Refactoring rizin type system: introduce type scope.

[rockrid3r]

Addresses #4421 .

• Introduce RzTypeScope.
• Make the *_by_offset hashtables in dwarf_process.c being attached to 1 compilation unit: allocate at the beginning, free before the end of processing current CU.
• Change implementation of "by name" type requests.
• Refactor API usage of "by name" type requests.
• Additional API

Introduce RzTypeScope

Should be a tree-like independent structure like this:

┌──────┐                                          
│ src  │                                          
└──┬───┘                                          
   │     ┌────────┐                               
   ├─────┤ headers│                               
   │     └───┬────┘                               
   │         │         ┌►struct A;                
   │         └►inc.h───┤                          
   │                   └►enum E;                  
   │     ┌────────┐                               
   └─────┤  src   │                               
         └───┬────┘                               
             │                                    
             │   ┌────────┐                       
             ├───┤   a    │                       
             │   └───┬────┘                       
             │       └───► a.c───►struct A;       
             │                                    
             ├──► b.c─┬─►struct B;                
             │        │                           
             │        └─►typedef struct B A;      
             │                                    
             └──► c.c (includes headers/inc.h)    
                   │                              
                   ├───► struct A; (headers/inc.h)
                   │                              
                   └───► enum E; (headers/inc.h)  

Change API implementation of "by name" type requests.

• rz_type_db_get_base_type
• rz_type_db_get_compound_type
• rz_type_db_save_base_type
• rz_type_db_update_base_type
• rz_type_db_delete_base_type
• rz_type_db_get_base_types_of_kind
• rz_type_db_get_base_types
• rz_type_db_edit_base_type
• rz_type_db_base_type_unwrap_typedef

Refactor API usage of "by name" type requests.

typedb->types usage:

• base.c
• serialize_types.c
• type.c
• typeclass.c
• c_cpp_parser.c
• test_type.c

rz_type_db.*base_type usage:

• dwarf_process.c
• pdb_process.c
• cpdb.c
• ctypes.c
• base.c
• format.c
• helpers.c
• path.c
• type.c
• typeclass.c
• test_dwarf_integration.c
• test_pdb.c
• test_project_migrate.c
• test_analysis_function.c
• test_serialize_types.c
• test_type.c

Additional API

• get all types for the given scope
• remove all types for the given scope

Your checklist for this pull request

• I've read the guidelines for contributing to this repository
• I made sure to follow the project's coding style
• I've documented or updated the documentation of every function and struct this PR changes. If not so I've explained why.
• I've added tests that prove my fix is effective or that my feature works (if possible)
• I've updated the rizin book with the relevant information (if needed)

Detailed description

...

Test plan

...

Closing issues

...

[XVilka]

You will need to extract CU/scope from PDB too

[XVilka]

Add Doxygen

[imbillow]

It's a bit awkward here.

[rockrid3r]

Kind of. But there is no need to have it until the rizin shuts down.

• it is not used anywhere except dwarf_process.c.
• we have copy of all types in typedb->types.

so it just takes RAM for nothing.
The same for debug_info->base_type_by_offset and debug_info->base_types_by_name.

[rockrid3r]

The way scope is defined right now is wrong.

Firstly, it uses the RzBinDwarfCompUnit->name, which is basically filename of CU, not the place where struct was defined (struct may be included).

Secondly, we want the RzTypeScope to be an independent tree-like structure. And RzBaseType should contain pointers to it's nodes.