Built-in pagination for making it easier to deal with graph-node limitations?

[dotansimha]

The goal of this idea is to have an option to allow users to fetch more than the limit of graph-node, without being concerned about orchestrating calls.

Today, the limit is 1000 records per page.

For example, if a user needs to have more than 1000 (for example 5000), they should be able to specify it in their GraphQL request, and then graph-client can orchestrate the 5 calls to get the list of data, and then merge it into a single response.

[dotansimha]

@ardatan let's explore this option to have an Executor that is aware of this feature? :)

[schmidsi]

Important here is that the recommended approach of paginating through a bigger data set is using the ..._gt filters on numerical attributes like block number or timestamp. Instead of using the first:, skip: arguments.

[dotansimha]

Important here is that the recommended approach of paginating through a bigger data set is using the ..._gt filters on numerical attributes like block number or timestamp. Instead of using the first:, skip: arguments.

Interesting, thanks @schmidsi ! I created a separate issue for that: #48 (since it's a different flow for fetching more data, that happens serially and cannot run concurrently like page-based pagination)

[dotansimha]

This is now available in @graphprotocol/client-cli@0.0.5, docs: https://github.com/graphprotocol/graph-client/blob/main/packages/auto-pagination/README.md