test(publisher-ers): fix fs-extra mocks

[erickzhao]

• I have read the contribution documentation for this project.
• I agree to follow the code of conduct that this project follows, as appropriate.
• The changes are appropriately documented (if applicable).
• The changes have sufficient test coverage (if applicable).
• The testsuite passes successfully on my local machine (if applicable).

Summarize your changes: #2766 added some tests to the trunk that were broken because #2706 added new code paths that weren't accounted for in the mocks.

[codecov]

Codecov Report

Merging #2813 (656b65e) into master (4b07c1f) will increase coverage by 1.71%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master    #2813      +/-   ##
==========================================
+ Coverage   73.90%   75.62%   +1.71%     
==========================================
  Files          77       77              
  Lines        2357     2359       +2     
  Branches      436      438       +2     
==========================================
+ Hits         1742     1784      +42     
+ Misses        465      423      -42     
- Partials      150      152       +2     

Impacted Files	Coverage Δ
...lisher/electron-release-server/src/PublisherERS.ts	86.95% <100.00%> (+60.09%)	⬆️

---

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 6069ebe...656b65e. Read the comment docs.

[clavin]

Looks good, though one question: was the package-lock.json file for a test fixture meant to be committed?

[erickzhao]

@clavin yep, it's a bolt monorepo thing. I imported a sinon stub into the tests to mock out fs-extra, but sinon wasn't installed in the package's dependencies. Without this lockfile change, the linter detects the import as extraneous.

https://github.com/mysticatea/eslint-plugin-node/blob/master/docs/rules/no-extraneous-import.md