Feature request: Unicode bullets

I’ve been using plain-Unicode text for lists, e.g.:

Project Changelog
=================

Release 4 (2013-07-02)
• Connect the sanafratz to the grepsling
• Hypercharge the whizding

Release 3 (1988-05-04)
• Check for loose wires in the kugelator

Release 2 (1988-04-01)
• Collect runaway monkeys
• Clean up ketchup spill
 
Release 1 (1987-12-15)
• Launch satellite

This almost translates to AsciiDoc, except I need to downgrade the source from using real bullet points (U+2022 •) to ASCII asterisks. ☹

May I request that AsciiDoctor be extended to recognize Unicode bullets as input for un-ordered lists?

(This is in a way the converse to ‹https://github.com/asciidoctor/asciidoctor/issues/957›.)

mojavelinux wrote

jcsalomon wrote

May I request that AsciiDoctor be extended to recognize Unicode bullets as input for un-ordered lists?

I think this goes against the spirit of being able to describe the formatting using the ascii code range (or course, it still supports UTF-8 content).

Depends on how that spirit is understood, I suppose: If it’s “AsciiDoc is ASCII (but UTF8-safe),” then this is in direct opposition; OTOH, if it’s “Pure ASCII source is supported,” then that leaves itself open to compatibility extensions using Unicode.

mojavelinux wrote

I suppose we could consider adding one unicode character to be used as an alternative to the asterisk. Is the bullet the character you would prefer over all others? Is that pretty universal? In order words, can you find other cases of it being used this way?

It’s kinda hard to search for “•”, so I have no idea how prevalent this is. I know I’m not the only one with such a preference; I’ve found a parallel discussion on the Commonmark discussion site.

This is very much a ‘camel’s nose’ feature, though. I use U+2022 (•) because there’s an easy Windows keyboard shortcut for it (Alt+0149) and almost all fonts have a glyph for it, but I could definitely see a follow-up request for U+2023 (‣), or the other bullets, or the heart or fleuron or ballot-mark characters. So while I’m proposing this feature and could really use it, I’m not whole-heartedly advocating it.

I can still make a case that it’s useful to allow input characters that do not have other meanings within AsciiDoc: i.e., I’m not proposing that “foo •bar• bas” be treated as identical to “foo *bar* bas”. Also, since CSS only knows about disc, circle, & square bullets, there might not be such a temptation toward proliferation of bullet-input characters.

—Joel

Dan “mojavelinux” Allen wrote

Understand that the downside of supporting U+2022 for lists is the matching overhead that's incurred. Because of the rules of the AsciiDoc syntax, the list item regex {1} has to be run on nearly every line in the AsciiDoc document (except for preformatted regions). If we determine the overhead is minimal (or, even better, negligible), then we can consider adding this character.

Can you file an issue report {2} so we can track it?

Filed as asciidoctor/asciidoctor#1177, including a suggested implementation.

—Joel