BaseX-Talk

basex-talk@mailman.uni-konstanz.de

4 participants
5268 discussions

Possible lingering issues with database optimization in v11
by Eliot Kimber 04 Mar '24

04 Mar '24

I’m continuing to test my process for loading data which depends on optimizing databases, some of which are pretty large (100+MB with 100s of 1000s of elements). I’m testing on both macOS and linux using the 27 Feb build on macOS and the 29 build on linux (just a matter of when I downloaded them). What I’m seeing is that when I test with a relatively small content set the optimization completes reliably and everything works as it should. When I test with a realistically large data set, the optimization either takes a very long time (as much as an hour to complete) or never completes with the server at 100% CPU utilization. It seems to be worse on macOS but it’s difficult to verify, partly because a test takes several hours to run. I have the BaseX source code available locally, although I’m unable to compile with maven due to internal maven issues (we have a pretty locked down maven proxy and I don’t know maven well enough to know how I might configure around that). Is there anything I can do to diagnose this issue to at least confirm or deny that there are still deadlock issues with the optimization? I assume that it should not take 10s of minutes to optimize even a large database. Here's the details for a typical database that has failed to optimize on macOS: [cid:image001.png@01DA6D49.2FC41230] Thanks, Eliot _____________________________________________ Eliot Kimber Sr Staff Content Engineer O: 512 554 9368 M: 512 554 9368 servicenow.com<https://www.servicenow.com> LinkedIn<https://www.linkedin.com/company/servicenow> | Twitter<https://twitter.com/servicenow> | YouTube<https://www.youtube.com/user/servicenowinc> | Facebook<https://www.facebook.com/servicenow>

2 2

Optimize database never returns, leaves database in "opened by another process" state
by Eliot Kimber 04 Mar '24

04 Mar '24

Using Basex 10.7 on Linux. I’m running a sequence of jobs to update and optimize a set of databases following loading a number of documents created dynamically (as opposed to being read from disk). I’m seeing a new behavior, which is that the optimization step never completes but also doesn’t show any error in the log. The database shows no items and is in the locked by another process state if I try to drop it. This behavior seems to be consistently repeatable with my current code base (I’m working on some code updates, so it’s possible I’ve introduced something that would cause this behavior but I haven’t changed the code that leads up to the failing optimize). The server has plenty of disk space, etc. Optimization code is: try { if (db:exists($database)) then ( util:logToConsole('dbadmin:optimizeDatabase', ``[Optimizing database `{$database}`]``), db:optimize($database, true(), $dbadmin:dbOptimizeOptions) ) else util:logToConsole('dbadmin:optimizeDatabase', ``[Database '`{$database}`' does not exist. Nothing to optimize.]``) } catch * { util:logToConsole( 'dbadmin:optimizeDatabase', ``[Exception optimizing database '`{$database}`': `{$err:code}` - `{$err:description}`]``, 'error') } And the optimization options are: declare variable $dbadmin:dbOptimizeOptions as map(*) := (: Turn on all the indexes :) map { 'attrindex' : true(), 'tokenindex' : true(), 'textindex' : true(), 'ftindex' : true() }; This code has been working fine for a long time and I’ve been running 10.7 for a least a couple of months, so I’m wondering: A) Would would cause this behavior? B) How can I diagnose it short of debugging the Java code (which I can do but it’s non-trivial for me to set up). Thanks, Eliot _____________________________________________ Eliot Kimber Sr Staff Content Engineer O: 512 554 9368 M: 512 554 9368 servicenow.com<https://www.servicenow.com> LinkedIn<https://www.linkedin.com/company/servicenow> | Twitter<https://twitter.com/servicenow> | YouTube<https://www.youtube.com/user/servicenowinc> | Facebook<https://www.facebook.com/servicenow>

3 10

Re: [basex-talk] Item Separator
by Leo Studer 29 Feb '24

29 Feb '24

Dear Martin Thank you for investigating. > The "text1, , text2" I only see if oXygen shows the result as a sequence in its result viewer; I would suggest to ask the oXygen support why you get that result for that case, it seems like some quirk in their result viewer (it shows two xs:string items but for reasons I can't investigate serializes the two items for the result viewer as "text1, , text2"). > This is incredible news. I will get to Oxygen about that. Really troubling … Blessings, Leo

