[Serializer] Fix AbstractObjectNormalizer to allow scalar values to be normalized

[Hanmac]

Q	A
Branch?	6.4
Bug fix?	yes
New feature?	no
Deprecations?	no
Issues	Fix #58056
License	MIT

It still does work when the Normalizer is used with Serializer (like it normally would?),

[carsonbot]

Hey!

Thanks for your PR. You are targeting branch "6.4" but it seems your PR description refers to branch "7.2 for features / 5.4, 6.4, and 7.1 for bug fixes".
Could you update the PR description or change target branch? This helps core maintainers a lot.

Cheers!

Carsonbot

[carsonbot]

It looks like you unchecked the "Allow edits from maintainer" box. That is fine, but please note that if you have multiple commits, you'll need to squash your commits into one before this can be merged. Or, you can check the "Allow edits from maintainers" box and the maintainer can squash for you.

Cheers!

Carsonbot

[Hanmac]

"Allow edits from maintainer" doesn't work because that is a Group Repository

[andreybotanic]

Your fix is very specific. Why ScalarNormalizer allows only booleans? You probably should create BooleanNormalizer instead. And it also mast supports denormalization, so you have to add appropriate methods and fix demormalization method in AbstractObjectNormalizer

[Hanmac]

@andreybotanic because this is only the start, for #58429 I wanted to add something to format numeric/float/int too,
but I first need to think what the best way would be. Probably "sprintf" Format Strings?

Yeah, denormalization might need to be done too. My use-case was for normalizing into stuff that can be denormalized too. (like "true/false" for boolean.)

I will try to update the denormalize way soon. (for example, "yes/no" for boolean)

[Hanmac]

@nicolas-grekas similar to #59501 can you look into this MR?

I need to find a way to normalize boolean into "true" and "false" instead of "1" and "0"

[Hanmac]

hm, I need to think about the test cases.

For example: testChildContextKeepsOriginalContextCacheKey
thanks to the change, it creates a new child-context for the scalar attributes

[Hanmac]

Finally, the tests are green again 😂

[Hanmac]

for reference, @stof said this in Slack:

looking at a git blame, I suspect that this was working between 2.7 (when serializer added support for normalizing scalars) and 6.1, and got broken in 6.2 (when the handling of scalars was changed in AbstractObjectNormalizer to fix an issue in #49525)

[nicolas-grekas]

Thank you @Hanmac.

[nicolas-grekas]

Merged as c5a661a