feat(webpack-cli): import flags from webpack core

[snitin315]

What kind of change does this PR introduce?
Feature.

Did you add tests for your changes?
Yes
If relevant, did you update the documentation?
No
Summary

Follow up #1629
Refers #717
Closes #1241

Does this PR introduce a breaking change?

Other information

TODO

• Write tests
• remove duplicate flags
• register --no-name negated flags.
• fix help output ( currently showing type string for every flag )
• allow parsing type: 'number'

[snitin315]

Right now works well -

[snitin315]

list of broken flags -

• --amd fixed
• --module-no-parse
• --module-no-parse-reset
• --module-rules-side-effect
• --output-library-reset
• --output-library-* related flags
• --watch-options-poll

[snitin315]

Draft list of flags we can drop ( need discussion ) -

• all flags related to module.rules i.e, --module-rules-use module-rules-use-options etc. (total 18 flags) because it ts not feasible to configure such options via CLI https://webpack.js.org/configuration/module/#ruleuse

[anshumanv]

I'm not sure of the feasibility of removing this on the core, should we keep a list of unsupported flags and remove them after importing core flags?

[snitin315]

I'm not sure of the feasibility of removing this on the core, should we keep a list of unsupported flags and remove them after importing core flags?

Yes, I think we can filter flags on the CLI side.

[snitin315]

--amd is broken on our side. it accepts only false value and right now CLI doesn't support parsing of boolean flags like webpack-cli --amd=false.
Looking into it if it's possible to parse it like this with commander.js

[alexander-akait]

@snitin315

--amd is broken on our side. it accepts only false value and right now CLI doesn't support parsing of boolean flags like webpack-cli --amd=false.

Not sure we should support it, because we have --amd and --no-amd support, boolean as string maybe some misleading, for example --entry=false can be mean:

• loading false.js as entry point
• no entry configuration (possible with webpack@5)

What we should use? I don't know, so better:

• --no-entry means no entry
• --entry=false uses false.js for entry

[snitin315]

@evilebottnawi I think we can specify --no-amd in cli-flags and we will be good.

[alexander-akait]

@snitin315 For all (not sure for all but for most of them) boolean flag we should register --no-name flag

[snitin315]

Let's implement --no-name flag for all now, we can filter them later (if needed).

[alexander-akait]

@snitin315 Can we move these changes:

• allow parsing type: 'number'
• fix help output ( currently showing type string for every flag )

to other PRs, to avoid unnecessary changes in this PR?

[rishabh3112]

Looks good.

[anshumanv]

💯

[ghost]

There were the following issues with this Pull Request

• Commit: a006725
  • ✖ message may not be empty
  • ✖ type may not be empty

You may need to change the commit messages to comply with the repository contributing guidelines.

---

🤖 This comment was generated by commitlint[bot]. Please report issues here.

Happy coding!

[ghost]

There were the following issues with this Pull Request

• Commit: a006725
  • ✖ message may not be empty
  • ✖ type may not be empty

You may need to change the commit messages to comply with the repository contributing guidelines.

---

🤖 This comment was generated by commitlint[bot]. Please report issues here.

Happy coding!

[webpack-bot]

@jamesgeorge007 Thanks for your update.

I labeled the Pull Request so reviewers will review it again.

@anshumanv Please review the new changes.

[anshumanv]

@jamesgeorge007 please don't commit yourself, suggestions suffice

[alexander-akait]

@webpack/cli-team Merge?

[anshumanv]

@webpack/cli-team Merge?

yep we can merge 👍

[alexander-akait]

Big thanks for all, it is great feature, I think we should resolve serve command and do release, when focus on perf and refactor