refactor: Refactor numeric sign and padding in Spark format_string

[JSOD11]

Which issue does this PR close?

• Closes Refactor numeric sign and padding in Spark format_string #24058

Rationale for this change

format_decimal and format_float had duplicated logic, and it'd be easy to modify one function and forget to change the other, causing bugs. This PR forces one source of truth for that logic.

What changes are included in this PR?

Created a new helper.

Are these changes tested?

Yes,

cargo t -p datafusion-spark format_string --lib

passes.

Are there any user-facing changes?

No user facing-changes.

[JSOD11]

cc @kosiew let me know your thoughts

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 81.02%. Comparing base (3e3a92d) to head (75e31d0).
⚠️ Report is 3 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #24115      +/-   ##
==========================================
+ Coverage   80.91%   81.02%   +0.10%     
==========================================
  Files        1103     1105       +2     
  Lines      377219   379692    +2473     
  Branches   377219   379692    +2473     
==========================================
+ Hits       305244   307654    +2410     
- Misses      53775    53814      +39     
- Partials    18200    18224      +24     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[kosiew]

@JSOD11,

Thanks for the cleanup here. Pulling the shared prefix, number, suffix assembly into a single helper makes the formatting logic much easier to follow and reduces the chance of these paths drifting apart again. I also like that the finite-only zero padding behavior for floats is preserved explicitly through the helper.

I have one small suggestion around regression coverage, but nothing that should block this PR. Nice work!

[kosiew]

Nice extraction. One thing I'd consider adding is a couple of focused regression tests around the helper's policy boundaries.

In particular, it would be good to verify that %0...f formats NaN and Infinity using space padding rather than zero padding, and that a negative decimal formatted with (, 0, and a width still keeps the closing ) after the digits.

The existing tests cover finite zero padding and parenthesized decimals independently, but these combinations exercise the behavior that write_numeric_parts is now responsible for.

[JSOD11]

Thanks for the review!

Good idea, just added these tests. Let me know if the change looks like what you had in mind.