# Cerebrum

> OpenWolf's learning memory. Updated automatically as the AI learns from interactions.
> Do not edit manually unless correcting an error.
> Last updated: 2026-06-03

## User Preferences

<!-- How the user likes things done. Code style, tools, patterns, communication. -->

## Key Learnings

- **Project:** netgazet — converts NETgazet print PDFs (InDesign-laid-out, ~A3, 16pp) into a static HTML/CSS site under `output/`. Entry point is `main.py`; runs via `python main.py` or `uv run python main.py`. Inputs live in `examples/`.
- **Conversion approach:** page-as-JPEG via `pdftocairo -jpeg -singlefile -r 150` + responsive `img { max-width: 100% }`. `pdftohtml` was deliberately rejected because its complex-mode output is fixed-pixel and not mobile-responsive. Selectable text is provided as a `<details>` block per page from `pdftotext -layout`, split on form-feed (`\f`).
- **`pdftocairo` filename quirk:** without `-singlefile` it appends `-<page>.jpg` to the output stem; with `-singlefile` it appends just `.jpg`. The script uses `-singlefile` per page so file paths are predictable.
- **No third-party Python deps.** Conversion shells out via `subprocess.run([...], check=True)`; `pyproject.toml` keeps `dependencies = []`.

## Do-Not-Repeat

<!-- Mistakes made and corrected. Each entry prevents the same mistake recurring. -->
<!-- Format: [YYYY-MM-DD] Description of what went wrong and what to do instead. -->

## Decision Log

<!-- Significant technical decisions with rationale. Why X was chosen over Y. -->