LinkGlancer

LinkGlancer is a desktop application for human URL review.

It imports rows from an xlsx workbook into SQLite, opens review targets in a controlled browser session, records manual review results, and exports a new xlsx file with the final data.

Status

• The project is under active development.
• The current architecture is task-driven.
• Windows is the primary supported platform today.
• macOS support is in progress and not yet validated as a release target.

Core Workflow

1. Create a task from an xlsx source file.
2. Select a browser configuration and review settings for that task.
3. Import source rows into the application database.
4. Launch a controlled browser session and confirm the browser is ready.
5. Review URLs, record results, and continue through the task buffer.
6. Export a new xlsx file with the configured output fields.

Tech Stack

• Python
• PySide6
• Playwright
• SQLite
• openpyxl

Development

The repository uses a local virtual environment at .venv.

Typical setup on Windows:

py -3.12 -m venv .venv
.venv\Scripts\Activate.ps1
python -m pip install --upgrade pip
pip install -e .[build]
python -m playwright install
python -m ruff check .
python -m compileall -q src

Run the app:

python -m link_glancer.main

Packaging

Windows packaging assets are included:

• packaging/link_glancer_windows.spec
• scripts/build_windows.ps1

macOS packaging assets are included for ongoing work:

• packaging/link_glancer_macos.spec
• scripts/build_macos.sh
• docs/plans/macos-build.md

License

MIT