July 2021 - BaseX-Talk - mailman.uni-konstanz.de

Preserving whitespace for XQuery Updates
by Jonathan Robie 16 Jul '21

16 Jul '21

I am doing some transformations of datasets, then submitting pull requests to upstream sources on GitHub. For instance, today I am inserting some attributes, but I may be restructuring in various ways or enhancing data in various ways. To make upstreams happy, I need to be disciplined about not changing whitespace. What do I have to do? Is it sufficient to preserve whitespace when importing, do an XQuery update, and export, or can that change whitespace beyond what the update operations explicitly say? Thanks! Jonathan

2 10

BaseX 9.5.2 simple query has slow Compilation time
by arcogabbo 13 Jul '21

13 Jul '21

Hi, I admit this is my first experience with BaseX and XQuery language, but this seems weird and i can't find anything on docs or issues related to this. I created a random dataset of calls and i wanted to query this dataset using XQuery, however it turns out compile time takes too long for a relatively simple query. The dataset can be found here <https://gist.github.com/arcogabbo/8042d08b1feb5c860f86ff46d4333e84> . The query is: doc("basex/etc/dataset25k.xml")/users/user[number= "605 6230540" and calls/call/location/@city = "Sandy Springs"]/fullName Timing: - Parsing: 0.61 ms - Compiling: 382.43 ms - Evaluating: 2.1 ms - Printing: 0.14 ms - Total Time: 385.28 ms Compiling: - rewrite fn:doc(uri) to document-node() item: doc("basex/etc/dataset25k.xml") -> document { "file:///home/arcogabbo/Scrivania/basex/etc/dataset25k.xml" } - rewrite to predicate: ((number = "605 6230540") and (calls/call/location/@city = "Sandy Springs")) - apply attribute index for "Sandy Springs" Even if i try a simple query on the example dataset of the application(factbook.xml) the compilation time takes too long, for example: Query: doc("basex/etc/factbook.xml")/mondial/country[@name="Norway"]/province/city Timing: - Parsing: 0.34 ms - Compiling: 82.29 ms - Evaluating: 1.05 ms - Printing: 1.49 ms - Total Time: 85.17 ms Compiling: - rewrite fn:doc(uri) to document-node() item: doc("basex/etc/factbook.xml") -> document { "file:///home/arcogabbo/Scrivania/basex/etc/factbook.xml" } - apply attribute index for "Norway" I also tried BaseX version 9.5 and on another computer, yet compilation times are approximately the same. Any help is appreciated.

2 1

Can the full text tokenizer be instructed not to tokenize at hyphens?
by Imsieke, Gerrit, le-tex 13 Jul '21

13 Jul '21

A customer wants to include composite terms such as 'third-generation' as single tokens so that they may be offered in a completion list. I don’t think this is configurable, or is it? Gerrit

2 2

Fuzzy full-text query
by Tim Thompson 06 Jul '21

06 Jul '21

Hello, I am trying to match a local dataset against entries in a full-text index. I ran into the following case and was confused about the fuzzy querying performance: <x>Baxter, Richard, 1615-1691</x>[text() contains text "Baxter, Richard, 615-1691" using fuzzy] The BaseX documentation says that the default fuzzy search will preserve a minimum of 1 error. However, the query above returns no result. If I add the "errors" parameter with a value of 1, it does work: <x>Baxter, Richard, 1615-1691</x>[text() contains text "Baxter, Richard, 615-1691" using fuzzy 1 errors] Is this expected behavior? Thanks in advance, Tim -- Tim A. Thompson Librarian for Applied Metadata Research Yale University Library

2 1

automatically running an unknown number of queries on a variable content set
by Graydon Saunders 06 Jul '21

06 Jul '21

Hello! So the idea is to -- via some automatic process, so "the command line" -- load a content set into BaseX and then run some arbitrary list of queries on that content set, returning an aggregated report. It's important that it's easy to add a query; we expect to be adding queries as we find new ways for the content to be inappropriate. (We're guaranteed it's valid; we are not guaranteed it's sensible, in the "and just how do you expect to publish forty seven levels of nested lists?" sense.) My first thought is to have a directory of individual queries and have the main query walk through them with xquery:eval(), but there might be a better way. Anyone got any suggestions? Thanks! Graydon

3 2

Increase automatic logout time
by Josselin Morvan 02 Jul '21

02 Jul '21

Hi everyone, We are using BaseX for the development of our web application and we are having issues with automatic logouts. We have developed forms with Xforms that allow us to process historical documents and we use Basex to manage our data directly in XML. Recently we put our application online to be able to share our data and work with the whole team more easily. For the management of the users and the permissions, we relied on the BaseX doc (https://docs.basex.org/wiki/Permissions <https://docs.basex.org/wiki/Permissions>). Globally everything works fine, except that we encounter a kind of automatic logout of the users maybe about 1 hour after their login, maybe even less, it’s hard to say… In fact, our forms are quite big and it often takes more than an hour to finish processing a single historical resource, and when the user save his work, if he is logged out, he loses all what he’ve done. Would you know a way to increase this logout time? On our local BaseX instance, we tried changing the database options, including Keepalive (https://docs.basex.org/wiki/Options#KEEPALIVE) We changed the value 600 to 60, just to make sure it was this option that needed to be changed, but it didn't seem to have any effect… For the moment we set up an automatic backup system, but this solution introduces other issues and is therefore not ideal. Thank you in advance for your help!! Take good care, Josselin.

3 2

Copy/modify, Update, or file:append - help appreciated
by Bridger Dyson-Smith 02 Jul '21

02 Jul '21

Hi all - I'm trying to keep track of some HTTP POST results in a file, something akin to the following: file:write-text( "/tmp/bridger-test.csv", "book" || "," || "title" || "," || "sip" || out:nl() ), for $i in 1 to 20 (: some http:send-request stuff :) return( prof:sleep(250), if ($request[1]//@status/data() = 201) then file:append( "/tmp/bridger-test.csv", "book: " $i || "," || "title: " $i || "," || "" ) else "something went wrong with the request; the response was: " || $request[1]//@status/data() I have something that kinda works, but I'd like it to be better XQuery, avoid the two step create-file-then-append, and leverage more of built-in awesomeness in the language. I'm sure that I need to have a better understanding of the Update syntax, and/or %updating annotation, but would someone be willing to give some hints and/or suggestions for a better way to handle situations like this? Thanks in advance for your time and trouble. Best, Bridger

2 1