Fix: Handle error when running init without npm

[GAumala]

Fix init command crashing on machines that don't have npm installed #9102 .

Before:

After:

What is the purpose of this pull request? (put an "X" next to item)

[ ] Documentation update
[ X ] Bug fix (template)
[ ] New rule (template)
[ ] Changes an existing rule (template)
[ ] Add autofixing to a rule
[ ] Add a CLI option
[ ] Add something to the core
[ ] Other, please explain:

What changes did you make? (Give an overview)
Modify fetchPeerDependencies() function in npm-util to retun null if the npm process exited with an ENOENT error.
Modify installSyncSaveDev() function in npm-util to log an error message if the npm process exited with an ENOENT error.
Write new tests for fetchPeerDependencies() and installSyncSaveDev() to cover these scenarios.
Modify the sync stub in other installSyncSaveDev() tests to return an object more similar to the child_process object specified in the node.js docs, to avoid undefined field errors.

Is there anything you'd like reviewers to focus on?

I'm not sure if returning null in fetchPeerDependencies() might be the best way of error handling. A consequence of this is that getPeerDependencies() also has to return null in this scenario. Let me know if you think there are better alternatives.
With this change, hasESLintVersionConflict() could potentially throw an error when trying to access the eslint field of a null object. I decided check for null and use an empty object instead, which results in the function returning false in this scenario. I'm not sure what consequences might this change have, and I couldn't figure out a way to test it since it is a private method.

[eslintbot]

LGTM

[jsf-clabot]

All committers have signed the CLA.

[mention-bot]

@GAumala, thanks for your PR! By analyzing the history of the files in this pull request, we identified @mysticatea, @IanVS and @not-an-aardvark to be potential reviewers.

[not-an-aardvark]

Looks good to me. Thanks for contributing!