URL params are not handled cleanly when defined

[rickheil]

In the current SimpleMDM.py, a starting_after=0 is force-applied to each URL in _get_data, and any passed parameters tacked on the end. When overriding the starting value (e.g. for Logs, found while testing #25) this ends up with two start values in the URL, eg: https://a.simplemdm.com/api/v1/logs?limit=100&starting_after=0&starting_after=aldskjrgawop4fjasdlkfjasfw93f4war

Obviously the SimpleMDM service is honoring the second one of these because it works, but we should add some more intelligent parsing of the parameters and use the existing start_id value if specific params are passed from the various functions.

[carlashley]

Probably the easiest way to ensure params are appropriately dealt with is to utilise the ability of requests to parse the params=dict_obj out into an encoded URL in the same manner that urllib.parse.urlencode does.

For example:

    def paginate(self, url: str, limit: int = 100, starting_after: str = "0", **kwargs) -> Optional[List[Any]]:
        """Paginate over results from API query.
        :param url: url string to perform pagination on
        :param limit: maximum number of objects returned in response when paginating
        :param starting_after: initial ID value to use for starting the pagination from,
                               the default is '0' so as to start from the first page in the
                               pagination process"""
        result = list()

        if not kwargs:
            kwargs = dict()

        kwargs["limit"] = limit
        kwargs["starting_after"] = starting_after
        has_more = True

        while has_more:
            req = self.session.get(self._urljoin(url), params=kwargs, timeout=self.timeout)

            if has_more and req.json() and req.json().get("data"):
                result.extend(req.json()["data"])
                kwargs["starting_after"] = str(req.json()["data"][-1].get("id"))  # Update next ID value
                has_more = req.json().get("has_more", False)
            else:
                if req.json().get("data"):
                    result.extend(req.json()["data"])
                    has_more = req.json().get("has_more", False)
                else:
                    break

        return result

In the case of querying devices using list_all from the API, the URL would end up as such using the default arg values provided to the paginate method:
https://a.simplemdm.com/api/v1/devices?limit=100&starting_after=0

Any methods that then call on the paginate method could then subsequently also have an optional starting_after arg passed through.

[MagerValp]

I think this should finally be fixed now with the merge of PR #59