Replace the crawler by an XSLT transformation
Opened this issue · 2 comments
AlainVagner commented
The current way to generate the html definition of a word is to crawl it from the website. This approach is highly inefficient and unreliable but was the fastest one for an MVP. The idea here is to remove the crawler and replace it with an (XSLT?) transformation.
By having more control on the html generation, this would let us improve the accessibility of the generated code, notably for headings and lists.
AlainVagner commented
Maybe the XSLT in this project can be reused:
https://github.com/tigran-a/lod2xdxf
AlainVagner commented
The crawler cannot work anymore, as the website has been totally revamped. Maybe the API can also help here: https://lod.lu/api/doc