Open MyriaCore opened 4 years ago
Here's a mock up of what the new SQL schema might look like under 3:
Advantages:
INSERT INTO ShortCodes (emoji, shortcode)
VALUES ("🎉", "tada")
INSERT INTO KeyWords (emoji, keyword)
VALUES ("🎉", "celebration")
SELECT keyword FROM KeyWords
WHERE emoji = "🎉"
emoji.name
in the old schema, Emojis.codepoint
in this one) is more self-evident, and is structurally validated by SQLITEname_search
, shortcodes
, etc.), and instead use individual tables, and SELECT
/ UNION
to get the data togetherCaveats:
Emojis
are dependent on Emojis
. This means:
codepoint
does exist in Emojis table. CASCADE
seems like a good choice. I like the ideas you outlined here :+1:
ATM EmojiSpider.py pulls from 2 sources - a unicode website, and emojipedia.org. The unicode website appears to be built primarily for human viewing, and NOT for parsing. It also has codepoints that don't really match up with the actual emoji representations, since the
fe0f
sequence (which can be used to control for text vs emoji presentation) is often omitted.I propose a few changes:
These changes would make the code base more readable, and it would also enable users to more easily hack the SQLITE DB to add things they want. It might also make it easier to deliver new features.