Skip to content

Latest commit

 

History

History
120 lines (89 loc) · 4.92 KB

09-EncodingsUnicode.md

File metadata and controls

120 lines (89 loc) · 4.92 KB

Encodings and Unicode

Vim is capable of using different character encodings when editing and saving files. I find UTF-8 the most useful encoding.

Character Encodings

Vim is capable of working with different character encodings.

Command Description
:set encoding list current character encoding
:set encoding=utf-8 set current character encoding vim displays
:set fileencoding list current encoding in which to save files
:set encoding=latin1 edit using LATIN-1 character encoding
:set fileencoding=latin1 save files with LATIN-1 encoding
:set fileencodings list encodings to try when loading a file

The two main encodings are UTF-8 for UNIX, MAC-OS and HTML/Web, and UTF-16 for Microsoft Windows. Windows tends to use a mix of UTF-16, UTF-8, and legacy encodings.

On POSIX systems I used to put

    set encoding=utf-8
    set fileencoding=utf-8

In my ~/.vim/vimrc. Now I do a similar configuration via init.lua.

When I go native on Windows, vim automatically figures out whether the edited file is in UTF-8 or UTF-16LE with \r\n line endings.

Aside: GIT, for text files, does the conversion to the correct format depending on the OS. In the olde days, we used to FTP files in text mode to convert format between different OS's.

Vim Digraphs

Following the recommendations of RFC-1345, vim allows users to enter characters within whatever encodings they are using via 2 character "diagraph" sequences. While in insert mode, type <C-k> followed by a two character sequence.

Command Description
:dig list diagraphs available for current encoding
<C-k>a^ enter the character â
<C-k>o: enter the character ö
<C-k>i' enter the character í
<C-k>l* enter the character λ

The choice of characters you can enter this way depends on the encoding you are using.

Using Unicode Code Points

In modern Unix terminal emulators and Libre Office, input and display of Unicode code points just works. Terminals are fixed width font beasts, but Libre Office handles the variable width code points just fine.

When using gvim, or vim/nvim with a Unicode aware terminal emulator like rxvt-unicode or gnome-terminal, code points can be entered while in insert mode via

Command Description
<C-S-u>03b2<CR> enter the character β
<C-S-u>3bb<CR> enter the character λ

Where <C-S-u> means holding down CTRL+SHIFT+u.

Note: Using <S-C-u> will not work.

Note: <C-S-u> does not work while on the Linux console.

Note: Defining vim key mapping with "key chords" like <C-S-u> will not work, neither will making it the target of the mapping.

Aside: Be aware that for both gvim and vim/nvim running in a terminal emulator, you are not interacting with the "terminal" being emulated, but the underlying GUI application. As far as a real terminal is concerned, CTRL+d and CTRL+D are the same control character, ASCII 4. That is what is coming down the RS-232 cable. The Graphical User Interface (GUI) is seeing all your keystrokes and can distinguish whether you pressed <C-d> or <C-S-d>.

Aside: Terminals in the vt102/vt220 family are keyboard/monitor interfaces to RS-232 cables. Holding the CTRL key down electrically zeroed out bits 6 and 7 of the 7-bit character key typed. Likewise, holding the down the SHIFT key zeroed out bit 6. The ASCII values were chosen to make this work.

ASCII Decimal Octal Binary Symbol Description
107 64+32+11 153 01101011 k Lowercase A
75 64+11 113 01001011 K Uppercse A
11 11 013 00001011 VT Vertical Tab
100 64+32+4 144 01100100 d Lowercase D
68 64+4 104 01000100 D Uppercase D
4 4 004 00000100 EOT End of Transmission
123 64+32+27 173 01111011 { Curly Left Bracket
91 64+27 133 01011011 [ Square Left Bracket
27 27 033 00011011 ESC Escape

Pro-trick: <C-[> is the same as <Esc>

| prev: Mult Neovim Windows | Home | next: Regular Expressions |