I plan to put some resources here in the future:

(1) R and STATA Codes to Utilize Industrial Censuses(1976, 1981,...2001) in Taiwan: 

-Original raw data released by Academia Sinica is ASCII (that is, sequences like 01001001...).

- I've written (1) R codes to translate these ASCII raw data and (2) a STATA.do file to generate variables and labels, clean some bugs, crosswalks of townships from 1976 to 2001(where the number of township =365) . I am working on constructing crosswalks to link  1905 census during the colonial era and 2001 townships.