feat(dockerfile): add support for escape chars and ARG instructions

[Churro]

Changes

This rework of the dockerfile extractor causes image parts in ARG values (referenced in FROM statements) to be assembled to one string value. The code in this PR enables auto updates for images that occur as a whole within an ARG value (example here) or split into multiple ARG values (example here).

• Added support for parsing and evaluating ARG name/value definitions that are referenced in FROM instructions
  • As their sequence matters, parsing was changed to a line-wise approach. This is favorable also to prevent false negative matches that currently occur due the position-independent RegEx search, e.g.,:

    COPY --chown=root --from=image11 / ./ # image11 is detected as stageName, although defined only afterwards
    FROM nginx as image11

• Added support for ` as escape character for multi-line instructions in Dockerfile
• Slightly simplified FROM and COPY Regex patterns
  • Improved to cover additional, legit cases:

    FROM \
    a as stage1

    and

    FROM \ # tab or whitespace after \
    a as stage1

  • Side note: not sure how to reproduce refactor: unexpected behaviour when re2 is used #12875. The used RegExes don't include expressions marked as unsupported here and tests also work with re2 but to be on the safe side, I kept RegExp.

No existing tests were changed/rewritten, therefore I don't expect any breaking changes.

Known Limitations

• No depName replacements
  • Replacing deprecated dependencies for docker datasource only replace version #12623)
• No variable shadowing for the subsequent stage: ONBUILD ARG name=value
• Nested variables (example here)
• No advanced parameter interpolation, e.g.,

 ARG FROM_REGISTRY=ghcr.io
 ARG FROM_IMAGE=${FROM_REGISTRY:+$FROM_REGISTRY/owner/sample-image}
 FROM ${FROM_IMAGE:-scratch}

Context

• Solves Support parsing and evaluating build arguments specified before the FROM #5626
• Solves fix(dockerfile): improve FROM/COPY regex matching #8331 (see viceice's comment regarding escapeChar)

Documentation (please check one with an [x])

• I have updated the documentation, or
• No documentation update is required

How I've tested my work (please tick one)

I have verified these changes via:

• Code inspection only, or
• Newly added/modified unit tests, or
• No unit tests but ran on a real repository, or
• Both unit tests + ran on a real repository

Test repositories:

• https://github.com/Churro/renovate-dockerfile-test
• https://github.com/Churro/renovate-dockerfile-test2

[PhilipAbed]

LGTM, but i'm not 100% familiar with docker files and this flow so others should review too.

[rarkins]

Approving based on the quality of code, extensiveness of tests, and the fact that existing tests were left as-is, hopefully indicating backwards compatibility

[zharinov]

Probably, it's the last set of changes before resorting to tokenizer/parser

[rarkins]

@Churro thank you for this amazingly good first-time contribution, both with quality of code as well as extensive testing. I will leave to @viceice to merge once he is available to confirm he has no further comments.

[renovate-release]

🎉 This PR is included in version 32.75.0 🎉

The release is available on:

• GitHub release
• 32.75.0

Your semantic-release bot 📦🚀