fix OGC API coverages subset and datetime not forwarded

[fmigneault]

Fixes needed for OGC TB-20 TIEs.

Some parameters are not forwarded to the coverage endpoint, causing requests to be refused by servers that consider the subset to be too large to return reasonably.

[geographika]

@fmigneault @tomkralidis - the failing test is due to making an external request in the offline testing. Adding the decorator @pytest.mark.online to test_ogcapi_build_url should fix this:

OWSLib/tests/test_ogcapi_records_pycsw.py

Line 76 in 8c7effc

def test_ogcapi_build_url(path, expected):

It looks like creating any OGC API class will make an HTTP directly on __init__ - I think this differs from WMS/WFS where the class can be created and then HTTP requests made with a different call.

OWSLib/owslib/ogcapi/__init__.py

Line 68 in 8c7effc

response = http_get(self.url, headers=self.headers, auth=self.auth).json()

[tomkralidis]

@fmigneault @tomkralidis - the failing test is due to making an external request in the offline testing. Adding the decorator @pytest.mark.online to test_ogcapi_build_url should fix this:

OWSLib/tests/test_ogcapi_records_pycsw.py

Line 76 in 8c7effc

def test_ogcapi_build_url(path, expected):

+1

It looks like creating any OGC API class will make an HTTP directly on __init__ - I think this differs from WMS/WFS where the class can be created and then HTTP requests made with a different call.

OWSLib/owslib/ogcapi/__init__.py

Line 68 in 8c7effc

response = http_get(self.url, headers=self.headers, auth=self.auth).json()

Not sure. WMS support always does a GetCapabilities on init.

[fmigneault]

I've added more tests (all passed) for the various kw updates related to OGC API coverages.
Failing tests are unrelated to this PR (records and csw tests).

[fmigneault]

@tomkralidis
Thanks for merging.
Are there plans for a new release in a near future?

[tomkralidis]

@tomkralidis Thanks for merging. Are there plans for a new release in a near future?

Probably a good time for 0.31.1 0.32.1 then (@geographika thoughts?).

[geographika]

@tomkralidis @fmigneault - yes, I can make a new release this week. There are quite a few changes, but all appear so far to be bug fixes, so it would be 0.32.1 - although if #963 is merged I think it would bump it to 0.33.0 as it is a change in functionality.

[tomkralidis]

@tomkralidis @fmigneault - yes, I can make a new release this week. There are quite a few changes, but all appear so far to be bug fixes, so it would be 0.32.1 - although if #963 is merged I think it would bump it to 0.33.0 as it is a change in functionality.

Sorry, typo, yes 0.32.1.

[fmigneault]

Hi @geographika
Doing a follow-up about this. The other PR doesn't seem to be receiving the requested feedback. Is it being worked on or should this be released separately?
Thanks

[geographika]

@fmigneault - no #963 not yet merged, but this pull request has been released in https://github.com/geopython/OWSLib/releases/tag/0.32.1 and is available at https://pypi.org/project/OWSLib/
You should have got a ping from the release as you're tagged (and also posted to the mailing lists https://lists.osgeo.org/mailman/listinfo/owslib-users and https://lists.osgeo.org/mailman/listinfo/owslib-devel).

[fmigneault]

@geographika Thanks