An extremely hacked together script for helping find dependency relationships between libraries in the purescript and purescript-contrib orgs.

You need to provide an api_key.txt file with a GitHub API access token.

Running node go will just create a graph.json containing the dependency edges.

To avoid hammering the API the responses are cached in /cache on the first run, from then on files will be loaded from there if present, so that should probably be deleted each time you come back to this tool.

Running node go some-name will list the dependants of some-name, along with related dependants. For example:

> node go purescript-maps

purescript-argonaut
    purescript-sets
purescript-graphs
    purescript-sets
purescript-sets

Meaning if we're making a breaking change release of purescript-maps then we also need to update purescript-argonaut, purescript-graphs, and purescript-sets, but we shouldn't update purescript-argonaut or purescript-graphs until purescript-sets has also had a version bump first.

I should probably rewrite this in PureScript so I can take advantage of purescript-graphs and just output a toposorted list instead...