Small simplification removing extra layer (closes #8)

[eddelbuettel]

This PR removes the try / catch block from four alignment helpers. Rcpp automatically adds such a block by expanding the BEGIN_RCPP and END_RCPP macros inserted by compileAttributes(), and also uses an approach to report the error back that has seen some refinement. This should help with the small valgrind leak reported at CRAN -- and does so in local testing.

[codecov]

Codecov Report

Patch coverage: 100.00% and no project coverage change.

Comparison is base (ea49578) 100.00% compared to head (f924615) 100.00%.

Additional details and impacted files

@@            Coverage Diff            @@
##            master        #9   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files            2         2           
  Lines          307       304    -3     
=========================================
- Hits           307       304    -3     

Impacted Files	Coverage Δ
src/align.cpp	100.00% <100.00%> (ø)

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report in Codecov by Sentry.
📢 Do you have feedback about the report comment? Let us know in this issue.

[eddelbuettel]

Actually, updating to data.table 0.14.3 as it is right now at its repo as opposed to the variant I had changes things again so I market this as draft, and will update the issue #8.

[eddelbuettel]

Yikes. I had forgotten about this. Still an issue we need to get to.

I can reproduce with data.table from CRAN as well as the dev version. A simple R CMD check --use-valgrind is enough; it does not stop the run but file dtts.Rcheck/tests/tinytest.Rout has something.

[eddelbuettel]

Ok, undrafted the PR. We both know it does not improve the valgrind issue by itself, but it helps with a marginal small improvement.