perf: optimize spark_cast_float64_to_utf8 (~40% faster)

[andygrove]

Which issue does this PR close?

N/A

Rationale for this change

Optimize existing expression.

What changes are included in this PR?

Format float-to-string casts straight into a GenericStringBuilder with a reused scratch buffer instead of allocating a String per row (and a second one in the scientific-notation path), removing one malloc/free and one copy per value.

How are these changes tested?

Existing tests.

Benchmark (criterion):

• cast_f32_to_utf8: 42.453% faster (base 408097ns -> cand 234847ns)
• cast_f64_to_utf8: 41.57% faster (base 418096ns -> cand 244294ns)

Full criterion output:

cast_float_to_string/cast_f64_to_utf8
                        time:   [243.85 µs 244.03 µs 244.23 µs]
                        change: [−41.767% −41.570% −41.425%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 4 outliers among 100 measurements (4.00%)
  2 (2.00%) high mild
  2 (2.00%) high severe
cast_float_to_string/cast_f32_to_utf8
                        time:   [234.71 µs 234.86 µs 235.08 µs]
                        change: [−42.515% −42.453% −42.390%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 6 outliers among 100 measurements (6.00%)
  2 (2.00%) low mild
  1 (1.00%) high mild
  3 (3.00%) high severe