal>VAL_PROXY</literal>. Each of these new copies is then
1642 passed down through the remaining filters in the route. (The
1643 copies are handled in parallel though the
1644 spawning of new threads.) Since the copies each have only one
1645 <literal>VAL_PROXY</literal> otherInfo, they can be handled by the
1646 <literal>z3950_client</literal> filter, which happily deals with
1647 each one individually. When the results of the individual
1648 searches come back up to the <literal>multi</literal> filter, it
1649 merges them into a single Search response, which is what
1650 eventually makes it back to the client.
1655 <imagedata fileref="multi.pdf" format="PDF" scale="50"/>
1658 <imagedata fileref="multi.png" format="PNG"/>
1661 <!-- Fall back if none of the images can be used -->
1663 [Here there should be a diagram showing the progress of
1664 packages through the filters during a simple virtual-database
1665 search and a multi-database search, but is seems that your
1666 tool chain has not been able to include the diagram in this
1671 <para>A picture is worth a thousand words (but only five hundred on 64-bit architectures)</para>
1678 <chapter id="sru-server">
1679 <title>Combined SRU webservice and Z39.50 server configuration</title>
1681 Metaproxy can act as
1682 <ulink url="&url.sru;">SRU</ulink> and
1683 web service server, which translates web service requests to
1684 <ulink url="&url.z39.50;">ANSI/NISO Z39.50</ulink> packages and
1685 sends them off to common available targets.
1688 A typical setup for this operation needs a filter route including the
1692 <table id="sru-server-table-config" frame="top">
1693 <title>SRU/Z39.50 Server Filter Route Configuration</title>
1697 <entry>Filter</entry>
1698 <entry>Importance</entry>
1699 <entry>Purpose</entry>
1705 <entry><literal>frontend_net</literal></entry>
1706 <entry>required</entry>
1707 <entry>Accepting HTTP connections and passing them to following
1708 filters. Since this filter also accepts Z39.50 connections, the
1709 server works as SRU and Z39.50 server on the same port.</entry>
1712 <entry><literal>sru_z3950</literal></entry>
1713 <entry>required</entry>
1714 <entry>Accepting SRU GET/POST/SOAP explain and
1715 searchRetrieve requests for the the configured databases.
1716 Explain requests are directly served from the static XML configuration.
1717 SearchRetrieve requests are
1718 transformed to Z39.50 search and present packages.
1719 All other HTTP and Z39.50 packages are passed unaltered.</entry>
1722 <entry><literal>http_file</literal></entry>
1723 <entry>optional</entry>
1724 <entry>Serving HTTP requests from the filesystem. This is only
1725 needed if the server should serve XSLT stylesheets, static HTML
1726 files or Java Script for thin browser based clients.
1727 Z39.50 packages are passed unaltered.</entry>
1730 <entry><literal>cql_rpn</literal></entry>
1731 <entry>required</entry>
1732 <entry>Usually, Z39.50 servers do not talk CQL, hence the
1733 translation of the CQL query language to RPN is mandatory in
1734 most cases. Affects only Z39.50 search packages.</entry>
1737 <entry><literal>record_transform</literal></entry>
1738 <entry>optional</entry>
1739 <entry>Some Z39.50 backend targets can not present XML record
1740 syntaxes in common wanted element sets. using this filter, one
1741 can transform binary MARC records to MARCXML records, and
1742 further transform those to any needed XML schema/format by XSLT
1743 transformations. Changes only Z39.50 present packages.</entry>
1746 <entry><literal>session_shared</literal></entry>
1747 <entry>optional</entry>
1748 <entry>The stateless nature of web services requires frequent
1749 re-searching of the same targets for display of paged result set
1750 records. This might be an unacceptable burden for the accessed
1751 backend Z39.50 targets, and this mosule can be added for
1752 efficient backend target resource pooling.</entry>
1755 <entry><literal>z3950_client</literal></entry>
1756 <entry>required</entry>
1757 <entry>Finally, a Z39.50 package sink is needed in the filter
1758 chain to provide the response packages. The Z39.50 client module
1759 is used to access external targets over the network, but any
1760 coming local Z39.50 package sink could be used instead of.</entry>
1763 <entry><literal>bounce</literal></entry>
1764 <entry>required</entry>
1765 <entry>Any Metaproxy package arriving here did not do so by
1766 purpose, and is bounced back with connection closure. this
1767 prevents inifinite package hanging inside the SRU server.</entry>
1773 A typical minimal example <ulink url="&url.sru;">SRU</ulink>
1774 server configuration file is found in the tarball distribution at
1775 <literal>etc/config-sru-to-z3950.xml</literal>.
