Fixed format test

[BaptisteVandecrux]

Bug origin: first we make all the format .lower() and then we test whether it matches with "STM"
Fixed now.

It is making the test of new logger files fail on AWS-L0: https://geusgitlab.geus.dk/glaciology-and-climate/promice/aws-l0/-/pipelines/7178/failures

[BaptisteVandecrux]

I'm actually unsure about what this part is supposed to do:

pypromice/src/pypromice/process/aws.py

Lines 94 to 102 in e445661

	formats = {dataset.attrs["format"].lower() for dataset in self.L0}
	if "raw" in formats:
	self.format = "raw"
	elif "stm" in formats:
	self.format = "STM"
	elif "tx" in formats:
	self.format = "tx"
	else:
	raise ValueError(f"Unknown formats from l0 datasets: {','.join(formats)}")

It takes all the formats from all the L0 datasets and then applies a general format attribute to the AWS object that is :

• raw if at least one raw L0 is in the AWS object
• STM if the AWS object doesn't contain any raw L0 dataset but at least one STM L0 dataset
• TX if the AWS object doesn't contain any raw or STM L0 dataset but at least one TX L0 dataset

Do we need to have an overall format for the AWS object?
Doesn't make sense that it should have an overal raw format if it composed of a mix of raw, STM and TX files.