Re: [SPAM UNSURE] Re: AdaCore Ada_Language_Server via eglot

[Stephen Leake]

Troy Brown <brownts@troybrown.dev> writes:

> Is this something you're considering releasing officially?  

No, because indent doesn't work well. Ada_Language_Server refuses to
indent any code with syntax errors.

On the other hand, ada-mode used to be that way as well, so I suppose it
would be nice to support that option.

> Also, is there any reason this wouldn't be applicable to lsp-mode in
> addition to eglot?

Ada_Language_Server does respect the LSP protocol, so it will work with
lsp-mode. I have no idea what it would take to set that up;
ada-eglot-mode.el might be enough. In which case it should be renamed to
ada-lsp-mode.el.

> One of the reasons I ask is that I often find myself disabling/overriding
> some features of the current ada-mode (e.g., skeletons in favor of
> YASnippet, require-final-newline setting, etc.) or not using them at
> all.

I'd be open to providing more ada-mode flags to turn some things off
more easily. Can you show what you do here?

I guess only loading ada-skel if requested, or if some other skeleton
package is not provided, might be easy to support.

ada-mode should not set require-final-newline. But removing that now
would be a regression, so I think we're stuck with it. It is easy to
override.

Similarly for comment-padding.

> For example, I don't use the packaged method for building...I'm often
> interfacing with makefiles or have multiple permutations of the build
> and multiple targets/compilers, such that the built-in features don't
> really work well in my workflow. 

Me too, but I just ignore those commands; they don't get in the way.

-- 
-- Stephe