Improve translation of Spectra object to python data structures

[jorainer]

At present (version 0.2.0) the Python list of matchms Spectrum objects is created in a loop (https://github.com/rformassspectrometry/SpectriPy/blob/main/R/conversion.R#L100-L101), i.e. the R Spectra is converted one element by one element into a Spectrum. Ideally, we should make use of the strength of Spectra, i.e. that it is a collection of spectra, not a single spectrum and convert the full object at once.

Options:

• extract the full data from a Spectra as plain R objects, translate them to Python object using reticulate and iterate over the items in Python to create the list.
• any other efficient way to create a list of Spectrum, not one by one, but using a large set of data?

See also issue #3 for a discussion of the original implementation.

[jorainer]

Instead of improving the translation efficiency we're also investigating the possibility of not having to translate the data all the time. This is related to issue #33

[jorainer]

Improved conversion is implemented in PR #50