bug#31624: 27.0; (elisp) `Syntax Flags'

bug#31624: 27.0; (elisp) `Syntax Flags'

[Drew Adams]

Something seems to be missing. This node refers to "a" style, but that
seems to be defined nowhere.  The node defines "comment style" this way:

  A comment style is a set of flags 'b', 'c', and 'n', ...

No mention of style "a" there.  What is style "a"?  I don't see it
specified anywhere.

Presumably style "a" is some set composed of the flags 'b', 'c', and
'n', but what set?

And what about the relation between those "flags" and the "styles"?  If
a flag is an element in a set that is a style, why does the doc use
chars b and c to name both flags and styles - that just confuses things.
If you want to relate style "b" to flag 'b' then maybe use "B", not "b",
as the style name.

(Yes, I realize that this text is very old - at least as old as Emacs
20.  But it seems like it could/should be clearer.)




In GNU Emacs 27.0.50 (build 3, x86_64-w64-mingw32)
 of 2018-03-21
Repository revision: e70d0c9e66d7a8609450b2889869d16aeb0363b5
Windowing system distributor `Microsoft Corp.', version 6.1.7601
Configured using:
 `configure --without-dbus --host=x86_64-w64-mingw32
 --without-compress-install -C 'CFLAGS=-O2 -static -g3''

bug#31624: 27.0; (elisp) `Syntax Flags'

[Alan Mackenzie]

Hello, Drew.

In article <mailman.627.1527518890.1292.bug-gnu-emacs@gnu.org> you wrote:
> Something seems to be missing. This node refers to "a" style, but that
> seems to be defined nowhere.  The node defines "comment style" this way:

>   A comment style is a set of flags 'b', 'c', and 'n', ...

> No mention of style "a" there.  What is style "a"?  I don't see it
> specified anywhere.

> Presumably style "a" is some set composed of the flags 'b', 'c', and
> 'n', but what set?

When neither flag 'b' nor flag 'c' is set, you have style "a".  

> And what about the relation between those "flags" and the "styles"?

When 'b' is set, you have style "b", when 'c' is set, you have style
"c".  I think having both of them set is undefined (but it does leave
room for a style "d" (shudder!)).

> If a flag is an element in a set that is a style, why does the doc use
> chars b and c to name both flags and styles - that just confuses
> things.  If you want to relate style "b" to flag 'b' then maybe use
> "B", not "b", as the style name.

I think that would be less confusing if something like my first sentence
were in the doc somewhere.

> (Yes, I realize that this text is very old - at least as old as Emacs
> 20.  But it seems like it could/should be clearer.)

I though style "c" was somewhat younger.  Maybe not.  But maybe it could
be clarified.

> In GNU Emacs 27.0.50 (build 3, x86_64-w64-mingw32)
>  of 2018-03-21
> Repository revision: e70d0c9e66d7a8609450b2889869d16aeb0363b5
> Windowing system distributor `Microsoft Corp.', version 6.1.7601
> Configured using:
>  `configure --without-dbus --host=x86_64-w64-mingw32
>  --without-compress-install -C 'CFLAGS=-O2 -static -g3''

-- 
Alan Mackenzie (Nuremberg, Germany).

bug#31624: 27.0; (elisp) `Syntax Flags'

[Drew Adams]

> When neither flag 'b' nor flag 'c' is set, you have style "a".

If that's the definition of style a then the doc should state that.

> > And what about the relation between those "flags" and the "styles"?
> 
> When 'b' is set, you have style "b", when 'c' is set, you have style
> "c".  I think having both of them set is undefined (but it does leave
> room for a style "d" (shudder!)).
 
> > If a flag is an element in a set that is a style, why does the doc use
> > chars b and c to name both flags and styles - that just confuses
> > things.  If you want to relate style "b" to flag 'b' then maybe use
> > "B", not "b", as the style name.
> 
> I think that would be less confusing if something like my first sentence
> were in the doc somewhere.

Both changes are needed, I think.  It just confuses to use the
same name, with the only difference being single vs double quote
chars around it, for two different concepts/things.

> > (Yes, I realize that this text is very old - at least as old as Emacs
> > 20.  But it seems like it could/should be clearer.)
> 
> I though style "c" was somewhat younger.  Maybe not.  But maybe it could
> be clarified.

You're right - c is not in Emacs 20, for instance.

bug#31624: 27.0; (elisp) `Syntax Flags'

[Eli Zaretskii]

> Date: 28 May 2018 21:39:44 -0000
> From: Alan Mackenzie <acm@muc.de>
> Cc: 31624@debbugs.gnu.org
> 
> In article <mailman.627.1527518890.1292.bug-gnu-emacs@gnu.org> you wrote:
> > Something seems to be missing. This node refers to "a" style, but that
> > seems to be defined nowhere.  The node defines "comment style" this way:
> 
> >   A comment style is a set of flags 'b', 'c', and 'n', ...
> 
> > No mention of style "a" there.  What is style "a"?  I don't see it
> > specified anywhere.
> 
> > Presumably style "a" is some set composed of the flags 'b', 'c', and
> > 'n', but what set?
> 
> When neither flag 'b' nor flag 'c' is set, you have style "a".  

I added that to the text.

> > And what about the relation between those "flags" and the "styles"?
> 
> When 'b' is set, you have style "b", when 'c' is set, you have style
> "c".  I think having both of them set is undefined (but it does leave
> room for a style "d" (shudder!)).

I added a sentence that explains how styles and flags are related.

Thanks.