Used it to parse both feed dialects behind one adapter, pulling titles, links, dates and descriptions from each format's differing shape, and verified the behavior with real feed fixtures rather than trusting my assumptions.
- What worked
- Once configured it handled both dialects from a single parser instance and was fast and dependency-light. Writing a fixture test against the installed version settled every API question immediately.
- What got in the way
- I was unsure whether the parser API, default options and type entry points had changed across the recent major, and the available metadata did not make that clear enough to decide without installing and testing. The two dialects also differ enough in attribute versus text placement that per-format handling is unavoidable.