1 0

Odd "Access denied" messages in the log
by Eliot Kimber 28 Feb '24

28 Feb '24

I have an application (our Mirabel system) running on a server inside our firewall (so not visible to the open Internet). I’ve recently started seeing messages like this in the log: Access denied: ��. Access denied: PRI * HTTP/2.0 SM . Where the value reported can be quite varied, but is often unrenderable characters or other stuff. (In this case the characters are all \uFFFD). The log messages all report the same IP address. This server does not use named users, so there’s no authentication required to access it. The IP address is not one of my own servers, so I don’t think it’s something generated by my own code. Any idea what this might be? It’s started relatively recently, which makes me think it might be some sort of penetration test. Cheers, E. _____________________________________________ Eliot Kimber Sr Staff Content Engineer O: 512 554 9368 M: 512 554 9368 servicenow.com<https://www.servicenow.com> LinkedIn<https://www.linkedin.com/company/servicenow> | Twitter<https://twitter.com/servicenow> | YouTube<https://www.youtube.com/user/servicenowinc> | Facebook<https://www.facebook.com/servicenow>

2 1

Item Separator
by Leo Studer 27 Feb '24

27 Feb '24

Hello I am serialising the following expressions on oxygen/saxonEE12.3 as text (item-separator=“”) and wonder why the differente outputs let $xml:=<xml><element>text1</element><element>text2</element></xml> return $xml/element/text() gives text1text2 And let $xml:=<xml><element>text1</element><element>text2</element></xml> return $xml/element/string() text1 text2 When I set the item-separator =“, “ then I get text1, text2 Or text1, , text2 Any explanations??? Thanks, Leo

3 3

file:write, formatting, and maps in select attributes
by Graydon Saunders 24 Feb '24

24 Feb '24

Hello -- I'm trying to regularize the formatting of some XSLT files by loading them (via doc()) and writing them back out via file:write() I know about the *indent* and *indent-attributes* serialization parameters, and mostly they're doing what I want. What I'd like is to preserve a one-line-per-entry format for a map in a select attribute. For example, I'd like to keep <xsl:variable as="map(xs:integer,xs:string)" name="hexNum" select=" map { 1: '0', 2: '1', 3: '2', 4: '3', 5: '4', 6: '5', 7: '6', 8: '7', 9: '8', 10: '9', 11: 'a', 12: 'b', 13: 'c', 14: 'd', 15: 'e', 16: 'f' }" static="yes" /> instead of <xsl:variable as="map(xs:integer,xs:string)" name="hexNum" select=" map { 1: '0', 2: '1', 3: '2', 4: '3', 5: '4', 6: '5', 7: '6', 8: '7', 9: '8', 10: '9', 11: 'a', 12: 'b', 13: 'c', 14: 'd', 15: 'e', 16: 'f' }" static="yes"/> I don't think there's a way to do this with the serialization parameters, but is there a way to do this? Thanks! Graydon -- Graydon Saunders | graydonish(a)fastmail.com Þæs oferéode, ðisses swá mæg. -- Deor ("That passed, so may this.")

3 3

basex failed to cast large numbers as xs:integer
by mx z 22 Feb '24

22 Feb '24

Hi, I found a bug in basex. Version: 1ce0618 (2024.02.22) Java: openjdk 17.0.9 2023-10-17 LTS OpenJDK Runtime Environment Corretto-17.0.9.8.1 (build 17.0.9+8-LTS) OpenJDK 64-Bit Server VM Corretto-17.0.9.8.1 (build 17.0.9+8-LTS, mixed mode, sharing) OS: Distributor ID: Ubuntu Description: Ubuntu 22.04.3 LTS Release: 22.04 Codename: jammy input.xml: <a xmlns:C="p"> <C:A id="1">1458820431</C:A>1438466813 </a> query.xq: //*[not(. castable as xs:integer)] Reproduce: java -cp basex-core/target/basex-11.0-SNAPSHOT.jar org.basex.BaseX -i input.xml query.xq It printed out <a xmlns:C="p"> <C:A id="1">1458820431</C:A>1438466813 </a> While basex should return empty, as both 1458820431 and 1438466813 are integers. Max Zhang

