epimorphics / appbase Goto Github PK

ResourceView is a simpler, more controllable way to handle configurable, locally cached views of possibly-remote resources.

Could we move the remaining functionality of WNode into it and deprecate the WSource/WNode machinery?

Would mean updating minibrowse and registry-dclib as well as the browse component of DMS.

Test for detecting action timeouts fails on fast machine under manual maven but works (mostly) under eclipse.

Is this a race condition in the test or a genuine issue?

To avoid needing to run the loop in #13, it would be handy if a resultset had a count() operation.

Please provide a copyright statement and a license

Suspicious that sometimes ConfigMonitor stops seeing changes if you don't have fileSampleLength set. All attempts to reproduce or debug have failed so far.

Write a randomized soak test to try to confirm.

Might relate to unexplained test failures on ijd machine.

• will need care to handle nested directories

ResourceViews look useful and may be a better approach to WNode.

Document and figure out how to position them.

Test under tomcat8

... once it is stable

Create a progress monitor that can aggregate progress across a set of sequential actions

Filter all incoming parameters injected into the velocity environment for XSS prevention

Get more exception traces in (passing) test runs than seems reasonable.

Review to make sure these are indeed all correct cases and if possible reduce the noise.

Have added update capabilities to the SparqlSource abstraction - need to bring the wiki documentation up to date

Can pass environment settings to the scripts from individual action specs but in some cases it would be more robust to have the ability to inject a set of global settings.

I have the following Velocity code, which is passed a WResultSetWrapper instance:

#set( $hasResultsBefore = $results.hasNext())
#set( $nResults = 0)
#foreach ($r in $results.copy())
  #set( $nResults = $nResults + 1)
#end
#set( $hasResultsAfter = $results.hasNext())

<p>results = $results<br />
hasResultsBefore = $hasResultsBefore <br />
hasResultsAfter = $hasResultsAfter
</p>

Giving:

results = com.epimorphics.appbase.data.impl.WResultSetWrapper@54cf6de3
hasResultsBefore = true 
hasResultsAfter = false

Expected outcome is that $hasResultsAfter and $hasResultsBefore are the same, but the observed result suggests that the loop is iterating over the base result set not a copy.