Hi Ashish,

Right now there are two active "branches" of ht://Dig releases.

The 3.1.x branch is for production servers and is very stable--there will
probably be at least one more release of this branch to take care of some
bugs reported in 3.1.6.

The 3.2 branch (from 3.2.0b1 to the current snapshots of 3.2.0b4) is the
"next generation" of ht://Dig and includes full support for phrase
searching, restricting queries to titles, authors, etc., per-URL and
per-server configuration options and builds databases as indexing
occurs--no need for a separate post-indexing "htmerge" phase. The 3.2 code
is based on the mifluz library underneath.

There are two major projects holding up 3.2 right now.
1) htsearch cleanup: The current htsearch isn't very modular and the query
parser is very inflexible. Quim Sanmarti contributed a new query parser
and I'm almost finished with code to use this framework in
htsearch.
Additional htsearch work will include: parsing "field" queries
(i.e. titles only), faster result sorting (using a heap), a cleaner
Display class, support for caching queries and results and more.

2) mifluz merge: The current ht://Dig code uses mifluz-0.19, which is
slower and buggier than the current 0.23 code. Changes in the mifluz CVS
tree need to be merged back into ht://Dig and our code will need to be
updated some to use the new API. This will improve indexing speed and
reliability, among others.

There are also a variety of other needs, including someone to look at
adding in Unicode/UTF-8 support to ht://Dig--the backend should support
it, but some of the code assumes 1-byte characters (e.g. when
parsing). Also, while the database backend now supports features for XML
indexing (i.e. user-defined word flags), the indexing frontend doesn't
use these.

Of course there's also a list of feature requests, bugs, etc. on
SourceForge, a pressing need to do some profiling and leak detection on
3.2 to improve performance, and any ideas you might have yourself that
would prove interesting.

If any of these sounds interesting in general, I'm sure myself or someone
else on the htdig-dev mailing list can fill you in on more details.

-Geoff

On Sat, 23 Mar 2002, ashish mohole wrote:

> Hi Geoff,
> 
> I looked at the website
> http://www.htdig.org/
> and it seems interesting..will you please give me more details about where 
> the project is going? And in what areas you are looking for help ?
> 
> thanks,
> Ashish


_______________________________________________
htdig-dev mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/htdig-dev

Reply via email to