fix: query-patterns-should-not-be-separated DEV-2473

[seakayone]

Pull Request Checklist

Task Description/Number

e351592 introduces a changed behaviour in how query patterns are transformed/optimized.

The create slow down reported in DEV-2473 is caused by this.
The gravsearch query which is used for retrieving the information of the resource to which a value should be created is containing a different optional block whilst the rest of the query still was the same.

The difference in runtime is enormous:
407 results in 0.083 seconds (fixed, old behaviour):

OPTIONAL {
  ?standoffNode <http://www.knora.org/ontology/knora-base#standoffTagHasInternalReference> ?targetStandoffTag .
  ?targetStandoffTag <http://www.knora.org/ontology/knora-base#standoffTagHasOriginalXMLID> ?targetOriginalXMLID .
}

and

88198 results in 5.43 seconds (faulty behaviour):

OPTIONAL {
  ?standoffNode <http://www.knora.org/ontology/knora-base#standoffTagHasInternalReference> ?targetStandoffTag .
}
OPTIONAL {
  ?targetStandoffTag <http://www.knora.org/ontology/knora-base#standoffTagHasOriginalXMLID> ?targetOriginalXMLID .
}

In the faulty example if a match is found for the first triple pattern, it will return results that include that match, even if there is no match for the second triple pattern, and vice versa.

Basic Requirements

Please check if your PR fulfills the following requirements:

• Tests for the changes have been added (for bug fixes / features)
• Docs have been added / updated (for bug fixes / features)

PR Type

What kind of change does this PR introduce?

• fix: represents bug fixes
• refactor: represents production code refactoring
• feat: represents a new feature
• docs: documentation changes (no production code change)
• chore: maintenance tasks (no production code change)
• test: all about tests: adding, refactoring tests (no production code change)
• other... Please describe:

Does this PR introduce a breaking change?

• Yes
• No
• Maybe (not 100% sure => check with FE)

Does this PR change client-test-data?

• Yes (don't forget to update the JS-LIB team about the change)
• No

[codecov]

Codecov Report

Patch coverage has no change and project coverage change: +4.33% 🎉

Comparison is base (12402f3) 18.00% compared to head (ded9718) 22.33%.
Report is 33 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2786      +/-   ##
==========================================
+ Coverage   18.00%   22.33%   +4.33%     
==========================================
  Files         281      242      -39     
  Lines       28899    23394    -5505     
==========================================
+ Hits         5202     5225      +23     
+ Misses      23697    18169    -5528     

see 93 files with indirect coverage changes

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[linear]

DEV-2473 SAVE PROCESS SLOW

From Stefan Münnich (Webern):

Description

Saving a new entry or even a small change (typo) sometimes takes a long time (more than 10 seconds).

https://dasch.atlassian.net/jira/servicedesk/projects/DSQ/queues/custom/3/DSQ-221

The resource to reproduce the bug with the label "Guido Adler tritt in den Ruhestand": https://ark.dasch.swiss/ark:/72163/1/0806/9pCn9mtsTmSzenpSL4YIcQg.20230806T085533017083949Z

Property label: "andere Quellen"

Screenshot from Flavie's console log (Firefox, staging server):

Screenshot 2023-08-07 at 10.20.04.png

[BalduinLandolt]

Thanks for spotting and fixing!

At some point we had a convention to name tests that use ZIO-Test ...ZSpec.scala - not sure if we still do that? But if we do, you may want to rename the files.

[seakayone]

Thanks for spotting and fixing!

At some point we had a convention to name tests that use ZIO-Test ...ZSpec.scala - not sure if we still do that? But if we do, you may want to rename the files.

I am happily ignoring this naming convention. Eventually all of our spec should be migrated to zio-test anyway. Also in the webapi/test source set zio-test should be the only available test framework.