Xaira is an XML Aware Indexing and Retrieval Application designed to
allow fast access to large corpora marked up in XML. More
information is available from the project website at http://www.xaira.org.

Xaira is distributed in source form for Unix, and as a pre-built
Windows installer file. The version number forms part of the name of
each download, so that for example the Unix source release for 1.17 is called
xaira-1.17.tar.gz. The version number may also be found on the Xaira
Windows client splash screen.

Packaged releases are made available from http://xaira.sourceforge.net
where a CVS repository including all source code (including the
Windows-only components) is also maintained.

When unpacked, the main distribution has the following sub-directories:

indexer: The standalone index program.  Type indexer -? for a summary of parameters.

server: The server shared library. This is the library that implements
the Xaira API, as documented in the html files in docs.

oldserver: This is the version of the server that implements the old
sara protocol. It will be withdrawn when testing of the new server is
complete.

xaira_proxy: This is another shared library that implements the same
protocol by making xmlrpc calls to a remote host.

xaira_daemon: This is the server process run on a remote host to serve
the proxy library.

samples: This directory contains sample applications in source.

xaira_test: This is a simple application to check that the built
indexer and server are working.

xaira_php: This is a PHP extension that makes the Xaira API available
within PHP 5.00 and above.

docs: This is the Xaira API documentation.

java: This is a set of java classes that work in exactly the same way
as xaira-proxy, thus allowing a Java application to interact with a
Xaira server.

test_corpus: This is the data used by the xaira_test program.

See the file INSTALL for installation instructions.

Xaira is made available under the GNU public licence. Details may be
found in the file COPYING. 

Xaira was developed at Oxford University's Research Technologies
Service, with funding from the British National Corpus project, and
from the Andrew W Mellon Foundation.



