[jira] [Commented] (TINKERPOP-1539) Create a ComplexTraversalTest with crazy nested gnarly traversals.

["Daniel Kuppitz (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/TINKERPOP-1539?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15710704#comment-15710704 ] 

Daniel Kuppitz commented on TINKERPOP-1539:
-------------------------------------------

[~okram] Do you prefer to use the good old toy graphs for these tests or should we create custom graphs (in the worst case one per test)? I would prefer if we don't limit ourselves to the toy graphs (it's actually just the GD graph when we talk about complex queries). Every now and then I end up writing real complex queries on the mailing list; these queries require graphs with certain structures that we mostly won't find in our toy graphs.

> Create a ComplexTraversalTest with crazy nested gnarly traversals.
> ------------------------------------------------------------------
>
>                 Key: TINKERPOP-1539
>                 URL: https://issues.apache.org/jira/browse/TINKERPOP-1539
>             Project: TinkerPop
>          Issue Type: Improvement
>          Components: process, test-suite
>    Affects Versions: 3.2.3
>            Reporter: Marko A. Rodriguez
>            Assignee: Daniel Kuppitz
>
> Our {{ProcessSuite}} has numerous tests verifying the semantics of the various steps. Unfortunately, these tests are on simple traversals focused on exposing the step in question in isolation.
> It would be good to add {{ComplexTraversalTest}} to the {{ProcessSuite}} which has traversals over the Grateful Dead graph (for complexity reasons) doing:
> 1. Numerous nests.
> 2. Match/Select/Where complexities.
> 3. Global side-effect access and unrolling and injecting.
> 4. ... just a bunch of nasty stuff.
> This will give us much more confidence as we add more strategies and potentially, mess up our algebra which isn't exposed by the simple "flat'-traversals we current test with.
> [~dkuppitz] --- would you be interested in doing this?



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)