- BaseX-Talk - mailman.uni-konstanz.de

Oops
by NewIntellectual 15 Jun '10

15 Jun '10

A couple of corrections to my previous post. First, I realized after posting that my statement "including retrieval and processing of title information for the documents" isn't quite true. I forgot that I now lazy-load that particular metadata for each book, on demand (which is also fast.) However, it is definitely true that BaseX is very rapidly finding all of those hits, with retrieval of book-level metadata for each hit. Also, apologies to Christian Grün for misspelling his last name as Guen in my last post.

1 0

Experience with BaseX (reply)
by NewIntellectual 15 Jun '10

15 Jun '10

Erik - My own experience with BaseX 6.0+ has been extremely positive. I needed a fast, multi-user (reading) database that implemented not just XQuery, but XQuery with standard Full Text extensions. And not for a toy database, but for a database of historical documents, initially with about 200,000 pages of text with significant internal structure. I initially stress-tested BaseX for multiple users/queries and found a problem, which Christian Guen took seriously and rapidly fixed. (When I say stress-tested, I mean I really hammered it.) The same has been true for a few other glitches and questions. Overall, BaseX has been very solid and very fast. It is crazy-fast for the complete indexing that it does, and it is really fast for queries. Fast enough to find tens of thousands of hits in a structured XQuery+full text search of about 200,000 book pages in under 2 seconds, including retrieval and processing of title information for the documents. So I really like BaseX for its practical utility, and as a programmer I also appreciate that the overall approach to the system is clean and pretty rigorous. They set out to implement actual W3C standards for XQuery and XQuery with Full Text extensions, and that's what you get, plus some useful extensions. The included simple GUI is actually very useful, not just a toy. I place BaseX in distinct contrast to the older and apparently more established eXist system. My experience testing with it, over time as the system has evolved, has not been positive. It is cumbersome to set up - especially for indexing - and it does not follow W3C standards for XQuery Full Text but instead uses its own syntax, and if you look at the mailing list for it, it seems to have a large number of ongoing bugs. It really feels like a big, unstable hack. From a performance perspective, a much smaller subset of my full database began to seriously slow down queries. As far the future goes, my biggest hope is that BaseX becomes better at being a true writable database system. While it appears to handle standard XQuery Update without problems (which itself is impressive!), this is restricted to one writer at a time, as I understand it. I would not try to use BaseX currently as a general purpose writable database, but I also don't need it. In my view, there is a place for XML databases alongside use of standard relational systems. In that regard, I highly recommend the H2 database system if you need a clean, powerful, fast Java relational DBMS - it is greatly superior to, e.g. Derby. So it probably goes without saying, but I am a big fan of BaseX and I hope to see it grow in popularity and features, as it deserves. Phil Oliver Oliver Computing LLC

1 0

XQuery Update syntax
by NewIntellectual 15 Jun '10

15 Jun '10

As an FYI for those starting with XQuery Update, I've found this to be a short but very practical and useful guide: http://www.xmlmind.com/_tutorials/XQueryUpdate/index.html I've successfully used BaseX to apply large numbers of updates to a database containing roughly 200,000 pages-equivalent of text.

2 1

exists(<xpath>) returns empty, database corruption
by Swaminathan S 14 Jun '10

14 Jun '10

Hi, I am using BaseX version 6.03 using client server mode. I face 2 kinds of problems: 1) sometimes, BaseX returns nothing for exists(<xpath>) function (only true or false is expected) - in these cases, I execute some XQueries(XQuery-updates and XQueries) using XQuery API.This starts happening consistently for a few xpaths but not for all xpaths. Just to give you the scenario, my xupdates are always done such that I export the database after every database change. Also, I operate on only on one database(one xml file). 2) also sometimes, the database gets corrupted (some elements get moved beyond the end of the root element). I want to know if you have ever encountered this problem, the cause and if they have been fixed in newer versions. Thanks, Swaminathan Sahiram

4 4

Search term highlighting?
by NewIntellectual 09 Jun '10

09 Jun '10

Is there a ready way to perform search term highlighting in a fulltext query result? I found a reference to a 'ft:highlight' function at http://www.xmlmind.com/qizx/_distrib/docs/manual/fulltext_extensions.htmlbut this does not apparently exist in BaseX or as a standard function.

2 1

Yahoo! Auto Response
by tameraso3per＠yahoo.com 08 Jun '10

08 Jun '10

Hi. Prescription medication or weight-|oss surgery is typica||y used in addition to dietary changes http://Kaileyx32.servegame.com?Jonty-Vandemortel.html Jonty Vandemortel

1 0

Confirming that your fix works well, and a question
by Stefano Santoro 07 Jun '10

07 Jun '10

Dear Lukas

2 1

Binding for BaseX in Rebol
by sf＠sabufrancis.com 05 Jun '10

05 Jun '10

Hi Thank you for your wonderful Native XML database system. I have been working with it for a few months now, and am quite happy to submit a Rebol binding for BaseX. This was based by my understanding of the Java Binding given by you. Please take a look here http://code.google.com/p/reb-basex Let me have your feedback! Thanks Regards Sabu Francis

2 1

Fwd: Having trouble inserting elements using XQJ two times with the same prepared expression
by Stefano Santoro 02 Jun '10

