xml-dev - Re: Coca-Cola Data [Was RE: Is this Impossible !!]

["Steve Muench" <smuench@us.oracle.com>]

If you are talking about querying a middle-tier
cache of XML documents, this technique might work,
but if you are talking about XML-enabling an
ebusiness whose systems run on ERP and CRM and
custom apps with lots of relational data lying
around, none of these customers is intested in
mass-converting their terabytes of mission-critical
data into brontobytes of persisted DOM documents
to search them in a fixed, hiearchical structure
through XPath.

A company's mission-critical business data represents
*lots* of virtual XML documents depending on the
particular point of view of the moment. Not a single,
fixed document with a single root.

Do you want a Bill of Materials for the part ordered
on line item 10 of Order #10001 ? Or an XML document
representing Order 10001 and all its line items?

The ability to pick any important thing in the
system and "make it the root" of a virtual XML
document" is a powerful concept.

________________________________________________________
Steve Muench, BC4J Development Team & XML Evangelist
http://technet.oracle.com/tech/java
http://technet.oracle.com/tech/xml
----- Original Message -----
From: Prakash Mandgi <pmandgi@hotmail.com>
To: Steve Muench <smuench@us.oracle.com>
Sent: Thursday, October 21, 1999 3:03 PM
Subject: Re: Coca-Cola Data [Was RE: Is this Impossible !!]


