Collector skips metrics that have no sources, but do have one or more issues

[fniessink]

This method checks whether the user has specified all mandatory parameters for all sources. Even if the metric has one or more issues, if not all mandatory parameters are present, the metric is skipped. If there are no sources at all, the metric is also skipped.

    def __can_collect(self, metric: dict) -> bool:
        """Return whether the user has specified all mandatory parameters for all sources."""
        sources = metric.get("sources", {})
        for source in sources.values():
            parameters = self.data_model.get("sources", {}).get(source["type"], {}).get("parameters", {})
            for parameter_key, parameter in parameters.items():
                if (
                    parameter.get("mandatory")
                    and metric["type"] in parameter.get("metrics")
                    and not source.get("parameters", {}).get(parameter_key)
                    and not parameter.get("default_value")
                ):
                    return False
        return bool(sources)

[fniessink]

Move the parameter check to SourceCollector.collect(), and if it fails, skip collecting and parsing the source responses.

    async def collect(self) -> SourceMeasurement:
        """Return the measurement from this source."""
        if self.__all_mandatory_parameters_ok(...):
            responses = await self.__safely_get_source_responses()
            measurement = await self.__safely_parse_source_responses(responses)
            measurement.api_url = responses.api_url
            measurement.landing_url = await self.__safely_parse_landing_url(responses)
        else:
            measurement = SourceMeasurement()
        return measurement