bug#60531: 29.0.50; Some fixes and suggestions for the docstrings of treesit-parent-while and treesit-parent-until

bug#60531: 29.0.50; Some fixes and suggestions for the docstrings of treesit-parent-while and treesit-parent-until

[Daniel Martín via Bug reports for GNU Emacs, the Swiss army knife of text editors]

The docstring of treesit-parent-while says:

Return the furthest parent of NODE that satisfies PRED.
Return nil if none was found.  PRED should be a function that
takes one argument, the parent node.

But, according to the source code, the first invocation of NODE passes
NODE itself, not its parent:

(let ((last nil))
    (while (and node (funcall pred node))

The same happens with the docstring of treesit-parent-until.

Another suggestion I have is that "the furthest parent of NODE" is a bit
ambiguous (furthest from NODE or from the root?).  I think it would be
more clear to say it will return the highest (ie. least deep) *ancestor*
of NODE.  In 37.3 Retrieving Nodes, we introduce the concepts of
"higher" and "lower" when explaining Tree-sitter nodes, so using this
terminology consistently would avoid confusion.

bug#60531: 29.0.50; Some fixes and suggestions for the docstrings of treesit-parent-while and treesit-parent-until

[Eli Zaretskii]

> Date: Tue, 03 Jan 2023 23:45:35 +0100
> From:  Daniel Martín via "Bug reports for GNU Emacs,
>  the Swiss army knife of text editors" <bug-gnu-emacs@gnu.org>
> 
> 
> The docstring of treesit-parent-while says:
> 
> Return the furthest parent of NODE that satisfies PRED.
> Return nil if none was found.  PRED should be a function that
> takes one argument, the parent node.
> 
> But, according to the source code, the first invocation of NODE passes
> NODE itself, not its parent:
> 
> (let ((last nil))
>     (while (and node (funcall pred node))
> 
> The same happens with the docstring of treesit-parent-until.

This is clearly a case of unnecessary details.  I made the doc string
say just "node".

> Another suggestion I have is that "the furthest parent of NODE" is a bit
> ambiguous (furthest from NODE or from the root?).  I think it would be
> more clear to say it will return the highest (ie. least deep) *ancestor*
> of NODE.  In 37.3 Retrieving Nodes, we introduce the concepts of
> "higher" and "lower" when explaining Tree-sitter nodes, so using this
> terminology consistently would avoid confusion.

Instead of changing a potentially confusing word to another
potentially confusing word, I decided to describe in more detail what
the functions do.