Sign inSign up

datapsycho/oh-my-liteparser

By datapsycho

•Updated 6 months ago

Batteries-included base runtime image with Node.js 22, LiteParse and friends.

Image
Machine learning & AI
Data science
0

1.3K

datapsycho/oh-my-liteparser repository overview

⁠Oh My LiteParser

datapsycho/oh-my-liteparser is the canonical batteries-included base image for the oh-my-liteparser document-parsing stack. It bundles every native runtime dependency required to convert and parse supported document formats without any network fetches or additional installs at container startup.

It is not meant to be used directly as an application image. Instead, it serves as the FROM layer for:

  • datapsycho/oh-my-liteparser-cli — the developer-facing CLI image
  • datapsycho/oh-my-liteparser-lambda — the AWS Lambda container image
⁠Base Features

Parsing multi-format documents (PDF, DOCX, PPTX, XLSX, images) inside a container typically requires assembling a fragile combination of native libraries, fonts, OCR models, and environment configuration. Getting all of these to coexist reliably is time-consuming and error-prone.

This base image solves that problem once, centrally. Any image that extends it inherits a fully configured, production-ready parsing environment without any extra setup.

⁠What Is Included
ComponentPurpose
Node.js 22 (Debian 12 bookworm-slim)Runtime
@llamaindex/liteparseDocument parsing engine
LibreOffice (headless)DOCX, PPTX, XLSX, ODT, RTF, ODS, ODP, CSV
ImageMagick + GhostscriptJPG, PNG, TIFF, WEBP, SVG, BMP, GIF, PDF rasterisation
Tesseract eng.traineddata (pre-downloaded)Offline OCR for scanned documents
fonts-liberationConsistent rendering of Office documents
Build toolchain (cmake, make, g++, autoconf)Native module compilation support

Environment variables set at build time:

TESSDATA_PREFIX=/app/tessdata
LITEPARSE_TMPDIR=/tmp/liteparse
NODE_ENV=production
⁠How to Use It

Pull the image:

docker pull datapsycho/oh-my-liteparser:1.5.2
# or
docker pull datapsycho/oh-my-liteparser:latest

Extend it in your own Dockerfile:

FROM datapsycho/oh-my-liteparser:1.5.2

WORKDIR /app
COPY dist ./dist

ENTRYPOINT ["node", "dist/your-entrypoint.js"]

Verify the runtime (quick smoke test):

docker run --rm datapsycho/oh-my-liteparser:1.5.2 node --version
docker run --rm datapsycho/oh-my-liteparser:1.5.2 libreoffice --version
⁠Tags
TagDescription
latestMost recent stable build
1.5.2Pinned release aligned with LiteParse v1.5.x
⁠Platform

linux/amd64 only.

⁠Credits

The document parsing capability in this image is powered by LiteParse⁠, developed by the LlamaIndex team. LiteParse provides structured, page-aware parsing for a wide range of document formats including PDF, DOCX, PPTX, XLSX, and images, with optional OCR support. All credit for the parsing engine goes to the LlamaIndex LiteParse contributors.

This image packages LiteParse together with the system dependencies it needs to run inside a container, making it straightforward to integrate document parsing into any Node.js container workflow.

Tag summary

Content type

Image

Digest

sha256:d03df8c18…

Size

449.7 MB

Last updated

6 months ago

docker pull datapsycho/oh-my-liteparser