@angadie/chittie-text

Smart text for chittie: print code-page text when the characters are representable, auto-rasterize complex scripts (Sinhala, Tamil, Thai, Devanagari, Arabic, …) via an injected rasterizer, and never silently print ?.

Platform-neutral and dependency-light — it never imports canvas or skia. You supply the rasterizer; this package decides when to use it.

Why: thermal printers have no font ROM or code page for Indic/complex scripts, and those scripts need shaping/reordering. The only correct path is to render the text to a bitmap on the host and send it as an image. chittie-text automates the decision and guards against the silent-? failure.

Install

pnpm add @angadie/chittie-text

API

needsRaster(text, codepage = 'cp437'): boolean

True if any character can't be encoded in codepage (it would print as ?).

import { needsRaster } from '@angadie/chittie-text';
needsRaster('Rs. 250');   // false
needsRaster('ආයුබෝවන්');  // true

TextRasterizer

The interface you implement with a platform rasterizer:

interface TextRasterizer {
  rasterize(text: string, options: RasterOptions): ImageData;
}

A web implementation, for example, draws to a <canvas> and returns ctx.getImageData(...). On React Native, use react-native-skia or a pre-rendered PNG decoded to ImageData.

smartText(encoder, text, options): void

Prints text the right way onto a chittie-core encoder:

import { smartText } from '@angadie/chittie-text';

smartText(encoder, 'මිල: Rs.250', {
  rasterizer,            // optional TextRasterizer
  codepage: 'cp437',     // optional
  raster: { fontSize: 24, maxWidth: 576 }, // passed to your rasterizer
});

• encodable → encoder.text(text)
• not encodable + rasterizer → encoder.image(...)
• not encodable + no rasterizer → throws (no silent ?)

With chittie-react

render(tree, { rasterizer, codepage }) wires this into every <Text> automatically — see @angadie/chittie-react.

License

MIT.