I think that by loading you mean consuming and parsing. Please see article How to parse page-based data using Refinitiv Real-time SDK Java. Notice how a page is comprised of rows, and each row is a field within record? The article comes with downloadable companion example on GitHub, listed at the end.
The same concepts will apply to the task of publication and contribution of pages (record-based).
Next, moving to contribution. It will differ, if you are contributing via your local Refinitiv Real-Time Distribution System (TRDS, formerly TREP) or directly via Contribution Channel.
The article Contributing your data to Refinitiv should be very helpful in explaining the required concepts and decisions involved on the part of the designer and also links to example code.
Hope you will find this information of help.