2 1

Found problems with map:for-each
by Vincenzo Cestone 20 Feb '24

20 Feb '24

Hi all, with the basex 10.7 version (but even in 9.6 version I have the same issue) I found that the following code wont work as expected: (: It wont work :) let $files := map {"hello1.txt" : xs:base64Binary('SGVsbG8gd29ybGQ='), "hello2.txt" : xs:base64Binary('SGVsbG8gd29ybGQ=')} let $w := map:for-each($files, function($filename, $content) { file:write-binary($filename, $content, 0) }) return () that is, it will not write the two files hello1.txt and hello2.txt in the basex_home/bin folder. But, if you return $w instead: (: It works :) let $files := map {"hello1.txt" : xs:base64Binary('SGVsbG8gd29ybGQ='), "hello2.txt" : xs:base64Binary('SGVsbG8gd29ybGQ=')} let $w := map:for-each($files, function($filename, $content) { file:write-binary($filename, $content, 0) }) return $w With a similar implementation, with the classic FLWOR, the issue does not arise, even if I return the empty sequence: (: It works :) let $files := map {"hello1.txt" : xs:base64Binary('SGVsbG8gd29ybGQ='), "hello2.txt" : xs:base64Binary('SGVsbG8gd29ybGQ=')} let $w := for $filename in map:keys($files) return file:write-binary($filename, map:get($files, $filename), 0) return () that is it will write two files hello1.txt and hello2.txt in basex_home/bin folder. Note that the files in the examples above are for demonstration purposes, however in my code they are actually binary files. My java version is a Oracle JDK 17 This problem also occur to others? Thanks, Vincenzo

2 2

Help with usage of Regex while deleting resources
by Deepak Dinakara 20 Feb '24

20 Feb '24

Hi, I wanted to know if it's possible to give a regex while deleting a resource. I have documents stored in a hierarchy of collections like {year}{month}/doc.xml. Eg: 202301/abc.xml, 202302/def.xml. If I want to delete a resource "abc.xml", Is it possible to issue commands like "*db:delete("db-name", '/*/abc.xml')*" ? Right now, I can do a XQuery with db:list and endsWith and get the complete path of "abc.xml". But regex would have been very handy. Similarly I also want to execute queries against a list of collections using regex. Something like "*for $document in collection('db-name/20230*')*" (First 9 months of 2023) Right now, I am doing something like "for $i in ('01', '02', '03', '04', ... '09') for $document in collection('test-collection/2023' || $i)" But if there are better ways, kindly let me know. Thank you, Deepak

7 10

Slow full-text querying
by Murray, Gregory 18 Feb '24

18 Feb '24

Hello, I have a database with several thousand XML documents, although I have tens of thousands I’d like to add. Each XML document contains a book — both the bibliographic metadata such as title, author, etc. (each in its own element) and the complete OCR text of all pages of the book. Each page of text from each book is in a <page> element with a single text node containing all words from that page in the book, resulting in large blocks of text. I’ve added a full-text index and optimized it. I am finding that full-text searching is very slow. The query shown below consistently takes about 20 seconds to run, even though there are only about 7400 documents. Obviously that’s far too slow to use the query in a web application, where the user expects a quick response. My first thought is whether the query is actually using the full-text index. Is there a way for me to determine that? I’m also wondering if my query is crude or is missing something. I don’t need the text nodes containing the search words; I only need to know which documents contain the words. let $keywords := "apple" for $doc in collection("theocom") let score $score := $doc contains text {$keywords} order by $score descending where $score > 0 return concat($score, " ", base-uri($doc)) As you can see, I’m searching all text in the entirety of each book. Is there a way to rewrite such a query for faster performance? Also, I’m wondering if the structure of the XML documents is such that the documents themselves need to have smaller blocks of text. For example, if the OCR text were contained in <line> elements, each containing only a single line of text, as printed in the original physical book, would full-text searching be noticeably faster, since each text node is much smaller? Thanks, Greg

2 2

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

BaseX-Talk