[
Lists Home |
Date Index |
Thread Index
]
Hi murali,
On Mon, 18 Aug 2003 07:54:57 -0700 (PDT)
Murali Mani <mani@CS.UCLA.EDU> wrote:
> what are the tools which currently do this XQuery/XPath -> SQL
> translation?
My product is one of the tool, supporting XPath 1.0.
http://gborg.postgresql.org/project/xpsql/projdisplay.php
> (a) What is the input for this tool? - Is it the query/path expression and
> the annotated XML Schema?
as for our approach, it is DTD(IWO, description of mapping) less approach.
users do not have to know the structure mapping.
There are mainly separated into the following two techniques
treating XML effectively in RDBMS.
(1) model-mapping approach (IOW, DTD less)
(2) structure-mapping approach (To say simple, which have DTDs)
#There is also a naive technique, storing XML data in CLOB as an atomic value.
XpSQL is a model-mapping approach, and node oriented.
XpSQL is based on XPath 1.0 data model, and have some supports for DOM L2.
> (b) Also what XQuery/XPath operators are supported by this tool? For
> example, do they support descendant, if so, how do they do the
> translation??
In XpSQL, all XPath-Axis is supported.
the way is same as NXDB, select using an ordered tree.
showing is my approch based on dewey decimal classification.
http://gborg.postgresql.org/images/projects/xpsql/screenshots/axis-dewey_big.png
> (c) Of course, the next question is: do any of these tools try to do
> minimization of the resulting SQL query, in other words, try to minimize
> the number of unnecessary joins or other wasteful operations..
yeah, we have a path table, storing path from root, like /aaa/bbb/ccc.
this approach reduces number of the joins.
+-------------------------------------------------------------------+
Makoto Yui <yuin@bb.din.or.jp>
Key fingerprint = 6462 E285 97D8 1323 40C4 F9E5 EB0F 9DE6 1713 219E
+-------------------------------------------------------------------+
|