1778 Off course, any other metaproxy modules can be integrated into a
1779 SRU server solution, including, but not limited to, load balancing,
1780 multiple target querying
1781 (see <xref linkend="multidb"/>), and complex RPN query rewrites.
1788 <chapter id="extensions">
1789 <title>Writing extensions for Metaproxy</title>
1790 <para>### To be written</para>
1796 <chapter id="classes">
1797 <title>Classes in the Metaproxy source code</title>
1800 <section id="classes-introductory-notes">
1801 <title>Introductory notes</title>
1803 <emphasis>Stop! Do not read this!</emphasis>
1804 You won't enjoy it at all. You should just skip ahead to
1805 <xref linkend="reference"/>,
1807 <!-- The remainder of this paragraph is lifted verbatim from
1808 Douglas Adams' _Hitch Hiker's Guide to the Galaxy_, chapter 8 -->
1809 you things you really need to know, like the fact that the
1810 fabulously beautiful planet Bethselamin is now so worried about
1811 the cumulative erosion by ten billion visiting tourists a year
1812 that any net imbalance between the amount you eat and the amount
1813 you excrete whilst on the planet is surgically removed from your
1814 bodyweight when you leave: so every time you go to the lavatory it
1815 is vitally important to get a receipt.
1818 This chapter contains documentation of the Metaproxy source code, and is
1819 of interest only to maintainers and developers. If you need to
1820 change Metaproxy's behavior or write a new filter, then you will most
1821 likely find this chapter helpful. Otherwise it's a waste of your
1822 good time. Seriously: go and watch a film or something.
1823 <citetitle>This is Spinal Tap</citetitle> is particularly good.
1826 Still here? OK, let's continue.
1829 In general, classes seem to be named big-endianly, so that
1830 <literal>FactoryFilter</literal> is not a filter that filters
1831 factories, but a factory that produces filters; and
1832 <literal>FactoryStatic</literal> is a factory for the statically
1833 registered filters (as opposed to those that are dynamically
1838 <section id="individual.classes">
1839 <title>Individual classes</title>
1841 The classes making up the Metaproxy application are here listed by
1842 class-name, with the names of the source files that define them in
1846 <section id="class-FactoryFilter">
1847 <title><literal>mp::FactoryFilter</literal>
1848 (<filename>factory_filter.cpp</filename>)</title>
1850 A factory class that exists primarily to provide the
1851 <literal>create()</literal> method, which takes the name of a
1852 filter class as its argument and returns a new filter of that
1853 type. To enable this, the factory must first be populated by
1854 calling <literal>add_creator()</literal> for static filters (this
1855 is done by the <literal>FactoryStatic</literal> class, see below)
1856 and <literal>add_creator_dyn()</literal> for filters loaded
1861 <section id="class-FactoryStatic">
1862 <title><literal>mp::FactoryStatic</literal>
1863 (<filename>factory_static.cpp</filename>)</title>
1865 A subclass of <literal>FactoryFilter</literal> which is
1866 responsible for registering all the statically defined filter
1867 types. It does this by knowing about all those filters'
1868 structures, which are listed in its constructor. Merely
1869 instantiating this class registers all the static classes. It is
1870 for the benefit of this class that <literal>struct
1871 metaproxy_1_filter_struct</literal> exists, and that all the filter
1872 classes provide a static object of that type.
1876 <section id="class-filter-Base">
1877 <title><literal>mp::filter::Base</literal>
1878 (<filename>filter.cpp</filename>)</title>
1880 The virtual base class of all filters. The filter API is, on the
1881 surface at least, extremely simple: two methods.
1882 <literal>configure()</literal> is passed an XML DOM tree representing
1883 that part of the configuration file that pertains to this filter
1884 instance, and is expected to walk that tree extracting relevant
1885 information. And <literal>process()</literal> processes a
1886 package (see below). That surface simplicity is a bit
1887 misleading, as <literal>process()</literal> needs to know a lot
1888 about the <literal>Package</literal> class in order to do
1893 <section id="class-AuthSimple">
1894 <title><literal>mp::filter::AuthSimple</literal>,
1895 <literal>Backend_test</literal>, etc.