| If you persist a DOM implementation in an object database you should be
able
| to query
| the DOM using XPATH without any hitch. ( if you have implemented XPATH but
| of course )
| The advantage of using Object databases is that you persist your object
data
| as is and
| retrieve it on the fly, without any expensive parsing or conversion of the
| relational data set
| into recreating the DOM document. ( mapping of relational data into the
DOM
| document
| or document fragment or node list etc )
|     The proper way to query a DOM document is to use XPATH and only XPATH
| and not
| OQL or SQL or any other query engine.
|
| Prakash
|
| ----- Original Message -----
| From: Steve Muench <smuench@us.oracle.com>
| To: Mark Birbeck <Mark.Birbeck@iedigital.net>
| Cc: <xml-dev@ic.ac.uk>
| Sent: Thursday, October 21, 1999 1:41 PM
| Subject: Re: Coca-Cola Data [Was RE: Is this Impossible !!]
|
|
| > Mark,
| >
| > I'm not advocating returning the SQL query results
| > in a SQL-query-results-looking format.
| >
| > By combining XSLT transformation with canonical
| > SQL query results (including richly nested
| > query results from an object/relational database)
| > you can achieve the situation where what the
| > user gets back is exactly the XML document you
| > want it to be.
| >
| > The point about a common XML-based query language
| > is right. Folks from many companies are now working
| > in the XML Query Working group to try and nail that
| > syntax for the future...
| >
| > ________________________________________________________
| > Steve Muench, BC4J Development Team & XML Evangelist
| > http://technet.oracle.com/tech/java
| > http://technet.oracle.com/tech/xml
| > ----- Original Message -----
| > From: Mark Birbeck <Mark.Birbeck@iedigital.net>
| > To: <xml-dev@ic.ac.uk>
| > Sent: Thursday, October 21, 1999 10:14 AM
| > Subject: Coca-Cola Data [Was RE: Is this Impossible !!]
| >
| >
| > | I think speed is not an issue. As a proportion of the time spent
sending
| > | the query, extracting the data and then packaging it up for return,
the
| > | time you take to convert the query from one syntax to another is
| > | negligible.
| > |
| > | However, what you lose by using SQL syntax IS an issue. You have data
| > | stored in a relational database which you query in a relational way,
but
| > | then return the results as a node set. This seems OK, until you
consider
| > | what you have presented to the outside world - a node set which can be
| > | queried using SQL syntax. Now, imagine that someone else stores their
| > | data in an object database, but also returns their results in XML.
They
| > | are presenting a node set which can be queried using some object query
| > | syntax.
| > |
| > | So we have two sets of data that to the outside world have similar
| > | structure - they are both a set of hierarchical nodes - but which have
| > | two different ways of being queried. Not good. Add to that, that the
| > | person who did the object database may convert you to his way of
| > | thinking and you then want to ditch your relational database, and
| > | suddenly all those queries that people have set up to access your
server
| > | have to change.
| > |
| > | My point is simply that just as you return data from your server in a
| > | common format - XML - so too you should accept requests in a common
| > | format. You should convert requests in, say, XPath, to SQL queries or
| > | object queries or whatever queries. Otherwise you are not really
| > | presenting an XML-interface to your data.
| > |
| > | And we'll never get the Coca-Cola effect (the whole world singing the
| > | same song, if you're wondering).
| > |
| > | Best regards,
| > |
| > | Mark Birbeck
| > | x-port.net Ltd.
| > |
| > | > -----Original Message-----
| > | > From: AlanM [mailto:AlanM@SYNECTICS.Soft.net]
| > | > Sent: Thursday, October 21, 1999 5:11 PM
| > | > To: xml-dev@ic.ac.uk
| > | > Subject: RE: Is this Impossible !!
| > | >
| > | >
| > | >
| > | > Hi guys,
| > | > Since speed is criteria is it faster to send Queries to the
| > | > server in SQL or
| > | > XML. If sent in XML we will have to parse it and then convert
| > | > it into SQL
| > | > again, which may slow down the process.
| > | >
| > | >
| > | > > -----Original Message-----
| > | > > From: Mark Birbeck [SMTP:Mark.Birbeck@iedigital.net]
| > | > > Sent: Thursday, October 21, 1999 8:29 PM
| > | > > To: xml dev mailing list
| > | > > Subject: RE: Is this Impossible !!
| > | > >
| > | > >
| > | > >
| > | > > Hi everyone,
| > | > >
| > | > > Very busy so can't give you in depth on this, but I note
| > | > that you are
| > | > > using VB scripting so here is a bit of code from a VB
| > | > script version of
| > | > > our SOAP server.
| > | > >
| > | > > It's not that complicated, but it provides you with an easy
| > | > way to create
| > | > > a results set from any values. For example, it could build
| > | > a node list
| > | > > from a three element array, with the first element being
| > | > another array
| > | > > that contains record sets, the second being an integer, and
| > | > the third
| > | > > being an XML DOM node (should you want to).
| > | > >
| > | > > Note that it copes with nested recordsets - which is great
| > | > if you are
| > | > > using the Microsoft data shaping facility in ADO. In fact using
this
| > | > > routine and data shaping is probably the easiest way to get an XML
| > | > > document with real structure out of a relational database.
