I’m working on adding some matchers support to libclang however I’m quiet confused with the preferred naming conventions
Current function naming:
- clang_<c_type>_ (e.g. clang_CXCursorSet_insert)
- clang_<c++type>_ (e.g. upper: clang_Cursor_getArgument lower: clang_index_getClientEntity )
- clang_get (e.g. clang_getDiagnosticInSet)
Dispose and create seems a little more consistent with prefix of clang_create/clang_dispose however there are some which use dispose/create suffix (e.g. clang_IndexAction_, clang_remap_)
Even variable names for functions are consistent:
- No names at all (clang_remap_dispose(CXRemapping))
- Underscore names (e.g. unsigned index_options)
- Camel case upper (e.g. unsigned NumTokens)
- Camel case lower (e.g. unsigned *isGenerated)
With all of these different alternatives what should be used for new code?