.readme.gotxt

`uni` queries the Unicode database from the commandline. It supports Unicode
15.1 (September 2023) and has good support for emojis.

There are four commands: `identify` codepoints in a string, `search` for
codepoints, `print` codepoints by class, block, or range, and `emoji` to find
emojis.

There are binaries on the [releases] page, and [packages] for a number of
platforms. You can also [run it in your browser][uni-wasm].

Compile from source with:

    % go install zgo.at/uni/v2@latest

which will give you a `uni` binary in `~/go/bin`.

README index:
- [Integrations](#integrations)
- [Usage](#usage)
  - [Identify](#identify)
  - [Search](#search)
  - [Print](#identify)
  - [Emoji](#emoji)
  - [JSON](#json)
- [ChangeLog](#changelog)
- [Development](#development)
- [Alternatives](#alternatives)

[uni-wasm]: https://arp242.github.io/uni-wasm/
[releases]: https://github.com/arp242/uni/releases
[packages]: https://repology.org/project/uni/versions

Integrations
------------

- [dmenu], [rofi], and [fzf] script at [`dmenu-uni`](/dmenu-uni). See the top of
  the script for some options you may want to frob with.

- For a Vim command see [`uni.vim`](/uni.vim); just copy/paste it in your vimrc.

[dmenu]: http://tools.suckless.org/dmenu
[rofi]: https://github.com/davatorium/rofi
[fzf]: https://github.com/junegunn/fzf

Usage
-----
*Note: the alignment is slightly off for some entries due to the way GitHub
renders wide characters; in terminals it should be aligned correctly.*

### Identify

Identify characters in a string, as a kind of a unicode-aware `hexdump`:

{{example "identify" "€"}}

`i` is a shortcut for `identify`:

{{example "i" "h€ý"}}

It reads from stdin:

     % head -c5 README.md | uni i
          CPoint  Dec    UTF8        HTML       Name
     '`'  U+0060  96     60          `    GRAVE ACCENT [backtick, backquote]
     'u'  U+0075  117    75          u     LATIN SMALL LETTER U
     'n'  U+006E  110    6e          n     LATIN SMALL LETTER N
     'i'  U+0069  105    69          i     LATIN SMALL LETTER I
     '`'  U+0060  96     60          `    GRAVE ACCENT [backtick, backquote]

    % echo 'U+1234 U+1111' | uni p
         CPoint  Dec    UTF8        HTML       Name
    'ᄑ' U+1111  4369   e1 84 91    ᄑ   HANGUL CHOSEONG PHIEUPH [P]
    'ሴ'  U+1234  4660   e1 88 b4    ሴ   ETHIOPIC SYLLABLE SEE

You can use `-compact` (or `-c`) to suppress the header, and `-format` (or `-f`)
to control the output format:

{{example "i" "-f" "%unicode %name" "a€🧟"}}

If the format string starts with `+` it will automatically be prepended with the
character, codepoint, and name:

{{example "i" "-f" "+%unicode" "a€🧟"}}

You can add more advanced options with `%(name flags)`; for example to generate
an aligned codepoint to X11 keysym mapping:

{{example "i" "-c" "-f" "0x%(hex l:auto f:0): %(keysym l:auto q:\":\",) // %name" "h€ý"}}

See `uni help` for more details on the `-format` flag; this flag can also be
added to other commands.

### Search

Search description:

{{example "search" "euro"}}

The `s` command is a shortcut for `search`. Multiple words are matched
individually:

{{example "s" "globe" "earth"}}

Use shell quoting for more literal matches:

{{trim 3 (example "s" "rightwards" "black" "arrow")}}

{{example "s" "rightwards black arrow"}}

Add `-or` or `-o` to combine the search terms with "OR" instead of "AND":

{{example "s" "-o" "globe" "milky"}}

### Print

Print specific codepoints or groups of codepoints:

{{example "print" "U+2042"}}

Print a custom range; `U+2042`, `U2042`, and `2042` are all identical:

{{example "print" "2042..2044"}}

You can also use hex, octal, and binary numbers: `0x2024`, `0o20102`, or
`0b10000001000010`.

General category:

{{trim 3 (example "p" "Po")}}

Blocks:

{{trim 5 (example "p" "arrows" "box drawing")}}

Print as table, and with a shorter name:

{{example "p" "-as" "table" "box"}}

Or more compact table:

{{example "p" "-as" "table" "box" "-compact"}}

### Emoji
The `emoji` command (shortcut: `e`) is is the real reason I wrote this:

{{example "e" "cry"}}

By default both the name and CLDR data are searched; the CLDR data is a list of
keywords for an emoji; prefix with `name:` or `n:` to search on the name only:

{{trim 3 (example "e" "smile")}}

{{example "e" "name:smile"}}

As you can see, the CLDR is pretty useful, as "smile" only gives one result as
most emojis use "smiling".

