Manuals publisher content doesn't account for accordion usage on manuals frontend

[owenatgov]

Manuals content is often organised into accordions, however this isn't accounted for at the publisher end and is instead hacked in via manuals frontend. At the moment, content is written and stored in continuous prose in manuals publisher, specifically as headings with associated content, and passed to manuals frontend as one mass of HTML. This content is then pre-processed to convert each heading and content pair into a data item that can be digested by an accordion before being presented to the user. This approach has lead to an until recently undetected divergence between the supported govuk accordion component and the accordion being used on manuals frontend, creating hidden tech debt and accessibility risk.

Ideally this pre-processing wouldn't need to be done in manuals frontend at all and instead would be handled by manuals publisher in how publishers actually construct content and how that data is handled and passed to manuals frontend. An example of an app that accounts for the separation of content via accordions (see content item).