Re: SQL and XML (was RE: Help!)

[Steve Muench <Steve.Muench@oracle.com>]

Forwarding Jim Melton's reply (which bounced due to his not being
a member of the list)...

_____________________________________________________________________
Steve Muench - Developer, Product Manager, XML Evangelist, Author
"Building Oracle XML Applications" - www.oreilly.com/catalog/orxmlapp

===/ Included Reply from Jim Melton /===



Gentlepeople,

Although I do not (yet?) subscribe to this list, the following note was
brought to my attention by Steve Muench and I am replying in the hopes that
my contribution might be useful.


  ----- Original Message -----
  From: Mike.Champion@SoftwareAG-USA.com
  To: xml-dev@lists.xml.org
  Sent: Friday, July 27, 2001 6:58 AM
  Subject: SQL and XML (was RE: Help!)



  > -----Original Message-----
  > From: Bullard, Claude L (Len) [mailto:clbullar@ingr.com]
  > Sent: Friday, July 27, 2001 9:22 AM
  > To: Ronald Bourret; Purvesh Sharma
  > Cc: xml-dev
  > Subject: RE: Help!
  >

  > Depending on the data source, it is likely that
  > SQL will be the most industrially used XML query
  > language for some time.  It has the advantage
  > of clarity, ease, and years of experience in
  > optimization.   The XPath syntax is gnomic and
  > prone to misinterpretation particularly when
  > using namespaces.

I certainly agree that SQL will continue to be the most used query language
for some time, at least in commercial applications.  SQL has proven itself
to be remarkably resilient and flexible and its on-going development and
extensions are certainly indicative of health and robustness.  As the
response copied below indicates, it is becoming well known that SQL is now
being extended to work better in XML contexts.  I have more on this below.
(Full disclosure: As my signature block below indicates, I am quite active
in the SQL standards committees and work for a well-known SQL vendor.)

However, regardless of the sometimes baroque syntax and semantics of XPath
(particularly, as you note, when using namespaces), I think that XPath does
a pretty reasonable job of concisely expressing certain types of relatively
simple queries against XML documents.  It is in the more complex queries
where I feel that XPath begins to lose attractiveness.  I should observe
that my employer (Oracle Corp.) has used XPath syntax quite successfully
within our SQL engine for querying XML documents managed by that engine; I
view that as (at least) proof of concept that it is possible and quite
useful.


  Does anyone know about the status of the ANSI/ISO SQL effort
  to add XML support to SQL?  Looking quickly at the latest working
  draft at
ftp://sqlstandards.org/SC32/WG3/Progression_Documents/Informal_working_draft
s/wd-xml-2001-06.pdf
  I see a lot of wrestling with the mapping of SQL data types to XML Schema
data types.  Is there more coming in the way of mapping SQL operations onto
XML hierarchies (e.g., incorporating XPath-like expressions?), am I missing
something, or what?

An open group of interested development and standardization engineers, with
the full support of their employers, has been formed to address the issues
of ensuring that SQL and XML can play well together.  This group, known as
the SQLX Group, has a web presence at www.sqlx.org (the website is
maintained in a very busy participant's "spare" time, so don't expect
anything remotely resembling a commercial site).  The SQLX Group develops
technical solutions to issues related to the use of SQL in XML environments,
with a fairly natural tendency to focus more on "data-oriented" applications
than pure "document-oriented" (e.g., from what might be the SGML world)
ones.  The specifications that the SQLX Group develop are forwarded to the
formal SQL standards groups for consideration (and, I may brag, very
successfully to date) to be included in a new part of the SQL standard under
developm! ent. < br>
The document cited above is the most recent draft of that new SQL standard.
This current version is made up mostly of "infrastructure", meaning
specification that address those niggling little details of mapping between
XML names and SQL identifiers, XML Schema types and SQL data types, etc.  In
process are proposals that would add more substantive capabilities that take
advantage of the infrastructure currently in place.  As new drafts are made
available, they will be found at the same site (although the filenames will
evolve to reflect the status and the date, and the final subdirectory used
will also evolve to become either "CD" or "FCD" --- my hope is for the
latter).

It is not yet clear whether or not specifications will be adopted that
incorporate XPath or XPath-like expressions directly into SQL statements.
That possibility is being discussed by some SQLX participants, but it has
proved to be somewhat more controversial than might have been hoped and I am
unable to guess whether anything will be actually proposed in this area.
All I can advise on this area is to stay tuned.


  Or perhaps are the SQL standards people primarily interested in presenting
the results of an RDBMS query in XML syntax, and not querying for XML
content in a database?

Clearly, the SQL (standards) people are quite interested in this particular
issue, but I assure you that our interests are vastly broader than this.
Whether that interest extends to "querying for XML content in a database" is
a very difficult issue.  I am positive that a large fraction of SQL people
(standards and otherwise) are quite interested in that subject; whether even
that large fraction wants to directly extend the SQL language to serve that
goal is rather less obvious.  It is not yet clear to me (speaking
*personally* and not as my company's representative) whether the XML Query
language under development in the W3C will or will not be sufficiently
successful to obviate the need for such SQL extensions.  Again, stay tuned
(and/or get involved).

As you will see on the SQLX website, membership is open to interested
participants.  As in every other analogous endeavor, merely signing up won't
give you much beyond the ability to lurk.  Only active participation (read
"travel, time, and thinking") will give you influence on the outcome!

Hope this helps,
   Jim
========================================================================
Jim Melton --- Editor of ISO/IEC 9075-* (SQL)     Phone: +1.801.942.0144
Oracle Corporation            Oracle Email: mailto:jim.melton@oracle.com
1930 Viscounti Drive          Standards email: mailto:jim.melton@acm.org
Sandy, UT 84093-1063           Personal email: mailto:jim.melton@acm.org
USA                                                Fax : +1.801.942.3345
========================================================================
=  Facts are facts.  However, any opinions expressed are the opinions  =
=  only of myself and may or may not reflect the opinions of anybody   =
=  else with whom I may or may not have discussed the issues at hand.  =
========================================================================

----- Original Message -----
From: <Mike.Champion@SoftwareAG-USA.com>
To: <xml-dev@lists.xml.org>
Sent: Friday, July 27, 2001 06:58
Subject: SQL and XML (was RE: Help!)


|
|
| > -----Original Message-----
| > From: Bullard, Claude L (Len) [mailto:clbullar@ingr.com]
| > Sent: Friday, July 27, 2001 9:22 AM
| > To: Ronald Bourret; Purvesh Sharma
| > Cc: xml-dev
| > Subject: RE: Help!
| >
|
| > Depending on the data source, it is likely that
| > SQL will be the most industrially used XML query
| > language for some time.  It has the advantage
| > of clarity, ease, and years of experience in
| > optimization.   The XPath syntax is gnomic and
| > prone to misinterpretation particularly when
| > using namespaces.
|
| Does anyone know about the status of the ANSI/ISO SQL effort
| to add XML support to SQL?  Looking quickly at the latest working
| draft at
|
ftp://sqlstandards.org/SC32/WG3/Progression_Documents/Informal_working_draft
| s/wd-xml-2001-06.pdf
| I see a lot of wrestling with the mapping of SQL data types to XML Schema
| data types.  Is there more coming in the way of mapping SQL operations
onto
| XML hierarchies (e.g., incorporating XPath-like expressions?), am I
missing
| something, or what?
|
| Or perhaps are the SQL standards people primarily interested in presenting
| the results of an RDBMS query in XML syntax, and not querying for XML
| content in a database?
|