This is a tool for searching ipfs, although it doesn't do that good of a job (yet). It uses lunr.js to perform the search, and the search index is built by scraping results from searx.netzplielplatz.de (a public instance of searx, which is itself a metasearch engine, aggregating search results from major search providers). As such, the only thing you're searching here are titles, URLs, and text snippets from about 1100 ipfs pages, because that's the information that's able to be scraped from searx, and those are the pages that have been indexed by major search providers. This is an alpha tool, and shouldn't be considered a good way to search ipfs.

The source for this searching tool is here, and the source for the scraping tool is here. (Be warned, this was hacked together quickly [and I'm a shit coder] so some of the code--espcially the scraping tool--is a bit gross!)

The main thing I'm left wondering after doing this: why are there so many PDFs of 1960's presidential checklists and daily briefings on ipfs?