#4 was a good idea and worked well on a 'simple' update. But I think it's not quite future-proof-able.
As we have had to make some changes to the fucntions to get it on CRAN (like adding a void argument) and included extra structs / typdefs so they can be exposed
So instead I propose updating the script to
#4 was a good idea and worked well on a 'simple' update. But I think it's not quite future-proof-able.
As we have had to make some changes to the fucntions to get it on CRAN (like adding a
voidargument) and included extrastructs/typdefsso they can be exposedSo instead I propose updating the script to
h3api.hh3api.h.inh3Assert.cneeds antypedef int unused;to avoid the "ISO C forbids an empty translation unit [-Wpedantic]" warningh3Index.c-h3ToString- remove 'sprintf' line and returnreturn E_FAILED;(void)fixes.h3libapi.h) and output differences in a .txt / log of some sortbuild_init.shwhich builds theinit.cfile)