Python f-strings syntax highlighting in Vim

Getting Python syntax highlighting to work in Vim requires very little code, to my surprise.

Here is everything that you need and an explanation below.

" in ~/.config/nvim/after/syntax or ~/.vim/after/syntaxsyn match pythonEscape +{{+ contained containedin=pythonfString,pythonfDocstringsyn match pythonEscape +}}+ contained containedin=pythonfString,pythonfDocstring syn region pythonfString matchgroup=pythonQuotes      \ start=+[fF]\@1<=\z(['"]\)+ end="\z1"      \ contains=@Spell,pythonEscape,pythonInterpolationsyn region pythonfDocstring matchgroup=pythonQuotes      \ start=+[fF]\@1<=\z('''\|"""\)+ end="\z1" keepend      \ contains=@Spell,pythonEscape,pythonSpaceError,pythonInterpolation,pythonDoctest syn region pythonInterpolation contained      \ matchgroup=SpecialChar      \ start=+{{\@!+ end=+}}\@!+ skip=+{{+ keepend      \ contains=ALLBUT,pythonDecoratorName,pythonDecorator,pythonFunction,pythonDoctestValue,pythonDoctest syn match pythonStringModifier /:\(.[<^=>]\)\?[-+ ]\?#\?0\?[0-9]*[_,]\?\(\.[0-9]*\)\?[bcdeEfFgGnosxX%]\?/ contained containedin=pythonInterpolationsyn match pythonStringModifier /![sra]/ contained containedin=pythonInterpolation hi link pythonfString Stringhi link pythonfDocstring Stringhi link pythonStringModifier PreProc

Declaring a syntax region for f-strings

The first two lines define a new syntax region (see :h syn-region) called pythonfString.

We then declare how it starts with the regex [fF]\@1<=\z(['"]\), which is equivalent to (?:<=[fF])(['"]) in Perl regular expressions (see :h \@<=). The second line just handles the case of a docstring.

The string will end how it starts, so we can just reference the captured group using \z1 (we need to prefix it with z because it is an external pattern, see :h \z().

The matchgroup argument tells Vim which highlight group it should use to highlight the start/end pattern. The group pythonQuotes come from the default syntax file.

Handling expressions inside f-strings

We also need to declare what this region contains.

For this we declare another region called pythonInterpolation, which starts with { (but not with {{, which will actually produce a literal {) and closes with }. With that in mind, we use the regex {{\@! because we don’t want a match if the preceding token is present (see :h \@!).

This region may contain only expressions, so stuff like a function declaration does not make sense (notice there is a syntax for that, :helpgrep ALLBUT)

Handling string modifiers

f-strings supports str.format syntax for formatting, for example:

import mathprint(f"The value of pi is approximately {math.pi:.3f}.")

So I read the Python docs and wrote a regex based on it, but regexes are always easier to write than to read so I wouldn’t recommend you trying.

It’s also possible to convert a value as if wrapping them in functions such as ascii(), repr(), str() with !a, !r, !s respectively, so we need to handle this also.

For this, I declared a syntax group with :h syn-match and pass the regexes that should be used.

It should only be highlighted inside a pythonInterpolation so we take advantage of the containedin argument (see :h syn-containedin).

Highlighting declared groups

Finally, we link these new highlight groups with an appropriate/whichever you like highlight group (see :h hi-link and :h group-name). I chose String for f-strings and PreProc for modifiers. And it should work as expected.