fix: improved session handling and logging

[kantorcodes]

Description:
Broken off from #318

This PR suggests a number of fixes to improve the dApp experience:

1. Enables dApps to specify a log level for better control of console logs.
2. Ensures that there is only one Signer per AccountId + ExtensionId combination. Having multiple signers with different ids created issues in pulling the correct record.
3. Explores better state management and automatic removal of stale sessions.

Improves

Related issue(s):

#254

Fixes #

Notes for reviewer:

This PR can be tested individually by installing the canary release:

Checklist

• Documented (Code comments, README, etc.)
• Tested (unit, integration, etc.)

[github-actions]

Coverage report

St.❔	Category	Percentage	Covered / Total
🟡	Statements	76.29% (+37.69% 🔼)	444/582
🟡	Branches	63.48% (+35.18% 🔼)	73/115
🟡	Functions	75.69% (+50.13% 🔼)	109/144
🟡	Lines	77.28% (+37.59% 🔼)	415/537

Show new covered files 🐣

St.❔	File	Statements	Branches	Functions	Lines
🟢	index.ts	100%	100%	100%	100%
🟢	lib/index.ts	100%	100%	100%	100%
🟢	lib/shared/index.ts	100%	100%	100%	100%
🟢	... / chainIds.ts	100%	100%	100%	100%
🔴	... / errors.ts	20%	0%	0%	25%
🟢	... / events.ts	100%	100%	100%	100%
🟢	... / methods.ts	100%	100%	100%	100%
🟢	lib/shared/utils.ts	86.96%	66.67%	86.21%	88.46%
🔴	... / extensionController.ts	30.77%	11.11%	16.67%	28.57%
🔴	lib/wallet/index.ts	37.74%	28.57%	50%	39%
🔴	... / provider.ts	38.46%	100%	27.27%	38.46%
🟢	lib/dapp/index.ts	86.83%	77.5%	83.33%	87.83%
🟢	... / logger.ts	100%	100%	100%	100%
🟢	... / DAppSigner.ts	100%	86.96%	100%	100%

Test suite run success

110 tests passing in 11 suites.

Report generated by 🧪jest coverage report action from d723d56

[codecov]

Codecov Report

Attention: Patch coverage is 8.00000% with 46 lines in your changes missing coverage. Please review.

Please upload report for BASE (main@5ae78a5). Learn more about missing BASE report.

Files with missing lines	Patch %	Lines
src/lib/dapp/index.ts	8.16%	45 Missing ⚠️
src/lib/dapp/DAppSigner.ts	0.00%	1 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #327   +/-   ##
=======================================
  Coverage        ?   36.05%           
=======================================
  Files           ?       14           
  Lines           ?      613           
  Branches        ?       79           
=======================================
  Hits            ?      221           
  Misses          ?      392           
  Partials        ?        0           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[franfernandez20]

I don’t see what has been split here, it still looks like code styling changes along with logging and some session management updates. The updates look solid, however, I still recommend splitting this PR into smaller, more focused ones for clarity. That said, I’m open to moving forward if others feel differently—I don’t want to hold up progress.

[tmctl]

I don’t see what has been split here, it still looks like code styling changes along with logging and some session management updates. The updates look solid, however, I still recommend splitting this PR into smaller, more focused ones for clarity. That said, I’m open to moving forward if others feel differently—I don’t want to hold up progress.

I think a net of 170 lines of code and 6 files changed is an okay sized PR. Here's an example PR into the javascript sdk that is similarily sized - https://github.com/hashgraph/hedera-sdk-js/pull/2631/files

I'd say we can probably move forward with this size PR

[justynspooner]

The code updates look good to me. It's hard to be confident with the updates though without test coverage especially considering it's mitigating against multiple connection states which would be hard to reproduce. Are there some tests we can add to help raise confidence with these fixes?

[kantorcodes]

The code updates look good to me. It's hard to be confident with the updates though without test coverage especially considering it's mitigating against multiple connection states which would be hard to reproduce. Are there some tests we can add to help raise confidence with these fixes?

Great flag. Just pushed a heap of tests to the PR. Could you check again and let me know your thoughts? We'll also have coverage on the connector which was previously not running.

[kantorcodes]

Note, these callbacks were refactored in order to make it easier to unit test

[jwtong]

Looks good to me. Is it possible to move this past the finish line quickly and merge it? It’ll be a major help for us at Saucerswap to address WC reliability on our platform. Thanks!