Chemotion Converter Profiles & Readers Index

This repository builds and publishes a GitHub Pages site that documents Chemotion Converter profiles and readers. It collects metadata from JSON profiles and reader code, enriches profile metadata with CHMO ontology labels, and produces a static HTML page in docs/.

What This Repo Does

• Generates a readable index of available profiles and readers (interactive tables).
• Publishes the index to GitHub Pages on every push to main.
• Mirrors profiles and reader source files into docs/atch/server/ for download links.
• Keeps profile metadata and readers in sync with the Converter package.

Installation

This project depends on the chemotion-converter-app package (installed directly from GitHub) plus local Python tooling.

Requirements:

• Python 3.12
• Git (to install the Converter package from GitHub)

Steps:

python -m venv .venv
source .venv/bin/activate
python -m pip install --upgrade pip
python -m pip install --no-cache-dir -r requirements.txt

The dependency on chemotion-converter-app is declared in requirements.txt and pulled from: git+https://github.com/ComPlat/chemotion-converter-app.

Folder Structure

.
├─ .github/workflows/        # GitHub Actions workflow for build + deploy
├─ build/                    # Generated Markdown artifacts (local output)
├─ data_files/               # Example and sample input files
├─ docs/                     # GitHub Pages output (generated)
├─ profile_manager/          # Build script, templates, and helpers
├─ profiles/                 # Profile JSON files (public + other sets)
├─ readers/                  # Local reader implementations (reference)
├─ requirements.txt          # Python dependencies
└─ README.md                 # Project documentation

Build Workflow

The index is rebuilt whenever profiles change or new profiles are uploaded.

Local rebuild:

python -m profile_manager build_index

This command:

• Reads profiles from profiles/public/.
• Scans reader classes from the installed converter_app package (via importlib.resources).
• Extracts reader metadata (class name, identifier, priority, and check() code) from AST parsing.
• Detects CHMO identifiers anywhere in a profile and fetches the ontology label from OLS4.
• Copies profiles and reader source files into docs/atch/server/ for download links.
• Renders an HTML index in docs/index.html using profile_manager/index_template.html.

Generated Output

Build artifacts are written to:

• docs/index.html (GitHub Pages landing page with interactive tables)
• docs/atch/server/profiles/ (downloadable profile JSON files)
• docs/atch/server/readers/ (downloadable reader source files)

AG Grid (Frontend)

The interactive tables in docs/index.html run entirely in the browser. AG Grid is not bundled in this repo; it is loaded at runtime from a CDN via external <link> and <script> tags that are injected by the build step (profile_manager.__main__.py in dict_to_ag_grid_html).

GitHub Pages Workflow (Rebuild on Push)

The GitHub Actions workflow in /.github/workflows/build_and_deploy.yml runs on:

• Every push to main.
• Manual workflow dispatch.

On each run, it installs dependencies, builds the index, and publishes the contents of docs/ to GitHub Pages. This means every profile upload or change that gets pushed to main automatically triggers a rebuild and redeploy.

---

Created: 2026-01-22
Last updated: 2026-01-29