BaseX-Talk

basex-talk@mailman.uni-konstanz.de

1 participants
5252 discussions

How to produce multipart/mixed results with RESTXQ?
by Imsieke, Gerrit, le-tex 22 Feb '22

22 Feb '22

I was faintly hoping that a sequence will be magically output as multipart/mixed with appropriate part content types when I annotate it with %rest:produces("multipart/mixed"): declare %rest:GET %rest:path("/test") %rest:produces("multipart/mixed") function my:test-multipart() { (<ok/>, json:serialize(map{'some':'json'})) }; But the response header says "Content-Type: application/xml", and the result is <ok/> { "some": "json" } Are there some other functions or annotations or rest:response elements that I can use in order to get multipart/mixed correctly assembled? Bonus: With the possibility to add file names to each chunk? Gerrit

2 1

REST API
by Paul L. Merchant Jr. 22 Feb '22

22 Feb '22

Hi everyone, I have a couple of questions about BaseX's REST API and how it executes stored XQuery. I'm evaluating how much effort would be involved in porting an existing XQuery code base to BaseX. The existing code is a directory of XQuery scripts and modules that are called over a REST interface. They aren't RESTXQ aware, although with some work they could be converted if needed. (I'd prefer to do as little modification as possible initially, however.) I've installed BaseX 9.6.4 on Mac OS in a directory I'll call BASEX_HOME. In $BASEX_HOME/.basex, I have WEBPATH set to $BASEX_HOME/webapp (the default) and RESTPATH = "" (the default). I also set an appropriate user and password. I've placed the following two scripts in $BASEX_HOME/webapp: -------- t1.xq -------- import module namespace t2="t2" at "t2.xq"; t2:response() -------- t2.xq -------- module namespace t2="t2"; declare function t2:response() as xs:string { "Hello World!" }; -------- When I run the first script using curl, I get an error back that the second cannot be located in the $BASEX_HOME directory not $BASEX_HOME/webapp as I expected: $ curl http://localhost:8984/rest?run=t1.xq Stopped at ., 3/1: [XQST0059] Could not retrieve module: $BASEX_HOME/t2.xq. Is this the expected behavior, or is BaseX resolving the path to module t2 incorrectly? In any case given this behavior, if I want to point RESTPATH outside the BaseX installation to a much larger library of XQuery, how can I share modules within the library without having to use absolute paths for everything? My second question is simply, is it possible to set a cookie in response to a REST (not RESTXQ) response? Thanks! -- Paul

3 3

Adding lots of files in a defined order
by Jonathan Robie 21 Feb '22

21 Feb '22

I have 929 files with names like these: 01-Gen-001.xml 01-Gen-002.xml 01-Gen-003.xml 01-Gen-004.xml 01-Gen-005.xml 01-Gen-006.xml 01-Gen-007.xml 01-Gen-008.xml 01-Gen-009.xml 01-Gen-010.xml I would like to add them to a new database in the same order that I see if I do $ ls in that directory, without using XInclude to put them inside another node. Is there an easy way to do that? Jonathan

2 2

Using %unit:after to verify database creation
by Eliot Kimber 20 Feb '22

20 Feb '22

I’m working out my new-and-improved orchestration mechanism using the job-chaining approach and I’m not getting the unit test behavior I was expecting. My current unit test is: declare %unit:after("test:runMultipleJobs") function test:afterRunMultipleJobs() { ( for $database in $test:dbs return unit:assert(db:exists($database), 'Expected database ' || $database || ' to exist') ) }; declare %unit:test function test:runMultipleJobs() { let $jobs as xs:string* := ( for $database in $test:dbs return dbadmin:makeJob('dbadmin:createDatabase', $database) ) let $jobIds as xs:string* := orch:runJobs($jobs) return ( unit:assert-equals(count($jobIds), 3, 'Expected 3 job IDs') ) }; When I run this test I expect it to either pass (because my job-runner caused all the databases to be created) or fail on “db2” (because I can see that the first database db1 is created by inspecting the Databases via the DBA console after running the test, but the others are not). What I’m getting, however, is that the assert in the test:afterRunMultipleJobs() fails on db1, meaning it doesn’t see that that database has in fact been created. This must be an issue with when the test runner can actually see that the database exists or a timing issue with when the job runs relative to the unit test code. I did experiment with having both the base unit test and the after function sleep before doing the existence assertion but that seemed to have no effect, which suggests that the runner simply can’t see databases created by the tests it runs. Is the inability to see the newly-created databases simply a limitation in the test runner or am I doing something wrong? Thanks, 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 9

