@tarogo/document-ir-parser
v0.2.2
Published
Optional Node parser adapter for structured document IR.
Maintainers
Readme
@tarogo/document-ir-parser
Optional Node adapter for parsing local documents into structured Document IR.
This package is intentionally separate from the browser-safe @tarogo/document-ir core package. The core package provides types, validation, outline helpers, and section conversion. This adapter provides the Node-only parser entry points.
Install
npm install @tarogo/document-ir-parserUse
import { parseDocumentIr } from "@tarogo/document-ir-parser";
const document = await parseDocumentIr("sample.docx");The adapter also provides a CLI:
document-ir-parser parse sample.docx -o sample.ir.jsonVersion 0.2.x owns the Node API, CLI, Python sidecar, and parser parity fixtures. It declares @tarogo/document-ir ^1.0.0 as a peer dependency so consumers install the browser-safe core separately; the repository's local development link is not included as a runtime package dependency. The parser implementation is bundled as an optional adapter and does not require an AIQC host repository PYTHONPATH; Python and format-specific runtime dependencies remain environment prerequisites. The adapter does not read AIQC Admin configuration or host storage.
Python, parser-library, PDF-engine, and optional OCR runtime requirements depend on the selected parser adapter. Browser applications should use @tarogo/document-ir instead.
The current source location is temporary:
jb-aigc-aiqc/packages/document-ir-parserIt will move with the public package source to an independent repository after the adapter/core split is accepted.
