refactor(errors): one file per error

[papb]

Pull Request check-list

Please make sure to review and check all of these items:

• Does npm run test or npm run test-DIALECT pass with this change (including linting)?
• Does the description below contain a link to an existing issue (Closes #[issue]) or a description of the issue you are solving?
• Have you added new tests to prevent regressions?
• Is a documentation update included (if this change modifies existing APIs, or introduces new ones)?
• Did you follow the commit message conventions explained in CONTRIBUTING.md?

Description of change

All errors were defined in a single huge file. This PR refactors it so that every error has its file, and uses a folder structure for the nested error types.

I encountered this when I was about to contribute with a new error, but I felt like refactoring this first.

[SimonSchick]

I don't really see the advantage of this, the error logic is far from complicated and all this does is increase load time.

[codecov]

Codecov Report

Merging #10355 into master will decrease coverage by 6.56%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master   #10355      +/-   ##
==========================================
- Coverage   96.28%   89.71%   -6.57%     
==========================================
  Files          68       90      +22     
  Lines        9775     9726      -49     
==========================================
- Hits         9412     8726     -686     
- Misses        363     1000     +637

Impacted Files	Coverage Δ
lib/errors/bulk-record-error.js	100% <100%> (ø)
lib/errors/connection-error.js	100% <100%> (ø)
lib/errors/database-error.js	100% <100%> (ø)
lib/errors/connection/access-denied-error.js	100% <100%> (ø)
lib/errors/database/timeout-error.js	100% <100%> (ø)
lib/errors/instance-error.js	100% <100%> (ø)
lib/errors/empty-result-error.js	100% <100%> (ø)
lib/errors/connection/host-not-reachable-error.js	100% <100%> (ø)
lib/errors/association-error.js	100% <100%> (ø)
...ib/errors/database/foreign-key-constraint-error.js	100% <100%> (ø)
... and 51 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 0f0255e...0c301e8. Read the comment docs.

[sushantdhiman]

Yeah I feel this file is too big and can be refactored without much impact on library. One change I would recommend to get rid of @extends directive from all errors for esdocs so they look like this

[papb]

@SimonSchick - I just tested on my machine here, using time-require:

• It took 345.7 ms on average (out of 10 samples) without this PR
• It took 354.5 ms on average (out of 10 samples) with this PR

I didn't compute the standard deviation for each case explicitly but the values ranged from ~330 to ~370 on both cases. To be honest I was surprised, I thought the load time would increase a lot more! Regardless, I believe this is a very small price to pay (the difference is eaten by the standard deviation!) for better readability and maintainability.

@sushantdhiman Nice catch! You're right. By removing it, esdoc parses correctly the inheritance chain by default! Nice! I just did it.

[SimonSchick]

@papb that is weird, when I did the mass cleanups before it increased load times by almost 10% which is why I omitted it from my PR, I guess this depends on the system it's running.

My previous test was using console.time on node v10 btw.

You should add trailing newlines btw.

Also @sushantdhiman what's the deal with the manual captureStackTrace, is it just to omit a stack frame?

[papb]

@SimonSchick - I see... I am on windows on Node 6 (old, I know, I'll do something about it later).

I've just added the trailing newlines as requested 😁

[sushantdhiman]

what's the deal with the manual captureStackTrace, is it just to omit a stack frame?

@SimonSchick Looks like that

[SimonSchick]

@sushantdhiman as a future note, this can be removed once support for v6 is dropped, iirc it was a workaround for prototype shenanigans pre v4.