I've made good progress with this, things work etc, I'm just wrestling
with SQL queries. One thing I noticed is that the current /api/patches/
list seems to go out to the database for the project linkname for every
patch:
...
/opt/pyenv/versions/3.8.1/lib/python3.8/site-packages/rest_framework/serializers.py
in to_representation(674)
return [
/opt/pyenv/versions/3.8.1/lib/python3.8/site-packages/rest_framework/serializers.py
in <listcomp>(675)
self.child.to_representation(item) for item in iterable
/home/patchwork/patchwork/patchwork/api/patch.py in to_representation(110)
data = super(PatchListSerializer, self).to_representation(instance)
/home/patchwork/patchwork/patchwork/api/base.py in to_representation(90)
data = super(BaseHyperlinkedModelSerializer, self).to_representation(
/opt/pyenv/versions/3.8.1/lib/python3.8/site-packages/rest_framework/serializers.py
in to_representation(526)
ret[field.field_name] = field.to_representation(attribute)
/home/patchwork/patchwork/patchwork/api/embedded.py in to_representation(57)
return self._Serializer(context=self.context).to_representation(data)
/home/patchwork/patchwork/patchwork/api/base.py in to_representation(90)
data = super(BaseHyperlinkedModelSerializer, self).to_representation(
/opt/pyenv/versions/3.8.1/lib/python3.8/site-packages/rest_framework/serializers.py
in to_representation(526)
ret[field.field_name] = field.to_representation(attribute)
/opt/pyenv/versions/3.8.1/lib/python3.8/site-packages/rest_framework/fields.py
in to_representation(1873)
return method(value)
/home/patchwork/patchwork/patchwork/api/embedded.py in get_web_url(81)
return request.build_absolute_uri(instance.get_absolute_url())
/home/patchwork/patchwork/patchwork/models.py in get_absolute_url(767)
kwargs={'project_id': self.project.linkname}) + (
I see about 22 of these, perfectly identical, on the existing code - and
the number goes up when I include the new code which displays more
embedded patches.
Thoughts?
Regards,
Daniel
Stephen Finucane <step...@that.guru> writes:
> On Sat, 2020-01-25 at 00:20 +1100, Daniel Axtens wrote:
>> I've spent some time tonight bashing this out, based on Stephen's API
>> design. I have got display to work, just working on writing now.
>>
>> Stephen, would you be able to write the API spec if I can send an
>> otherwise working implementation? I think the ramp-up time for me on
>> that might be a big unnecessary delay...
>
> Sure thing.
>
> Stephen
>
>> Kind regards,
>> Daniel
_______________________________________________
Patchwork mailing list
Patchwork@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/patchwork
