Capturing escape sequence sizing with categorized slices

[ncatelli]

What would be the best way to derive the escape code by size of a give CategorizedSlice? Having this information would be extremely useful for tasks like text wrapping.

[kurtlawrence]

Hi @ncatelli.

CategorisedSlice has a .text field which is a reference to the underlying string (without the escape codes). You could call .chars() on it to get a character 'width'.

Is that sufficient for your use case?

[ncatelli]

Hi Kurt,

Unfortunately no, specifically I was looking for the best way to either:

• Tie the original text, including the escape codes back to the categorized slice.
• Include the byte size of the escape codes for a given categorized slice.

That could be used, in conjunction with the .text field to gain the real width of a CategorizedSlice over its input text.

[kurtlawrence]

I think you could use the fact that the CategorisedSlice has start and end fields. Each slice can be considered to contain a leading 'style' which is the escape code.

For example, take the following:

let x = format!("{}, {}!", "Hello".blue().on_green(), "world".red().on_yellow());
println!("{x:?}");

let [hello, comma, world, bang]: [v3::CategorisedSlice; 4] = 
v3::categorise_text(&x).try_into().unwrap();

let hello_rng = 0..hello.end;
let comma_rng = hello.end..comma.end;
let world_rng = comma.end..world.end;
let bang_rng = world.end..bang.end;

dbg!(&x[hello_rng]);
dbg!(&x[comma_rng]);
dbg!(&x[world_rng]);
dbg!(&x[bang_rng]);

It will print out:

"\u{1b}[42;34mHello\u{1b}[0m, \u{1b}[43;31mworld\u{1b}[0m!"
[src/tests.rs:531] &x[hello_rng] = "\u{1b}[42;34mHello"
[src/tests.rs:532] &x[comma_rng] = "\u{1b}[0m, "
[src/tests.rs:533] &x[world_rng] = "\u{1b}[43;31mworld"
[src/tests.rs:534] &x[bang_rng] = "\u{1b}[0m!"

Which captures the original escape code sequences.

If you need a 'trailing' escape code sequence (say \u{1b}[42;34mHello\u{1b}[0m), you could use 0..comma.start.