Using the RSS feed for the content seems clever, and useful as a means to build the static search widget outside the content generation phase of site building (

Using the RSS feed for the content seems clever, and useful as a means to build the static search widget outside the content generation phase of site building (or for adding after the fact). I recently had a similar (less well fleshed-out) idea[0], but built the search corpus during the generation of my static site (this was made easy by the fact that I wrote the generator too). The results aren't nearly so impressive, but it was an interesting experience that lead to all sorts of interesting ways to improve searching, like stemming, lemmatization[1] and (removing) stop words (this was mostly to reduce the size of the text corpus). [0]: https://idle.nprescott.com/2017/text-search-on-a-static-blog.html https://idle.nprescott.com/2017/text-search-on-a-static-blog... [1]: https://nlp.stanford.edu/IR-book/html/htmledition/stemming-and-lemmatization-1.html https://nlp.stanford.edu/IR-book/html/htmledition/stemming-a...