enforce length n serial distribution

[sbfnk]

Fixes #111

[sbfnk]

@jamesmbaazam this fails snapshot tests which should not be affected by this PR. Do we need to set a seed in (some of) the testing code?

[jamesmbaazam]

There's something weird going on here that I am yet to diagnose. seeds have been set in each test_that() call but it seems they're being lost after each function call. I'm not getting the same errors in main and not sure why it's happening here. Will investigate further.

[jamesmbaazam]

Thanks for these changes, Seb. I've left a comment to hopefully help us fix the CI issue.

[sbfnk]

There's something weird going on here that I am yet to diagnose. seeds have been set in each test_that() call but it seems they're being lost after each function call. I'm not getting the same errors in main and not sure why it's happening here. Will investigate further.

Ohhh is the issue that in the tests we're generating a new random number (in check_serial_valid) and therefore all subsequent calls to random number generator calls different (pseudo-)random numbers than without that call. In that case all I need to do is accept the snapshot, right?

[jamesmbaazam]

Ohhh is the issue that in the tests we're generating a new random number (in check_serial_valid) and therefore all subsequent calls to random number generator calls different (pseudo-)random numbers than without that call.

Yes, exactly.

In that case all I need to do is accept the snapshot, right?

Yep. That should fix it.

[jamesmbaazam]

LGTM. I will rebase and merge.

One thing I've not enforced properly is to keep the version at v0.0.0.9999 until we release v0.1.0.

[jamesmbaazam]

Also, note incoming change in #116 to rename serials_dist to gen_interval.