bug#58098: Improve composition rules for Gumrukhi

The form

(chess-symbol #x1FA00 #x1FA67)

requires that _both_ #x1FA00 and #x1FA67 be supported by a font, for
it to be eligible to display chess-symbols. By contrast, the form

(chess-symbol . [#x1FA00 #x1FA67])

requires that _either_ of the two characters is supported. So my
question is: do we really want _both_ of the characters supported by a
font, and if not, do we really want Emacs to reject such a font?

I have updated the patch accordingly and replaced (chess-symbol #x1FA00 #x1FA67) with (chess-symbol . [#x1FA00 #x1FA67])

> You don't have Symbola installed?

I have the Symbola font installed via the gdouros-symbola-fonts package on my Fedora37 system, but it does not seem to support
the chess-symbols block as verified by hb-view.
Its description in the fedora wiki also does not mention the chess-symbols block:
Symbola covers the following scripts and symbols supported by The Unicode Standard 5.2: Basic Latin, Latin-1 Supplement, Latin Extended-A, IPA Extensions, Spacing Modifier Letters, Greek and Coptic, Cyrillic, Cyrillic Supplementary, General Punctuation, Superscripts and Subscripts, Combining Diacritical Marks for Symbols, Letterlike Symbols, Number Forms, Arrows, Mathematical Operators, Miscellaneous Technical, Control Pictures, Optical Character Recognition, Box Drawing, Block Elements, Geometric Shapes, Miscellaneous Symbols, Dingbats, Miscellaneous Mathematical Symbols-A, Supplemental Arrows-A, Supplemental Arrows-B, Miscellaneous Mathematical Symbols-B, Supplemental Mathematical Operators, Miscellaneous Symbols and Arrows, Supplemental Punctuation, CJK Symbols and Punctuation, Yijing Hexagram Symbols, Vertical Forms, Combining Half Marks, CJK Compatibility Forms, Specials, Tai Xuan Jing Symbols, Counting Rod Numerals, Mathematical Alphanumeric Symbols, Mahjong Tile Symbols, Domino Tile Symbols.

On Tue, Sep 27, 2022 at 7:38 PM Eli Zaretskii <eliz@gnu.org> wrote:

> From: समीर सिंह Sameer Singh <lumarzeli30@gmail.com>
> Date: Tue, 27 Sep 2022 19:22:09 +0530
> Cc: 58098@debbugs.gnu.org
>
> I don't understand this part: why should we require #x1FA67? Did you
> perhaps mean this:
>
> (chess-symbol . [#x1FA00 #x1FA67])
>
> Isn't the script-representative-chars used to improve font selection by requesting more characters?
> So does the specific character requested have any effect? because I chose it randomly.

The form

(chess-symbol #x1FA00 #x1FA67)

requires that _both_ #x1FA00 and #x1FA67 be supported by a font, for
it to be eligible to display chess-symbols. By contrast, the form

(chess-symbol . [#x1FA00 #x1FA67])

requires that _either_ of the two characters is supported. So my
question is: do we really want _both_ of the characters supported by a
font, and if not, do we really want Emacs to reject such a font?

Did you look at what #x1FA67 looks like? It is not a "traditional"
chess symbol. And neither are the characters that are its neighbors.

> We already have the setup for chess-symbol, below this line where you
> are making changes. Is it not enough for some reason?
>
> Prior to this patch the chess symbol block was not rendered on my machine despite having its font (Noto
> Sans Symbols2)

You don't have Symbola installed?

From:	समीर सिंह Sameer Singh
Subject:	bug#58098: Improve composition rules for Gumrukhi
Date:	Tue, 27 Sep 2022 20:54:22 +0530