How to browse/cleanup components/assets in groovy script with H2 instead of OrientDB (e.g. without StorageFacet)

Sonatype Nexus Repository
, ,
1

Hi,

There are many examples how to browse/cleanup components/assets via groovy scripts like so - GitHub - PhilSwiss/nexus-cleanup: Remove old releases from a Maven2-repository on Nexus3 - but after switching to H2 instead of OrientDB with nexus.datastore.enabled=true I receiving error unable to resolve class org.sonatype.nexus.repository.storage.StorageFacet

Please explain me why I can’t use StorageFacet with H2 and how to browse components/assets in this case?

2

The internals for SQL were largely re-written.

Each repository will have a ContentFacet attached which can provide access to components and assets. Note that interacting with the internals of Nexus directly may not result in appropriate book keeping occurring.

3

Thank you!

I’m trying to use StorageFacet in similar way:

import org.sonatype.nexus.repository.storage.ContentFacet

def response = []

def repo = repository.repositoryManager.get(args)
def tx = repo.facet(ContentFacet).txSupplier().get()

try {
    tx.begin()
    tx.browseAssets(tx.findBucket(repo)).each { asset -> response << "asset:${asset.name()}" }
    tx.browseComponents(tx.findBucket(repo)).each { component -> response << "component:${component.name()}" }
}
finally {
    tx.close()
}

return response.join('\n')

But I’m receiving error even on import:

$ ./run.py browse-content.groovy 
UPDATE: <Response [204]> 
RUN: <Response [400]> {
  "name" : "browse-content",
  "result" : "org.codehaus.groovy.control.MultipleCompilationErrorsException: startup failed:\nScript6.groovy: 1: unable to resolve class org.sonatype.nexus.repository.storage.ContentFacet\n @ line 1, column 1.\n   import org.sonatype.nexus.repository.storage.ContentFacet\n   ^\n\n1 error\n"
}

run.py is simple python script to update and run any groovy script via REST API

So, why ContentFacet can’t be imported and how to use it correctly?

4

I will suggest something like this below

import org.sonatype.nexus.repository.content.facet.ContentFacet

chunkSize = 100

repo = repository.getRepositoryManager().get(args)
content = repo.facet(ContentFacet.class)
response = []

for (page = content.assets().browse(chunkSize, null);
     !page.isEmpty();
     page = content.assets().browse(chunkSize, page.nextContinuationToken())) {
  page.each(asset -> response << "asset:${asset.path()}" } 
}

return response.join('\n')
5

Something to be careful with - if your page is ever empty the continuation token will be null and you may infinitely loop.

This class contains some helpers: nexus-public/components/nexus-common/src/main/java/org/sonatype/nexus/common/entity/Continuations.java at main · sonatype/nexus-public · GitHub

1 Like
6

Thanks! All above works good enough

7

Hi @eugene.prokopiev @wiget.
If you know the way to correctly remove docker components with H2 groovy API, could you please share it.