Improvements for matchms cleaning pipeline

[niekdejonge]

When working on the mathms cleaning pipeline, these issues were not implemented but might be good to implement before publishing.

General matchms changes:

• We still have filters that are already run during loading of spectra. Maybe remove the running of these filters from Metadata? Because this feels inconsistent.
• Add automatic test that checks that the tests use the same naming and folder structure? (if feasible)(now the naming is often not updated in tests)

Pipeline improvments:

• Remove Default filters. Since this was by now replaced by the default pipelines
• Create a tutorial for library cleaning
  • The poetry install does not seem to match with the readme. Does pip install -e .[dev] still work? It does not install pytest
• Make progressing report use the order of filter orders for clarity (now it is on order of affecting a spectrum)
• Make a FilterFunction class. That does the filter function loading, instead of doing this in spectrum processor.