fovea-stream

Incremental fixation-point emphasis for text that arrives in pieces and is already coloured — the transform behind fovea.

npm install fovea-stream

import { BionicTransform } from "fovea-stream";

const transform = new BionicTransform({ style: "dim-tail" });
for await (const chunk of source) process.stdout.write(transform.feed(chunk));
process.stdout.write(transform.flush());

There is a Node stream wrapper too:

import { foveaStream } from "fovea-stream";

process.stdin.pipe(foveaStream()).pipe(process.stdout);

Why this is not just toAnsi in a loop

Transforming a finished string never has to deal with either of these:

• Chunks split anywhere. A word or an escape sequence can straddle the boundary. Incomplete tails are held back and emitted once whole, capped so a long run of non-word bytes cannot stall the stream.
• The text is already styled. SGR 22, which ends a bold stem, also clears dim. Emphasizing one word inside dim text would silently un-dim the rest of the line, so the active attributes are tracked and re-asserted after every stem. Bytes inside escape sequences are never treated as text.

Attributes are held in fixed slots rather than an accumulated string, so a long session with thousands of colour changes stays bounded.

Options

Everything fovea-core accepts, plus:

| Option | Default | Meaning | | --- | --- | --- | | style | "bold" | bold, dim-tail, or both | | codeDetection | "auto" | auto, fences, or off |

codeDetection keeps whole code blocks unemphasized. fences follows Markdown fences, which survive in piped output. auto adds a painted background as a second signal, for agents that have already rendered their Markdown to ANSI and left no fences behind — a heuristic, so right most of the time rather than always. Classification reads only the first few characters of a line, so output is never buffered to the next newline.

dim-tail fades the back of each word instead of bolding the front. On terminals whose bold is a brighter colour rather than a heavier weight, it is often the only one that reads at all.

MIT