AyaMe

May 2021

Context

Working with some XML-based synoptic, I had to often search for a specific sensor or filter them by location.

I wrote an XML parser in Python that would extract information from files, transform them into a search-friendly format and finally index them into a small database.

Realisation

Using some fetch promise I gathered the up-to-date index and loaded it into a really simple Svelte application. Using the LunrJS search engine, I built a simple pagination and ranking system.

TailwindCSS helped me to get a modern-looking interface and quick responsiveness. Their list component worked well with the results list and allowed me to set a structured page that have nowadays best practises.

Information update

Last step was to create an automatic task that would regenerate the database on demand. Since all the site is static with no server-side generation, I had to thinker on how to update the main database. I changed my simple script to be called remotely using XMLRPC tasks protected by passwords.

Finally, I created a simple front-end for the script using Python TKinter, that would simplify running script management. I re-used it on the REImu Watch project to automate its generation behaviour since both shared the same architecture !