Unleash / unleash-edge / #1322

78%

push

refactor: store less information per streaming client (#618)



This PR refactors the broadcaster and what it stores per client:

    I've created a new StreamingQuery struct that has only the data we need,
    I've moved the token storing from client group level to individual client level (and created a ClientData struct)

I realized when setting up the tests for this that the Query, EdgeToken, and FilterQuery all contain more or less the same bits of data. But all we really need is:

    The actual token string (so that we can boot clients if the token expires)
    Name prefix, projects, and env.

In the Unleash Types Query type, projects and env are optional, but we need them to be present to perform the calculation.

So I created a StreamingQuery struct, which consolidates the data we need from the Query and EdgeToken. I also copied in the methods we use for this elsewhere in Unleash and slightly adapted them. I've added notes inline.

36 of 46 new or added lines in 2 files covered. (78.26%)

1409 of 1891 relevant lines covered (74.51%)

1.83 hits per line