Find the Unicode character you want by drawing it

Originally published at:

The great thing about procrastination and the internet is the longer you put off doing interesting little projects you’ve thought of doing the greater the chance someone else will do it for you.

hoping someone gets to work on that make me rich project soon.


All I get is

eta: dafuq? Discourse replaces Unicode chars with an emoji? I’m not a huge fan of the automagic replacement stuff.


If you have a page book marked where you can draw a unicode character, why not just book mark a page that has all the unicode characters to cut an paste?

I was pretty excited until I read the ‘Chinese Japanese and Korean characters not supported’ disclaimer. That would be truly useful.


Bookmarking the unicode page might be useful if your drawing skill is so bad that it’s quicker to search the page manually, I guess.

If that’s your preferred method, more power to you, but it’s nice to have the option.


Found that symbol after drawing a smiley face

so not bad.

1 Like

Doesn’t seem to support Babylonian cuneiform, although they are part of Unicode, but not in the basic block.

You want a page with the 100,000 or so allocated Unicode code points? You think that would be easier than this?

1 Like

Not sure why I’d use this, but seemed to work pretty well at understanding my half-remembered § approximation.

As a bonus, I know now there’s a Georgian letter har ( ჴ ) and a Lao vowel sign ay ( ໃ ) that can come in handy for stereotypical Mexican Pirate uses. ໃ ໃ ໃ! ჴ ჴ!


It arguably depends on what you know about the glyph you are searching for.

In the references provided by the Unicode Consortium each codepoint has a description like “POWER ON-OFF SYMBOL
• IEC 5010 power on-off symbol”

If you have a textual idea of what you are looking for; a single page with all those descriptions would be ideal: just hit CTRL+F and start typing.

If you have a visual idea of what you are looking for, 100,000 code points in a somewhat idiosyncratic order dictated largely by historical quirks would be about as close to useless as one could reasonably imagine.

It’s like most search problems, really: the best way to find something depends on what you do know and what gaps in knowledge you are trying to fill.


There is no Yuzz. Or Wum, or Um, or Humpf. I think it’s broken.!


I ▢ Unicode


Fine. Close enough.


hihihihi… boobies! Ꙭ

After you wait 10 minutes for the page the of 100,000 descriptions to load. If you want images too, add another 30 minutes.

On a Mac you can set the Keyboard pref. to ‘Show Keyboard, Emoji, and Symbol Viewers in menu bar’
Then you have a picker you can select that has a search function:

you can double-click a glyph to enter it. It’s pretty handy for odd symbols and emoji.


And having the mentality of a 10-year-old with a proper dictionary for the first time, my attempts at crudely drawn obscenties produced:

which is a bishop symbol.



But only once, unless the caching mechanism involved is seriously braindead. Unicode is big and hairy; but it’s not that big: If you want just the Unicode Character Definitions, those weigh in(zipped) at 6.3MB for CJK ‘unified’ and 5.4MB for everything else. If you want the nice, attractive, PDF; that’ll be 104MB.

Not the sort of thing one just casually loads on their cellphone as the fancy strikes them; but barely a blip by the standards of impractically large files.

1 Like