Replace Unicode / LaTeX conversion table with R script

[nanxstats]

Suggestion from @yihui: https://github.com/Merck/r2rtf/pull/217/files?diff=unified&w=0#r1564779269

The goal is to put the unicode_latex variable in an R script, to replace the current binary form stored in R/sysdata.rda.

This would also avoid the need of the globalVariables() call for unicode_latex as a hack to bypass R CMD check notes.

[elong0527]

Based on our discussion, it might trigger issues with special characters. Would consider to keep it as is.

[yihui]

Special characters written in the form \Uxxxx should be fine for R CMD check (here is an example). Alternatively, they can be encoded to integers via utf8ToInt() and converted back to UTF-8 via intToUtf8(). Technically I guess the problem is solvable, but I understand it's no big deal, so I'm okay with leaving it alone.