Bjolang reads """...""" as a raw string: three or more quotes open it, the
same number close it, and nothing in it is an escape. The grammar has a
string.quoted.triple.bjolang scope for it, so quotes inside one stay inside,
and one over several lines is a string on every line. tests/demo.bjo checks
both, and passes.
🤖 Generated with [ECA](https://eca.dev) (anthropic/claude-opus-5-5)
Co-Authored-By: eca-agent <git@eca.dev>
TextMateSharp runs VS Code's TextMate grammars and themes; this hands
the result over once, as Bjolang values, so nothing using it needs to
know TextMateSharp is there. `tokenize` gives each line's tokens and
their scopes, `highlight` each line's spans in the theme's colours, and
`highlight->node` and `code->node` give (text xml) nodes: a pre and a
code element around spans, coloured inline by the theme or by class
names taken from the scopes. `highlighter-css` writes the stylesheet
for the classes from a theme, choosing each class's rule the way
TextMateSharp's tokenizer chooses a token's.
Bjolang's own grammar is in grammars/bjolang, laid out as a VS Code
extension, and every highlighter reads it from beside the package's
src/. It follows what Bjolang's reader reads, including interpolated
strings with expressions in their holes.
There is no C# shim: everything goes through import/class and
import/extern. The package is kept within TextMateSharp 2.0, because
token colours are decoded with a class in its Internal namespace.
🤖 Generated with [ECA](https://eca.dev) (anthropic/claude-opus-4-7)
Co-Authored-By: eca-agent <git@eca.dev>
