rahulaloth/goodle

This is a lightweight local search tool that runs on your computer and lets you search through files in a folder you choose, without sending anything to the internet. It scans PDFs, text files, Word documents, images, and CSV files, then gives you a browser-based interface where you can type a keyword and instantly find matching content, even text inside images. The tool is built for people who save lots of files in one place and later struggle to find a specific note, screenshot, or reference document. Students, office workers, and anyone who accumulates files will recognize the problem: you saved something but cannot remember the filename. Goodle indexes the folder and lets you search by content rather than filename. The recommended way to run it is through Docker, a tool for running software in an isolated container. You pull the Goodle image, point it at your folder, and open a browser to a local address. No Python installation is required if you use Docker. For those who want to modify or develop the app, a manual setup path is also available: clone the repository, install Python dependencies, and run the Streamlit app directly. Supported file types include PDFs, plain text files, Markdown files, Word documents, images (PNG and JPG), and CSV files. Excel spreadsheets are intentionally not recommended because they contain structured data rather than plain text. The folder being indexed can be changed at any time by adjusting the Docker volume mount or pressing a Re-Index button in the interface. The project is described as intentionally simple so others can build on top of it. Planned future improvements include faster indexing, more file types, better ranking, and a plugin system. There is a known issue where the import and export feature fails on Windows due to how Docker handles file paths on that platform.