Use recursive approach for indexing header files

[VadimGush]

Currently this scripts just scans all header files in the specified directory. This approach doesn't work when we want to use only one header file as our starting point to build a new one. So it might be reasonable to go to the content of one header file instead and then index all of its dependencies (and for every of those dependencies we will perform indexing of all of its dependencies).

The workflow should be as follows:

stateDiagram-v2
  s1: Scan header file
  i: Add it to the index
  c: Does it include header?
  s4: Scan current directory
  s5: Does it have another header file
  s6: Build header from the index

  [*] --> s1
  s1 --> i
  i --> c
  c --> s4: no
  c --> s1: yes

  s4 --> s5
  s5 --> s1: yes
  s5 --> s6: no

Loading