02 Jun '10

Hi Lukas, just to be thorough I tried all the steps I have described in my earlier e-mail on a different workstations running a different OS (linux), and the attempt yielded the same Exception. Ciao Stefano On Fri, May 28, 2010 at 1:14 PM, Stefano Santoro <stefano(a)santoro.org> wrote: > Hi Lukas, > > Interesting. I have gotten all your maven repository jar files and > included them in the class path as follows: > > $ echo $GVYCP > C:\opt\jlib\basex-api\xmldb-api-1.0.jar;C:\opt\jlib\basex-api\xqj-api-1.0.jar;C:\opt\jlib\basex-api\basex-api-6.1.3.jar;C:\opt\jlib\basex-api\basex-6.1.3.jar > > the same but more clearly listed > > $ dvcp GVYCP > 0) C:\opt\jlib\basex-api\xmldb-api-1.0.jar > 1) C:\opt\jlib\basex-api\xqj-api-1.0.jar > 2) C:\opt\jlib\basex-api\basex-api-6.1.3.jar > 3) C:\opt\jlib\basex-api\basex-6.1.3.jar > > I have run the same java class I have enclosed in the earlier version > as follows: > > $ java -version > java version "1.6.0_17" > Java(TM) SE Runtime Environment (build 1.6.0_17-b04) > Java HotSpot(TM) Client VM (build 14.3-b01, mixed mode, sharing) > > $ java -cp ".;${GVYCP}" NPEReproducer > java.lang.ArrayIndexOutOfBoundsException: 5 > at org.basex.query.up.Primitives.add(Primitives.java:52) > at org.basex.query.up.DBPrimitives.add(DBPrimitives.java:40) > at org.basex.query.up.Updates.add(Updates.java:90) > at org.basex.query.up.Insert.atomic(Insert.java:102) > at org.basex.query.up.Insert.atomic(Insert.java:30) > at org.basex.query.expr.Expr.iter(Expr.java:58) > at org.basex.query.QueryContext.iter(QueryContext.java:322) > at org.basex.query.QueryContext.iter(QueryContext.java:290) > at org.basex.api.xqj.BXQDynamicContext.execute(BXQDynamicContext.java:246) > at org.basex.api.xqj.BXQPreparedExpression.executeQuery(BXQPreparedExpression.java:54) > at NPEReproducer.reproduceNPE(NPEReproducer.java:51) > at NPEReproducer.main(NPEReproducer.java:60) > > I have also checked out your basex trunk and was able to build > basex-6.1.3.jar, and even with that jar the same exception was thrown. > > I was not able to build the api maven project as it could not resolve > the jax-rx dependency.(see attached log) > > I was able though to build an api jar from your trunk that contained > only the xml:db and XQJ drivers from trunk, and with that too the same > exception was thrown. > > So I have tried with the jars that I have downloaded from: > > http://www.inf.uni-konstanz.de/dbis/basex/maven/org/xmldb/xmldb-api/1.0/xml… > http://www.inf.uni-konstanz.de/dbis/basex/maven/javax/xml/xquery/xqj-api/1.… > http://www.inf.uni-konstanz.de/dbis/basex/maven/org/basex/basex-api/6.1.3/b… > http://www.inf.uni-konstanz.de/dbis/basex/maven/org/basex/basex/6.1.3/basex… > > I have also built basex-6.1.3.jar and basex-6.1.3.jar (w/o jaxrx) from > your trunk, and yet I was not able to run it as you were able to do. > > Is there anything I am missing? > > Ciao > Stefano > > > > On Fri, May 28, 2010 at 9:15 AM, Lukas Kircher > <lukaskircher1(a)googlemail.com> wrote: >> Hey Stefano, >> >> thanks for your report. >> >>> I am enclosing a java class file that reproduces this problem against >>> 6.1.3 basex jar files >> >> >> Unfortunately I wasn't able to reproduce the exception by running your >> code - neither with version 6.1.3 nor with our current repository version. >> It runs fine on my machine. Do you have a clue where I could've missed >> something? >> >> Or - if you have another case at hand where you encounter the same problem, >> just let us know and we'll see what we can do. >> >> Kind regards, >> Lukas >> >

2 2

Fwd: Re: dropping DataBase
by Leonard Wörteler 30 May '10

30 May '10

Hi Bea, i didn't realize that you're not on the basex-talk list. Sorry for that. Leo -------- Original-Nachricht -------- Betreff: Re: [basex-talk] dropping DataBase Datum: Sun, 30 May 2010 18:34:42 +0200 Von: Leonard Wörteler <leo(a)woerteler.de> An: basex-talk(a)mailman.uni-konstanz.de Hi again, Just in case you're interested: A simple way to create a main-memory based DB in BaseX, given the path to an XML document, would be final Data db = CreateDB.xml(IO.get(path), CTX.prop); CTX.openDB(db); // open DB in current Context By the way, this also works if you feed IO.get(..) an XML string instead of a path. Cheers again, Leo Am 30.05.2010 17:10, schrieb Leonard Wörteler: > Hello Beatriz, > > Am 30.05.2010 16:32, schrieb Beatriz Sancho: >> static public Vector queryExtraeDatosCabecera(final String query) throws >> BaseXException, QueryException, IOException { >> WorkSpace sp1 = new WorkSpace(); >> String workspace = sp1.getWorkspace(); >> new CreateDB(workspace + "\\cabecera.xml", "DBCabecera").execute(CONTEXT); >> [...] >> new DropDB("DBCabecera").execute(CONTEXT); >> return datos; >> } > > wouldn't it be much easier to use an in-memory DB in this case? No need > to explicitly drop it and you probably would even get better performance > because of reduced I/O. > > Just my 0.02$...

2 1