On Mon, Aug 29, 2016 at 07:07:09AM +0200, Thomas Herve wrote:
> On Sun, Aug 28, 2016 at 11:58 PM, Steven Hardy <sha...@redhat.com> wrote:
> > Hi all,
> >
> > I have a need to merge a list of maps of lists:
> >
> > heat_template_version: 2016-10-14
> >
> > outputs:
> > debug:
> > value:
> > yaql:
> > # dict(vms=>dict($.vms.select([$.name, $])))
> > expression: dict($.data.l.select([$.keys().toList()[0],
> > $.values().toList()[0]]))
> > data:
> > l:
> > - a: [123]
> > - b: [123]
> > - a: [456]
> >
> >
> >
> > I want to end up with debug as:
> >
> > a: [123, 456]
> > b: [123]
> >
> > Perhaps we need a map_deep_merge function, but can this be done with yaql?
> >
> > I suspect it can, but can't currently figure out how the assignment to the
> > intermediate "a" value is supposed to work, any ideas on the cleanest
> > approach appreciated!
>
> I believe you don't need the intermediate value, and can rely on what
> you'd do in Python with setdefault:
>
> dict($.groupBy($.keys().toList()[0], $.values().toList()[0][0]))
>
> ought to work, I believe?
So, as it turns out, my example above was bad, and groupBy only works if
you have a list of maps with exactly one key, we actually need this:
# Example of tripleo format
# We need an output of
# "gnocchi_metricd_node_names": ["overcloud-controller-0"]
# "tripleo_packages_node_names": ["overcloud-controller-0",
"overcloud-compute-0"]
# "nova_compute_node_names": ["overcloud-compute-0"]
debug_tripleo:
value:
yaql:
expression: dict($.data.l.groupBy($.keys().toList()[0],
$.values().toList()[0][0]))
data:
l:
- "gnocchi_metricd_node_names": ["overcloud-controller-0"]
"tripleo_packages_node_names": ["overcloud-controller-0"]
- "nova_compute_node_names": ["overcloud-compute-0"]
"tripleo_packages_node_names": ["overcloud-compute-0"]
So, I'm back to wondering how we make the intermediate assignement of
tripleo_packages_node_names
Thanks for the ideas - I think this probably does highlight the value of a
map_deep_merge function
so perhaps that's something we can consider for ocata :)
Thanks!
Steve
__________________________________________________________________________
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
