I'm overjoyed to be able to announce the release of Xapian 1.0.0. This release features Unicode support, updated stemming algorithms (with new languages: Hungarian, Romanian, and Turkish), a new and improved indexing strategy, improvements to the remote backend (which now supports all backend features and works on MS Windows), new QueryParser features (ranges, pure NOT queries, support for incremental searches), and major enhancements to the Python bindings. Flint supports using zlib to compress tags in the record and termlist trees, and is now the default backend. The documentation has also been improved. For a broader overview see: http://wiki.xapian.org/ReleaseOverview/1.0.0 The full list of user-visible changes, can be seen via the "[news]" links on the download page: http://www.xapian.org/download.php If you're using Xapian 0.9.x, you may need to plan your migration to 1.0.x. The classes which care about encoding (Xapian::Stem, Xapian::QueryParser and the new Xapian::TermGenerator) now expect UTF-8, and Omega and its indexers now store everything in UTF-8 internally. Also, a few features have been marked as deprecated, and features marked as deprecated since 0.9.0 or earlier have been removed. As always, if you encounter problems, please report them here, or to the bugtracker. And a big thankyou to everyone who's help make this release a reality! Cheers, Olly
Congratulations and thanks a lot to everyone involved! JL Olly Betts wrote:> I'm overjoyed to be able to announce the release of Xapian 1.0.0. > > This release features Unicode support, updated stemming algorithms > (with new languages: Hungarian, Romanian, and Turkish), a new and > improved indexing strategy, improvements to the remote backend (which > now supports all backend features and works on MS Windows), new > QueryParser features (ranges, pure NOT queries, support for incremental > searches), and major enhancements to the Python bindings. > > Flint supports using zlib to compress tags in the record and termlist > trees, and is now the default backend. The documentation has also been > improved. > > For a broader overview see: > > http://wiki.xapian.org/ReleaseOverview/1.0.0 > > The full list of user-visible changes, can be seen via the "[news]" > links on the download page: > > http://www.xapian.org/download.php > > If you're using Xapian 0.9.x, you may need to plan your migration to > 1.0.x. The classes which care about encoding (Xapian::Stem, > Xapian::QueryParser and the new Xapian::TermGenerator) now expect UTF-8, > and Omega and its indexers now store everything in UTF-8 internally. > Also, a few features have been marked as deprecated, and features marked > as deprecated since 0.9.0 or earlier have been removed. > > As always, if you encounter problems, please report them here, or to > the bugtracker. > > And a big thankyou to everyone who's help make this release a reality! > > Cheers, > Olly > > _______________________________________________ > Xapian-discuss mailing list > Xapian-discuss@lists.xapian.org > http://lists.xapian.org/mailman/listinfo/xapian-discuss >
Congratulations to Olly and all the great contributors. Xapian is truly a marvelous project and it reflects the hard work you guys have put into it. Regards, Andrew Olly Betts wrote:> I'm overjoyed to be able to announce the release of Xapian 1.0.0.
Nice, But I'm not sure if I understand the necessary changes needed for the change to UTF-8. As you know we generate a set of datafiles (if any encoding applies its probably ISO-8859-15) for scriptindex and afterwards omega to search through the data. Is there any way to let scriptindex know we're feeding it ISO-8859-15 rather than UTF-8? The UTF-8 support in normal php installations isn't very good. And is it also possible to let Omega know we are feeding it ISO-8859-15 and want that returned as well? Or are we required to supply those commands with UTF-8 data? Best regards, Arjen On 18-5-2007 2:59 Olly Betts wrote:> I'm overjoyed to be able to announce the release of Xapian 1.0.0. > > This release features Unicode support, updated stemming algorithms > (with new languages: Hungarian, Romanian, and Turkish), a new and > improved indexing strategy, improvements to the remote backend (which > now supports all backend features and works on MS Windows), new > QueryParser features (ranges, pure NOT queries, support for incremental > searches), and major enhancements to the Python bindings. > > Flint supports using zlib to compress tags in the record and termlist > trees, and is now the default backend. The documentation has also been > improved. > > For a broader overview see: > > http://wiki.xapian.org/ReleaseOverview/1.0.0 > > The full list of user-visible changes, can be seen via the "[news]" > links on the download page: > > http://www.xapian.org/download.php > > If you're using Xapian 0.9.x, you may need to plan your migration to > 1.0.x. The classes which care about encoding (Xapian::Stem, > Xapian::QueryParser and the new Xapian::TermGenerator) now expect UTF-8, > and Omega and its indexers now store everything in UTF-8 internally. > Also, a few features have been marked as deprecated, and features marked > as deprecated since 0.9.0 or earlier have been removed. > > As always, if you encounter problems, please report them here, or to > the bugtracker. > > And a big thankyou to everyone who's help make this release a reality! > > Cheers, > Olly > > _______________________________________________ > Xapian-discuss mailing list > Xapian-discuss@lists.xapian.org > http://lists.xapian.org/mailman/listinfo/xapian-discuss >
On 5/17/07, Olly Betts <olly@survex.com> wrote:> Flint supports using zlib to compress tags in the record and termlist > trees, and is now the default backend. The documentation has also been > improved.Is this zlib compression turned on by default? If it isn't how do you enable it? Can you enable it after the fact?
Bravo, Bravo and Congratulations! To all of you who are working on Xapian and are providing such a great tool : the more I use it, the more I understand the tag lines "highly adaptable toolkit" and "advanced indexing and search facilities"! I imagine that reaching the symbolic "1.0" version is an important step for such a project. Reading the mailing-lists since more than one year, I think I have a glimpse on the amount of work which was required... So, once again, all my kudos! -- Daniel M?nard