Sherlock Search

A while back I created this demo as a proof of concept to show how Algolia could index huge amount of text. I took all the Sherlock Holmes book from the Project Gutenberg website, and push them to an index.

I parsed HTML files (each file had a slightly different markup, forcing me to write different parsers) to generate records. Each record is a paragraph of the file, enriched with the author, book, chapter and page number.

I also included some “context” in the form of the previous and next paragraph. This, coupled with some front-end logic allowed me to group the display of several results if they were supposed to be consecutive paragraphs.

I did not spend much time on the display and re-used an old design. I’d like to try to add some skeuomorphism to this one to give it a library/old book feeling.

All the code is available on GitHub.

1 Like

I hadn’t seen this “experiment” :smile: Really cool and thank you for reminding me about the Gutenberg project, I need exactly that kind of test data for a project I’m working on :wink: