[
https://issues.apache.org/jira/browse/TINKERPOP3-465?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stephen mallette updated TINKERPOP3-465:
----------------------------------------
Description:
For profiling purposes, it would be great if there were a way to retain the
lineage from the original query to the one after applying the strategies.
For instance, suppose a user issues the query:
{code}
g.V.has('name','Marko').out('knows').name
{code}
then, after applying the strategies on TinkerGraph,
{code:g}.V.has('name','Marko'){code} will be compressed into a single step
(index call). After that we apply the profiling and so we record how much time
that index call took, but we won't be able to relate that information back to
the original query to say {code:g}.V.has('name','Marko'){code} took X
milliseconds.
was:
For profiling purposes, it would be great if there were a way to retain the
lineage from the original query to the one after applying the strategies.
For instance, suppose a user issues the query:
{code:groovy}
g.V.has('name','Marko').out('knows').name
{code}
then, after applying the strategies on TinkerGraph,
{code:g}.V.has('name','Marko'){code} will be compressed into a single step
(index call). After that we apply the profiling and so we record how much time
that index call took, but we won't be able to relate that information back to
the original query to say {code:g}.V.has('name','Marko'){code} took X
milliseconds.
> Retain strategy application lineage
> -----------------------------------
>
> Key: TINKERPOP3-465
> URL: https://issues.apache.org/jira/browse/TINKERPOP3-465
> Project: TinkerPop 3
> Issue Type: Improvement
> Components: process
> Reporter: Matthias Broecheler
>
> For profiling purposes, it would be great if there were a way to retain the
> lineage from the original query to the one after applying the strategies.
> For instance, suppose a user issues the query:
> {code}
> g.V.has('name','Marko').out('knows').name
> {code}
> then, after applying the strategies on TinkerGraph,
> {code:g}.V.has('name','Marko'){code} will be compressed into a single step
> (index call). After that we apply the profiling and so we record how much
> time that index call took, but we won't be able to relate that information
> back to the original query to say {code:g}.V.has('name','Marko'){code} took X
> milliseconds.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
