necessity of ctag's -n

[Konfekt]

Just wondering how reliant this plug-in is on ctag's -n option? Could --excmd=mixed suffice?

[MiguelBarro]

You got me on that 🙄.
The plugin only provides a special function that improves over the default 'tagfunc' profiting from new ctags metadata for C++ (classes, namespaces, inheritance, etc...). That metadata was not available for C (which is the default 'tagfunc' target).
The presence of that metadata requires some flags on ctags invocation like --fields=+iaeSK --fields-c++=* but the command line provided in the docs is merely an example (despite of the section name):

cpptags/start/cpptags/doc/cpptags.txt

Lines 88 to 90 in 4a99d3f

	$ ctags --extras=+f -R --languages=c++ --langmap=c++:+.c. --c++-kinds=+pl
	-n --fields=+iaeSK --fields-c++=* --options=<option-file-path>
	-f <output-tags-file> <source-dirs>

In fact, the value of --excmd is indeferent to the plugin, because the 'tagfunc' doesn't rely on the line numbers at all.
Furthermore, the values for --fields and fields-c++ could be refined.
To be sincere the settings -n and --fields=+e were added because I had another side project that used ctags metadata to identify scope from cursor position. This way a custom 'omnifunc' could hint variables, members and methods names like a lsp sever would.

[Konfekt]

That's reassuring to hear, as I naturally hoped this to work most independently of the ctags configuration as possible; I assumed that it cared more about ctags' output than its tags file, the former independent of --excmd.