Database 'name' is being updated, or update was not completed.
by Jonathan Robie 20 Feb '22

20 Feb '22

I have some processes that update databases, taking a long time to do so. I would like other processes to be able to read the last complete snapshot while this happens. Is that possible? Jonathan

2 2

Understanding Job Execution and Verification
by Eliot Kimber 18 Feb '22

18 Feb '22

I’m working out an “orchestration” module that manages running jobs that are functions from different modules, where the orchestrator dynamically constructs an XQuery that runs the specified function with the specified string parameters. I’m running jobs through a function that takes a sequence of jobs (XQuery strings) and calls jobs:eval() on each one in sequence: declare function orch:runJobs($jobs as xs:string*) as xs:string* { let $debug := util:logToLog('Have ' || count($jobs) || ' jobs:') let $debug := prof:dump($jobs) return for $job at $i in $jobs let $debug := (util:logToLog('Running job'), prof:dump($job)) return try { jobs:eval($job, (), map{'cache' : true(), 'log' : 'Running job: ' || $job}) } catch * { util:logToLog('orch:runJobs', ('Error running job ' || $err:code || ': ' || $err:description) ) } }; What I’m trying to achieve is a sequence of jobs run such that job 2 does not start until job 1 is finished and I don’t think that’s actually what I’m producing here. Most of these jobs are manipulating databases: dropping, creating, renaming, copying, etc., but a few will be longer-running—constructing indexes and whatnot. My question: What is the (best?) way to produce a sequence of jobs where each job waits for the preceding? Is it simply to do a jobs:wait(jobs:eval()) on each job or is there something better? Or am I missing a simpler way to perform this kind of job orchestration? Thanks, 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 3

specifying explict path to database
by Eric Levy 17 Feb '22

17 Feb '22

I am targeting an application that would incorporate BaseX databases into larger projects, each following a standard file tree. As such, it is necessary for the application to interact with a database given by some path relative to the working directory. The application is structured as to provide query files through the command line in a standalone invocation. The documentation explains that the normal case is to resolve a directory in which all databases are stored, with each separate database in a sub-directory having the same name as the database. What support is available for the case in which the application must resolve the path to the database independent of the normal rules?

2 2

Syntax-checker
by Ben Engbers 17 Feb '22

17 Feb '22

Hi, After I had formulated the query in Basex-GUI, I tried to execute the same multi-line query in R/RbaseX. Nada ;-( In R, one can use the paste function to concatenate strings. I use this function to build a string which is passed to the RbaseX-client. Example: Stmt_1 <- "for $i in 1 to 2 return $i" => OK Stmt_2 <- paste0("for $i in 1 to 2", => ERROR "return $i") It took 2 days of debugging before I found the error. Stmt_2 is concatenated to "for $i in 1 to 2return $i" and it is clear that this can't be executed. Instead of using the "paste0"-function I should have used "paste" which introdus a space between the strings to be concatenated. This works fine. My problem is that the server/my client does not give an error-message. And this leads me to following question: It would be helpfull if the syntax for the XQuery statement was checked before sending to the server. Where in the BaseX sources can I find the code for XQuery checking? Is it possible to translate tist code into R or is that way to difficult? Ben Engbers

4 4

Namespaces, replace, and performance
by Jonathan Robie 17 Feb '22

17 Feb '22

I wrote a query that replaced roughly 470153 text nodes with an element that had two namespaces declared on it. It ran for several hours. I didn't need or want these namespaces, they were in the original source. I rewrote the query to remove the namespaces from these nodes before replacing the text node. I don't know how quickly it ran, but I think not more than 15 minutes, the first time I checked on it, it had already finished. I made no other changes to the query. Is this surprising? Or is this a general lesson about performance and updates that add namespaces low down in the hierarchy? Or ... Jonathan

2 1

Long-running queries: command line or BaseX Gui?
by Jonathan Robie 17 Feb '22

17 Feb '22

Do long-running queries run faster from the command line, the GUI, or about the same from either? Jonathan

3 2

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

BaseX-Talk