Enter Unicode characters with fcitx5

The fcitx5 functionality I reach for the most, its Unicode module lets you access an astronomical number of characters, symbols and glyphs with the expected convenience. But beyond the obvious, it also bundles a pearl of ingenuity that lets you identify these invisible, combining or other peculiarly artful characters like a jeweller would delight in examining a rare gem.

Familiar code points entry

fcitx5 has a built-in Unicode input method that allows users to enter any Unicode character by its code point. Just start the daemon, verify that your journalctl --user --unit fcitx5 shows Loaded addon unicode, and you're all set.

To use it, simply switch to the Unicode input method—by default, Ctrl-Shift-U, and type the hexadecimal code point of the character you want to enter, followed by Enter or Space, cancelling out with Esc.

I am hugely fond of hyphens, en dashes and em dashes (and I do sorely resent the dreadful notion that writers that employ them may be of the artificial complexion, and you will not catch me present a range with anything other than the adequate punctuation mark.

• Ctrl-Shift-U, then 2013 and voilà, an en dash (–), transforming subtractions into numeric ranges.

  The first World War didn't "span negative four", or whatever 1914-1918 may mean; it actually raged on from 1914 through 1918, 1914–1918.

• Ctrl-Shift-U, then 2014 and you can wield the em dash (—) to aptly document your suspicion of unorthodoxy in the rhetoric of your fellow comrades:

  "Except—" began Winston doubtfully, and then stopped.
  It had been on the tip of his tongue to say "Except the proles," [...]

  — George Orwell, Nineteen Eighty-Four

• Ctrl-Shift-U, then 300 gives you the grave accent (`), which unlocks a Frenchman's ability to start a sentence with the preposition "to", À.

• 301, and you get to put the acute accent on any letter you fancy, possibly even spelling my name the way my parents intended, Éric.

• 302 gives you the circumflex, 303 the tilde, 304 the macron, 305 the overline, and so on... take your pick!

But while I find all of these notable examples to be surprisingly easy to remember, you do not have to index all of Unicode in your head—that valuable space is well known to be most suitable to π's digits after all.

Unicode search

What about the characters you won't reach for regularly enough to commit to (muscle?) memory their code point? With Ctrl-Shift-Alt-U (I like to call this "digital mouthful" a "handful"), you will be able to search by description.

• Drawing some TUI menu? Have enough boxes to make Brad Pitt's head spin:

  ┌───────────────────┐
  │  ╔═══╗ Some Text  │
  │  ╚═╦═╝ in the box │
  ╞═╤══╩══╤═══════════╡
  │ ├──┬──┤           │
  │ └──┴──┘           │
  └───────────────────┘
  

  from Wikipedia's article on box-drawing characters

• the ideograph for love? 戀 is all that you need—according to the Beatles

• a healthy serving of spaghetti? 🍝, delivered on a silver platter.

And even reverse-search

Lastly, here's the neat trick that originally sold me on better adopting that tool: the "digital mouthful" Ctrl-Shift-Alt-U pops up already displaying some elements before you even start searching: they're your current selection and the contents of your clipboard, for convenient reverse-lookups.

Will you need that often? It does have some dubious value in being only able to pop up where you can type, but I reckon that shouldn't be a problem for any serious CLI dwellers.

And if once in a while you tend to some finesse writing, you might very well save yourself some xxd invocations—unless you can distinguish a soft-hyphen from a non-breaking space, from a regular space...

Referenced tools