Refactoring for creating dummy instances for catching parsing errors in Match Lines

[CarlosCancino-Chacon]

In PR #332, dummy instances of match lines are created inside the prepare_kwargs_from_matchline method, in order to fix issues with regular expression pattern checking (and raising MatchErrors). It would be more efficient to create the dummy instances of the match lines outside of the method (i.e., as global variables), otherwise, the new instances of dummy_snoteand dummy_instance will slow down the parsing of match files.

However, as pointed by @sildater, creating dummy instances for checking match errors would require a larger refactoring and potentially introduce other issues. The current solution in PR #332 will stay for the time being, but if this solution ends up slowing down pipelines that require the loading of many match files (e.g., from the (n)ASAP dataset), we should consider a refactoring.

[sildater]

maybe PR #334 addresses this with minimal changes