1896 (<filename>filter_auth_simple.cpp</filename>,
1897 <filename>filter_backend_test.cpp</filename>, etc.)</title>
1899 Individual filters. Each of these is implemented by a header and
1900 a source file, named <filename>filter_*.hpp</filename> and
1901 <filename>filter_*.cpp</filename> respectively. All the header
1902 files should be pretty much identical, in that they declare the
1903 class, including a private <literal>Rep</literal> class and a
1904 member pointer to it, and the two public methods.
1907 The source file for each filter needs to supply:
1912 A definition of the private <literal>Rep</literal> class.
1917 Some boilerplate constructors and destructors.
1922 A <literal>configure()</literal> method that uses the
1923 appropriate XML fragment.
1928 Most important, the <literal>process()</literal> method that
1929 does all the actual work.
1935 <section id="class-Package">
1936 <title><literal>mp::Package</literal>
1937 (<filename>package.cpp</filename>)</title>
1939 Represents a package on its way through the series of filters
1940 that make up a route. This is essentially a Z39.50 or SRU APDU
1941 together with information about where it came from, which is
1942 modified as it passes through the various filters.
1946 <section id="class-Pipe">
1947 <title><literal>mp::Pipe</literal>
1948 (<filename>pipe.cpp</filename>)</title>
1950 This class provides a compatibility layer so that we have an IPC
1951 mechanism that works the same under Unix and Windows. It's not
1952 particularly exciting.
1956 <section id="class-RouterChain">
1957 <title><literal>mp::RouterChain</literal>
1958 (<filename>router_chain.cpp</filename>)</title>
1964 <section id="class-RouterFleXML">
1965 <title><literal>mp::RouterFleXML</literal>
1966 (<filename>router_flexml.cpp</filename>)</title>
1972 <section id="class-Session">
1973 <title><literal>mp::Session</literal>
1974 (<filename>session.cpp</filename>)</title>
1980 <section id="class-ThreadPoolSocketObserver">
1981 <title><literal>mp::ThreadPoolSocketObserver</literal>
1982 (<filename>thread_pool_observer.cpp</filename>)</title>
1988 <section id="class-util">
1989 <title><literal>mp::util</literal>
1990 (<filename>util.cpp</filename>)</title>
1992 A namespace of various small utility functions and classes,
1993 collected together for convenience. Most importantly, includes
1994 the <literal>mp::util::odr</literal> class, a wrapper for YAZ's
1999 <section id="class-xml">
2000 <title><literal>mp::xml</literal>
2001 (<filename>xmlutil.cpp</filename>)</title>
2003 A namespace of various XML utility functions and classes,
2004 collected together for convenience.
2010 <section id="other.source.files">
2011 <title>Other Source Files</title>
2013 In addition to the Metaproxy source files that define the classes
2014 described above, there are a few additional files which are
2015 briefly described here:
2019 <term><literal>metaproxy_prog.cpp</literal></term>
2022 The main function of the <command>metaproxy</command> program.
2027 <term><literal>ex_router_flexml.cpp</literal></term>
2030 Identical to <literal>metaproxy_prog.cpp</literal>: it's not clear why.
2035 <term><literal>test_*.cpp</literal></term>
2038 Unit-tests for various modules.
2044 ### Still to be described:
2045 <literal>ex_filter_frontend_net.cpp</literal>,
2046 <literal>filter_dl.cpp</literal>,
2047 <literal>plainfile.cpp</literal>,
2048 <literal>tstdl.cpp</literal>.
2054 <reference id="reference">
2055 <title>Reference</title>
2056 <partintro id="reference-introduction">
2058 The material in this chapter is drawn directly from the individual
2059 manual entries. In particular, the Metaproxy invocation section is
2060 available using <command>man metaproxy</command>, and the section
2061 on each individual filter is available using the name of the filter
2062 as the argument to the <command>man</command> command.
2068 <appendix id="license">
2069 <title>License</title>
2074 Metaproxy is free software; you can redistribute it and/or modify it under
2075 the terms of the GNU General Public License as published by the Free
2076 Software Foundation; either version 2, or (at your option) any later
2081 Metaproxy is distributed in the hope that it will be useful, but WITHOUT ANY
2082 WARRANTY; without even the implied warranty of MERCHANTABILITY or
2083 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
2088 You should have received a copy of the GNU General Public License
2089 along with Metaproxy; see the file LICENSE. If not, write to the
2090 Free Software Foundation,
2091 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
2099 <!-- Keep this comment at the end of the file
2104 sgml-minimize-attributes:nil
2105 sgml-always-quote-attributes:t
2108 sgml-parent-document: nil
2109 sgml-local-catalogs: nil
2110 sgml-namecase-general:t