| > | > >
| > | > > Two quick things that are slightly off theme:
| > | > >
| > | > > First - if you are returning XML you should really consider
| > | > querying in
| > | > > XML. The client should not really be sending SQL queries
| > | > since you are not
| > | > > returning a SQL result set. You should accept an XML query
| > | > (say, using
| > | > > XPath) and then convert that on the server to an SQL query.
| > | > That way the
| > | > > fact that your data is currently relational will be hidden
| > | > (since it may
| > | > > change).
| > | > >
| > | > > Second - for those interested - note how much better SOAP
| > | > structure is
| > | > > than XML-RPC! I've fallen for it!!
| > | > >
| > | > > Best regards,
| > | > >
| > | > > Mark Birbeck
| > | > > x-port.net Ltd.
| > | > >
| > | > >
| > | > >
| > | > > -----Original Message-----
| > | > > From: Goyal, Sanjeev [mailto:Sanjeev.Goyal@usa.xerox.com]
| > | > > Sent: Thursday, October 21, 1999 2:49 PM
| > | > > To: 'Abhishek Srivastava'
| > | > > Cc: xml dev mailing list
| > | > > Subject: RE: Is this Impossible !!
| > | > >
| > | > >
| > | > > Abhishek,
| > | > >
| > | > > Most of the XML Parser implementations provides mechanism to
| > | > > generate well formed XML documents from the DOM Tree. I
| > | > have used sun's
| > | > > XML parser, it provides APIs to generate well-formed XML from your
| > | > > Document Node.
| > | > >
| > | > > Hope it helps.
| > | > > Sanjeev
| > | > >
| > | > > -----Original Message-----
| > | > > From: Abhishek Srivastava [mailto:abisheks@india.hp.com]
| > | > > Sent: Wednesday, October 20, 1999 1:29 AM
| > | > > To: xml dev mailing list
| > | > > Subject: Is this Immpossible !!
| > | > >
| > | > >
| > | > >
| > | > > Hi All,
| > | > >
| > | > > I have the following requirement. A client
| > | > wants to query a
| > | > > database. It sends out an SQL Query to the Database server.
| > | > At the server
| > | > > side, instead of returning a recordset, it returns an XML File.
| > | > >
| > | > > The client on receiving the XML file, parsers it for
| > | > > "Validity" (to be sure that all elements it had asked for
| > | > are there).
| > | > >
| > | > > Problem is that at the server side, How to build the xml
| > | > > document.
| > | > > Presently, I am doing something like this
| > | > >
| > | > > String("<Name>") + rs.getField("auName") +
| > | > String("</Name>")
| > | > >
| > | > > However, this is a very inelegant approach as
| > | > the code is
| > | > > full of such string concatenations.
| > | > >
| > | > > Is there a more elegant solution to this ?  All
| > | > the material
| > | > > on XML talks about parsing and reading XML. What about
| > | > writing them ? Are
| > | > > there DOM / SAX Api extensions available that would
| > | > > create "VALID" XML documents without clutter in
| > | > the code ?
| > | > >
| > | > > Any help would be greatly appreciated.
| > | > >
| > | > > Thanks & Best Regards,
| > | > > Abhishek.
| > | > >
| > | > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
| > | > >     _/               Abhishek Srivastava
| > | > >    _/                Hewlett Packard ISO
| > | > >   _/_/_/   _/_/_/    -------------------
| > | > > _/   /   _/  _/     (Work)   +91-80-2251554 x1190
| > | > > _/  _/   _/_/_/      (Ip)     15.10.47.37
| > | > >         _/           (Url)
| > | > > <http://sites.netscape.net/abhishes/homepage>
| > | >
| > | > >        _/            You've heard it all by
| > | > now. Get wired
| > | > > or get whacked.
| > | > >                      You're networking or you're not
| > | > > working. Dot-com or die
| > | > >                      - SUN MICROSYSTEMS
| > | > >
| > | > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
| > | > >
| > | > >  << File: SOAP Return.asp >>
| > | >
| > | > xml-dev: A list for W3C XML Developers. To post,
| > | mailto:xml-dev@ic.ac.uk
| > | Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on
| > | CD-ROM/ISBN 981-02-3594-1
| > | To unsubscribe, mailto:majordomo@ic.ac.uk the following message;
| > | unsubscribe xml-dev
| > | To subscribe to the digests, mailto:majordomo@ic.ac.uk the following
| > | message;
| > | subscribe xml-dev-digest
| > | List coordinator, Henry Rzepa (mailto:rzepa@ic.ac.uk)
| > |
| > | xml-dev: A list for W3C XML Developers. To post,
mailto:xml-dev@ic.ac.uk
| > | Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on
| > CD-ROM/ISBN 981-02-3594-1
| > | To unsubscribe, mailto:majordomo@ic.ac.uk the following message;
| > | unsubscribe xml-dev
| > | To subscribe to the digests, mailto:majordomo@ic.ac.uk the following
| > message;
| > | subscribe xml-dev-digest
| > | List coordinator, Henry Rzepa (mailto:rzepa@ic.ac.uk)
| > |
| > |
| >
| >
| > xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev@ic.ac.uk
| > Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on
| CD-ROM/ISBN 981-02-3594-1
| > To unsubscribe, mailto:majordomo@ic.ac.uk the following message;
| > unsubscribe xml-dev
| > To subscribe to the digests, mailto:majordomo@ic.ac.uk the following
| message;
| > subscribe xml-dev-digest
| > List coordinator, Henry Rzepa (mailto:rzepa@ic.ac.uk)
| >
| >
|


xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev@ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To unsubscribe, mailto:majordomo@ic.ac.uk the following message;
unsubscribe xml-dev
To subscribe to the digests, mailto:majordomo@ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa@ic.ac.uk)