AW: XQuery frameworksHans-Juergen Rennau hrennau at yahoo.de
Mon Feb 1 23:49:56 PST 2010
Liam, 'extend standardization "beyond the query"' - that is an excellent formula, I think! As one of the reasons why the need is real I regard the challenge of data integration. I have a feeling that there is a remarkable contradiction between potential and reality. XQuery has a great potential as an integrator, what with the possibility to just *name* the resource (URI) and immediately have its structured content under your fingertips. But if the actual usage of XQuery does not yet live up to the potential, one of the reasons may be that it is difficult to set up a system of many queries to a robust whole, although the *single* queries can be written so easily and efficiently. Put differently: it makes a great difference if the focus of integration lies within a single query, or beyond. A movement "beyond the query": how about a processing model just one level above the query level, of moderate complexity, integrating the execution of single queries, for example by assigning to them properties which would allow to automate their squencing, input provision and output delivery? The XQuery spec provides for the possibility to write library modules, but it is not possible to formulate relationships between library functions. In general, we cannot say someting "about" a function. We cannot, for example, express that a query Q has a "logical" source document D, although the actual input is not (or not only) D itself, but the result produced by some other query Q0, likewise logically - or perhaps literally - processing D. A really simple XML language might make it clear that the command "apply Q to D" would have to be expanded by a framework into a chain: "apply Q0 to D, resulting in R; present Q to R". We do not have this level of abstraction which would allow for orchestrating the individual queries. The XQuery language level knows only XDM values and expressions. I wonder if just one level above the language a simple, standardized processing model might not be possible, simple enough to be attractive and really enhancing efficiency. With kind regards, Hans-Juergen Rennau ----- Ursprüngliche Mail ---- Von: Liam R E Quin <http://x-query.com/mailman/listinfo/talk> An: Hans-Juergen Rennau <http://x-query.com/mailman/listinfo/talk> CC: http://x-query.com/mailman/listinfo/talk Gesendet: Montag, den 1. Februar 2010, 2:14:47 Uhr Betreff: Re: XQuery frameworks On Fri, 2010-01-29 at 10:03 +0000, Hans-Juergen Rennau wrote: > Dear People, > > presently I am scribbling away at a little XQuery framework A few years ago when I did http://www.fromoldbooks.org/Search/ using XQuery, I wrote my own. It's far from perfect -- it is a perl script that checks the system load average (!), and the query parameters, and then reads an xml config file, and calls up an external xquery engine based on a variable -- so I could easily switch between galax, qizx and saxon, initially, and later dbxml too. It keeps a cache, to avoid firing up a JVM (two of those implementations are in Java) with each incoming HTTP request, and does get fairly high cache hit rates. It'd be a reasonable question, "why not use servlets", and a lot of the answer to that is that, when I wrote it, Sun's licence for Java was such that Linux distributions couldn't easily include it, and Java was a sufficient pain to set up that I avoided any architecture that depended on it. But then I started using Saxon, and, later qizx, both of which do use Java. More recently I've used dbxml, but not on the production system, as the packaging for dbxml is even weirder than for Java, and my server has changed operating systems several times over the years. All this is a long winded way of saying, a lot of it depends on the environment. I'm particularly interested (since I work at W3C...) in what we could do to extend standardisation "beyond the query" into more of a framework, and how, and what would be the value of that, but I think the XQuery Working Group right now does not have the resources to do that sort of work. Liam -- Liam Quin - XML Activity Lead, W3C, http://www.w3.org/People/Quin/ Pictures from old books: http://fromoldbooks.org/ Ankh: irc.sorcery.net irc.gnome.org www.advogato.org __________________________________________________ Do You Yahoo!? Sie sind Spam leid? Yahoo! Mail verfügt über einen herausragenden Schutz gegen Massenmails. http://mail.yahoo.com
PURCHASE STYLUS STUDIO ONLINE TODAY!
Purchasing Stylus Studio from our online shop is Easy, Secure and Value Priced!
Download The World's Best XML IDE!
Accelerate XML development with our award-winning XML IDE - Download a free trial today!
Subscribe in XML format