Remove multiple redefinitions of variables when reading equilibrium files?

Some of the parameters defined in the EFIT a- and g-files are re-defined multiple times when reading the files. This causes confusion in case there are issues with any such parameter.

E.g. simagx, rmagx, and zmagx are defined in multiple places in both the a- and g-file
https://github.com/LLNL/UEDGE/blob/d4926dfbe575b28b0612bc6aa61262a917a89383/flx/flxread.m#L333
https://github.com/LLNL/UEDGE/blob/d4926dfbe575b28b0612bc6aa61262a917a89383/flx/flxread.m#L440
https://github.com/LLNL/UEDGE/blob/d4926dfbe575b28b0612bc6aa61262a917a89383/flx/flxread.m#L441
https://github.com/LLNL/UEDGE/blob/d4926dfbe575b28b0612bc6aa61262a917a89383/flx/flxread.m#L442

Proposed solution:
Define each variable once during reading of the files, store other duplicate definition in dummy variables. Currently, the last instance of each variable in the g-file are used, although the EFIT files sometimes only define the variables in the first row. Thus, I suggest only the first occurrences in the g-file are used, the remainder being read into dummy variables. This way, the dependency on the a-file is minimized.

Alternatively, manual definitions could be used, using simple Python scripts to pick variables (such as magnetic axis and X-point locations) could be utilized.

Currently, only strike-point locations seem to be uniquely defined in the a-file, and are the only required values from the a-file. An internal routine (using e.g. Python) could be implemented to completely remove the need for an a-file.

Start with flx/flxread.m at the linked lines and trace how the EFIT a- and g-files assign simagx, rmagx, and zmagx. Determine which first occurrences are valid and which duplicate values should be read into dummy variables. Done means duplicate definitions no longer override one another and the required strike-point data remains available without unnecessary dependence on the a-file.