OpenMRS pagination

[josephjclark]

We've hit on a couple of problems with OpenMRS pagination.

In 4.3.0, if you do searchPatient({ q: 'foo', limit: '100' } (does limit really have to be a string??) , it'll auto-page all data in pages of 100. It will not download 100 search results and return.

In other words, limit is actually working as pageSize, not limit, which is super misleading.

Against the new 5.0 branch, what we need to ensure is this:

• The REST API should basically auto-page by default to download the desired number of records. It should provide limit pageSize and offset options. These do not map directly to url query parameters. pageSize controls how many are downloaded per page. limit controls how many in total are downloaded. The API needs to take care to not download more than the limit. I think by default we should download everything? Note that just 400 patient records seems to blow up the worker (although we're working on that)
• The http.* namespace should not auto-page by default. It should do whatever you ask of it.If you pass query.limit, it'll just send that to the server and give you a page of data
• later we can add either a autopage: true option to http.get? Or maybe we support options.limit and options.pageSize, and for http.get we'll honour them for you?. But if we disable autopage by default for http, we can add options to enable it as minor patches later.

[josephjclark]

@PiusKariuki I'd like you to look at this too please. Is the issue clear? Maybe we should talk it over?

[PiusKariuki]

Hi @josephjclark . Yeah I'd like to talk this over and see how we can solve this. Are you free for a call?

[josephjclark]

Great. I think it has to be tomorrow now, I need to get my head down on a few things. I'll propose a meeting time now

[PiusKariuki]

Awesome. I got the link. See you then.

[josephjclark]

To util we'll add

request()
paginatedRequest()

We should log with query params every individual request (shows a heartbeat to users)

Ensure each request logs with query parameters