Wizrad: add condition to useHostnameValidation in case of empty string

[mgold1234]

this adds condition to useHostnameValidation in case of empty string if user decide to delete the value in hostname field
FIX ISSUE: 2917

[codecov]

Codecov Report

Attention: Patch coverage is 90.32258% with 3 lines in your changes missing coverage. Please review.

Project coverage is 82.09%. Comparing base (06f4ce4) to head (c18a152).
Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
src/Components/CreateImageWizard/validators.ts	83.33%	2 Missing ⚠️
...rc/Components/CreateImageWizard/ValidatedInput.tsx	93.75%	1 Missing ⚠️

@@           Coverage Diff           @@
##             main    #2872   +/-   ##
=======================================
  Coverage   82.09%   82.09%           
=======================================
  Files         207      207           
  Lines       23291    23313   +22     
  Branches     2285     2290    +5     
=======================================
+ Hits        19120    19139   +19     
- Misses       4144     4147    +3     
  Partials       27       27           

Files with missing lines	Coverage Δ
...Components/CreateImageWizard/CreateImageWizard.tsx	97.24% <ø> (-0.01%)	⬇️
...mponents/CreateImageWizard/steps/Details/index.tsx	100.00% <100.00%> (ø)
...teImageWizard/steps/FileSystem/FileSystemTable.tsx	71.24% <100.00%> (+0.07%)	⬆️
...ents/CreateImageWizard/utilities/useValidation.tsx	84.70% <100.00%> (ø)
...rc/Components/CreateImageWizard/ValidatedInput.tsx	91.56% <93.75%> (+0.06%)	⬆️
src/Components/CreateImageWizard/validators.ts	92.56% <83.33%> (-1.25%)	⬇️

---

Continue to review full report in Codecov by Sentry.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 06f4ce4...c18a152. Read the comment docs.

[lucasgarfield]

Nice catch, the validation should not appear when there is no hostname.

I’m not sure about the approach, though. In this PR we’re hiding the validation by returning an error object with a fieldname of ‘default’. That technically works and is supported by the HookValidatedInput but it just feels kind of surprising/counter-intuitive.

I think it feels counter-intuitive because <HookValidatedInput> does not surface validation on pristine state components. So there is already an existing pattern for not displaying validation – that’s what I intuitively would expect to use. In other words, set isPristine to true when the field’s value returns to the default state (in this case, ‘’).

This might be a good opportunity to refactor a little bit. The double nested ternary operator is hard to understand. It definitely does support this ‘feature’ where if the fieldName is ‘default’ then but let’s take a step back and ask… should it? Maybe we can get rid of the nested ternary and the concept of a ‘default’ state. Having to manage both ‘default’ and ‘isPristine’ feels duplicitous.

In all the validation hooks, it only seems to be used in one other place, and there is a comment specifically calling it out as a hacky solution…

Therefore, I am a little hesitant to introduce this new pattern to the code base. I’d prefer refactoring a bit. What do you think?

[mgold1234]

after some investigation, the solution that I did here https://github.com/osbuild/image-builder-frontend/pull/2844/files#diff-5206f54a601e8c56290e1a412fbf3f35172f16b892c7674852c91b4ea846bcd9R62
will fix that problem, so I can close this pr

[mgold1234]

I saw your comment here #2844
so reopen this pr and refactor the code now

[regexowl]

One small bug in FSC - the input doesn't get validated without a value

this is how the error for invalid value looks like:

There's also the same problem with Blueprint name which should be required:

[mgold1234]

nice catch @regexowl , I am looking at stage, if user delete the value in Blueprint name we should see an error be cecause it's mandatory field, am I wrong?
I fixed the FSC respectivally