Custom Pagination of Cached and Remote Data

I've written a Wordpress plugin that utilizes Vimeo's API to display videos on a user's blog. The plugin retrieves a dataset of 50 videos from the API and caches it to a local file. Then, the plugin will display a portion of those 50 videos (say, 20 videos, but this can be a variable number based on which template is being used, 5, 10 etc.) on each paginated set in the user's template. - When the user clicks for the "next" page of video results, I want to retrieve the next 20 videos from the cached file and show them in the user's template. This might mean either making a request to the cached file, or preloading the cache when the page loads - that's up to you, the expert. - When the user clicks "next" again, the script should recognize that the cached local file does not have all of the data that the user is requesting. Therefore, the script should identify that a new request to Vimeo needs to be made to retrieve the next 50 videos in the dataset. The request endpoints are automatically generated by Vimeo and are returned with each request (so, the first endpoints can be found in the local cached file.) Those next videos (51-100) can be sent to the template, but the template should only show videos 41-60 (41-50 would be in the local cache, and 51-60 would be from the newly-made request.) The new request does not have to be cached to a file. - This can go on for an unlimited amount of videos, with a new request being made to Vimeo each time that the dataset reaches the end of what is local and not cached. - With each paging request, we should also change the URL to include which paginated set the user is currently viewing. This functionality is achieved by using the HTML5 history API. - Long term, I'd like for parameters in the url to bring the visitor to a specific state in the view, for example: visiting [login to view URL]:4 takes you to the fourth page of the dataset visiting [login to view URL] links you to the page in that dataset containing that specific video. In general most of my ideas fall around that concept of pre-warming your cache. Instead of caching when the user requests the data, attempt to anticipate what data they want, and have ways to build a cache after the user has been served data they initially requested. I'm basically looking to recreate the functionality of the gallery this page, but on a user's Wordpress site: [login to view URL] Notice how paging not only loads a new video set, but changes the URL of the page. This functionality is achieved by using the HTML5 history API. I'd like to recreate this as well. Also, visiting [login to view URL]:4 takes you to the fourth page, which I'd also like to achieve. Finally, visiting [login to view URL] links you to the page containing that specific video. I did my best to describe what it is that I am looking for, but I may not be completely on point with the best technical solution to this project. I'm leaving that up to you, because you're smarter than I am and can take the reigns and make decisions for the best software architecture and user experience. You should feel comfortable and have experience with the problems of pagination and how to best solve them, how to use jQuery to handle large datasets, and how to use AJAX and PHP to request and filter through those datasets. Wordpress experience is a definite plus. I have taken time to write this post, so to be sure that you've read and understand the requirements and are willing to take it on, please do not leave a generic message simply saying this will get done. Make me feel confident that you understand what I'm trying to achieve and try to provide evidence of being capable of my project by sharing a past experience that you may have had with a similar project. This will only help us both work best together and give you a better chance at getting the job.