Prefix with `group:` to search by group:

{{example "e" "group:hands"}}

Group and search can be combined, and `group:` can be abbreviated to `g:`:

{{example "e" "g:cat-face" "grin"}}

Like with `search`, use `-or` to OR the parameters together instead of AND:

{{example "e" "-or" "g:face-glasses" "g:face-hat"}}

Apply skin tone modifiers with `-tone`:

{{example "e" "-tone" "dark" "g:hands"}}

The handshake emoji supports setting individual skin tones per hand since
Unicode 14, but this isn't supported, mostly because I can't really really think
a good CLI interface for setting this without breaking compatibility (there are
some other emojis too, like "holding hands" and "kissing" where you can set both
the gender and skin tone of both sides individually). Maybe for uni v3 someday.

The default is to display only the gender-neutral "person", but this can be
changed with the `-gender` option:

{{example "e" "-gender" "man" "g:person-gesture"}}

Both `-tone` and `-gender` accept multiple values. `-gender women,man` will
display both the female and male variants, and `-tone light,dark` will display
both a light and dark skin tone; use `all` to display all skin tones or genders:

{{example "e" "-tone" "light,dark" "-gender" "f,m" "shrug"}}

Like `print` and `identify`, you can use `-format`:

{{example "e" "g:cat-face" "-c" "-format" "%(name): %(emoji)"}}

See `uni help` for more details on the `-format` flag.

### JSON

With `-as json` or `-as j` you can output the data as JSON:

{{example "i" "-as" "json" "h€ý"}}

All the columns listed in `-f` will be included; you can use `-f all` to include
all columns:

{{example "i" "-as" "json" "-f" "all" "h€ý"}}

This also works for the `emoji` command:

{{example "e" "-as" "json" "-f" "all" "kissing cat"}}

All values are always a string, even numerical values. This makes things a bit
easier/consistent as JSON doesn't support hex literals and such. Use `jq` or
some other tool if you want to process the data further.

ChangeLog
---------
Moved to [CHANGELOG.md](/CHANGELOG.md).

Development
-----------
Re-generate the Unicode data with `go generate unidata`. Files are cached in
`unidata/.cache`, so clear that if you want to update the files from remote.
This requires zsh and GNU awk (gawk).

Alternatives
------------
Note this is from ~2017/2018 when I first wrote this; I don't re-evaluate every
program every year, and I don't go finding newly created tools every year
either.

### CLI/TUI

- https://github.com/philpennock/character

  More or less similar to uni, but very different CLI, and has some additional
  features. Seems pretty good.

- https://github.com/sindresorhus/emoj

  Doesn't support emojis sequences (e.g. MAN SHRUGGING is PERSON SHRUGGING +
  MAN, FIREFIGHTER is PERSON + FIRE TRUCK, etc); quite slow for a CLI program
  (`emoj smiling` takes 1.8s on my system, sometimes a lot longer), search
  results are pretty bad (`shrug` returns unamused face, thinking face, eyes,
  confused face, neutral face, tears of joy, and expressionless face ... but not
  the shrugging emoji), not a fan of npm (has 1862 dependencies).

- https://github.com/Fingel/tuimoji

  Grouping could be better, doesn't support emojis sequences, only interactive
  TUI, feels kinda slow-ish especially when searching.

- https://github.com/pemistahl/chr

  Only deals with codepoints, not emojis.

### GUI

- gnome-characters

  Uses Gnome interface/window decorations and won't work well with other WMs,
  doesn't deal with emoji sequences, I don't like the grouping/ordering it uses,
  requires two clicks to copy a character.

- gucharmap

  Doesn't display emojis, just unicode blocks.

- KCharSelect

  Many KDE-specific dependencies (106M). Didn't try it.

- https://github.com/Mange/rofi-emoji and https://github.com/fdw/rofimoji

  Both are pretty similar to the dmenu/rofi integration of uni with some minor
  differences, and both seem to work well with no major issues.

- gtk3 emoji picker (Ctrl+; or Ctrl+. in gtk 3.93 or newer)

  Only works in GTK, doesn't work with `GTK_IM_MODULE=xim` (needed for compose
  key), for some reasons the emojis look ugly, doesn't display emojis sequences,
  doesn't have a tooltip or other text description about what the emoji actually
  is, the variation selector doesn't seem to work (never displays skin tone?),
  doesn't work in Firefox.

  This is so broken on my system that it seems that I'm missing something for
  this to work or something?

- https://github.com/rugk/awesome-emoji-picker

  Only works in Firefox; takes a tad too long to open; doesn't support skin
  tones.

### Didn't investigate (yet)

Some alternatives people have suggested that I haven't looked at; make an issue
or email me if you know of any others.

- https://github.com/cassidyjames/ideogram
- https://github.com/OzymandiasTheGreat/emoji-keyboard
- https://github.com/salty-horse/ibus-uniemoji
- https://fcitx-im.org/wiki/Unicode
- http://kassiopeia.juls.savba.sk/~garabik/software/unicode/ and https://github.com/garabik/unicode (same?)
- https://billposer.org/Software/unidesc.html
- https://github.com/NoraCodes/charpicker (rofi)