Currently webmention.io uses XRay to parse a page, then picks out individual parts of the XRay response into the database, and then when queried in the API, puts it back together in a format that looks very similar to XRay.
This causes problems, e.g. #177, as well as having to add special support for new post types even after XRay can already parse them.
Instead, webmention.io should store the raw XRay response, only pulling out fields when absolutely required, and use the raw XRay response when building the API response.
Currently webmention.io uses XRay to parse a page, then picks out individual parts of the XRay response into the database, and then when queried in the API, puts it back together in a format that looks very similar to XRay.
This causes problems, e.g. #177, as well as having to add special support for new post types even after XRay can already parse them.
Instead, webmention.io should store the raw XRay response, only pulling out fields when absolutely required, and use the raw XRay response when building the API response.