On 10. feb. 2014 17:35, Madushanka Fonseka wrote:
Hi,
I have been working with Apache Derby for last few months (for a
research) and I need to know more about following mechanisms.My
research is based on innovating Derby to support fuzzy-logy based queries
(Ex. select * from student where /Avg/ is /good, /select * from
employee where /salary /is /low/)
*1. How Qualifiers being created for different predicates ?*
Start looking at the WHERE clause code, e.g. in SelectNode.java, later
pushed in ProjectRestrictNodes (in SelectNode#preprocess)
*2. How a query or so as to speak something like "Salary > 55000"
being validated against a row collection ?*
The evaluation of predicates are done in generated code, cf. the code in
ProjectRestrictNode.java. Some predicates are
pushed down to the storage layer in the form of Qualifiers, cf. where
PRN#modifyAccessPath calls pushUsefulPredicates.
The machinery is quite complex, so using an IDE/debugger will help
understand what's happening.
*3. How heap + conglomerate machinery get connected with this
Qualifie*rs *?*
Cf above.
Hope this helps,
Dag
I would be pleased if you can give me some insight related to above
questions.
*And on a different note I'd like to submit a research paper in regard
to my concept. I'd also like to know what would be the process for that.*
Thanks,
Madushanka Fonseka
Computer Science Dept,
Faculty of Engineering
University of Moratuwa
Sri Lanka
