feat: upgrade to @hapi

[mkg20001]

This upgrades all packages and switches namespaces for old hapi packages to new namespace, to fix warnings and security issues.

[mkg20001]

Joi seems to have ditched the Joi.validate error which makes some tests fail

Also sinon is complaining about double-wrapped methods

Among other random test failures

[JKHeadley]

@mkg20001 This is great! Thanks for handling this. Definitely good to have loggin updated until #150 is handled.

Once we can get these tests fixed I'll go ahead and merge.

[mkg20001]

@JKHeadley Some deps are still downgraded (since I tried fixing the tests this way, but seems that most are broken because of hapi API changes). Also joi-objectid has a problem with the new joi, so that is forked now as well.

Btw, I have sent an email to the owner of the loggin package and CC'd npm support, so if it won't get updated I'll possibly get to take it over next month.

I'll do the same one for joi-objectid, since peeble is now bought by fitbit and I doubt they're doing much opensource for it now.

[JKHeadley]

@mkg20001 Awesome about loggin and joi-objectId, def let me know how it goes!

[mkg20001]

@JKHeadley I now own loggin (and the underlying gh-organization fistlabs)

In case of joi-objectid: I'll let the pr idle until the end of the month before contacting support.

(Edit: I've saved the email as a draft, could you possibly remind me in a few weeks to send it?)

[mkg20001]

@JKHeadley Besides that, could you look into the tests? I'm clueless as to what is broken, but since you've written it you might be able to get a better idea.

[JKHeadley]

Great news! Yeah I'll try to look into the tests.

[mkg20001]

Note that a possible bugfix for joi errors that include something about $._.alternatives is to wrap the Joi object {key: Joi.string()} with the joi function Joi.object({key: Joi.string()})

[codecov]

Codecov Report

Merging #201 into master will not change coverage.
The diff coverage is 97.21%.

@@          Coverage Diff           @@
##           master    #201   +/-   ##
======================================
  Coverage    94.3%   94.3%           
======================================
  Files          51      51           
  Lines       11827   11827           
======================================
  Hits        11154   11154           
  Misses        673     673

Impacted Files	Coverage Δ
tests/unit/handler-helper.tests.js	98.15% <ø> (ø)	⬆️
tests/unit/rest-helper-factory.tests.js	97.59% <ø> (ø)	⬆️
tests/e2e/basic-non-embed.tests.js	99.7% <100%> (ø)	⬆️
tests/unit/query-helper.tests.js	99.18% <100%> (ø)	⬆️
policies/track-duplicated-fields.js	92.5% <100%> (ø)	⬆️
tests/e2e/audit-log.tests.js	100% <100%> (ø)	⬆️
...o-3/models/linking-models/user_permission.model.js	100% <100%> (ø)	⬆️
policies/enforce-document-scope.js	92.25% <100%> (ø)	⬆️
tests/e2e/duplicate-field.tests.js	100% <100%> (ø)	⬆️
...e2e/test-scenarios/scenario-1/models/role.model.js	100% <100%> (ø)	⬆️
... and 41 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update ef5d919...bf9e261. Read the comment docs.

[mkg20001]

I've sent out the email to joi-objectid now instead of waiting forever. Response is that it's a bit complicated since the company that owns it is now gone.

[JKHeadley]

Pushed some changes to fix the tests. I'll leave it up to you, but I think for now keeping Joi to v14 will be a good idea so we can merge the rest. The jump from 14 to 16 looks to be pretty major:
hapijs/joi#2037

Maybe we can tackle that one in a different PR.

[mkg20001]

Sure, sounds pretty reasonable
Since they're both backwards compatible this shouldn't be much of an issue
Edit: That is, I'd also support doing this in a different pr

[mkg20001]

So LGTM?

[JKHeadley]

Merged! Thanks for all the work! 👍