Redundant distinction between rmc.unit.time option POSIX and character

[vincentvanhees]

What we have for option "POSIX" also works for option "character".

• Deprecate "character" inside read.my.acc function and in documentation, and in check_params re-assign to "POSIX" if user supplies "character".
• Add check that P$timestamp in read.myacc.csv does not become NA, and provide informative error. For example, by plotting rmc.format.time and the first timestamp before it got converted to NA.
• Add documentation on the imputation of missing decimal places, which is done but not in read.myacc.csv vignette.

[vincentvanhees]

I have added the "help wanted" label as I do not have time and am not paid to work on this at the moment.

[vincentvanhees]

Actually, because of the extra processing steps in POSIX it is much slower, so would be good to also investigate how to avoid slowing things down when merging the two.