在C＃中转换Fortran头文件.INC的最佳方法是什么

Question

I am currently working on a fortran to c# project & I have come across some header files with extension .INC Foll. is an example of the header file

**     MAIN.INC
   INTEGER uid
   INTEGER MathsMarks
   INTEGER AWAMarks
   INTEGER ScienceMarks
   LOGICAL isStudent
   REAL*8 percentage
   REAL avg
   CHARACTER*2 sName
   LOGICAL*1 sUpdateStat
   DOUBLE PRECISION d_var
   complex*8 c8_var

Can anyone guide me in the best strategy to be used to convert this header file (.INC) in c#

Answer 1

Do you know how main.inc is used? Does it appear in several files or just in one file? Is there a common block associated with these variable names?

If there are no common blocks associated with it, then it is probably a bunch of commonly used local variables that are not shared.

If there is a common block associated with it, then they are global variables. If it had been done properly, the common should also have been in the inc file. In C# if everything is in the same class, then these will be class level declarations.

Translations

• Integer = int
• logical = bool
• real*4 or real = float
• real*8 = double
• parameter = const
• dimension is an array declaration. Remember arrays start at 1 by default but this can be overridden in the declaration. They can start at any number. eg dimension(-23:10) will start at -23.
• equivalence is a union
• character*n is a string on n characters, space padded, not null terminated

[EDITED] Initially, everything is public and static in the initial translation until you figure out what the program is doing. Once you've done that, it can be refactored.

If it is named common, put it in a static class of the same name as the named common. If it is unnamed common, then put it in the main class as a static declaration. Normally if something is in a common block, it is shared across functions so don't put it in a struct.

Equivalence is a nasty one: depends on how it is used.

• Sometimes it is used for passing data to routines like passing a class. In that case you could just dispense with the equivalence and declare the structure as a class.
• Sometimes it is used as a global struct for moving values all over the place. You have to watch this one. It is normally a huge array of double precision. Every member is equivalenced to something completely different - some integers, some double precision, some characters. They always have the same index and always have the same name.
• Sometimes it is used for conversion
• Sometimes they are used for mapping one part of an array to another.