fix(snowflake): convert Decimal columns to float64 instead of object dtype

[fadil4u]

Issue

When querying a Snowflake table with NUMBER(p, s) columns where scale > 0 (e.g. NUMBER(38, 8)), the resulting pandas DataFrame has object dtype instead of float64: see #11990

t = con.table("MY_TABLE")  # has NUMBER(38, 8) column
df = t.execute()
print(df["my_col"].dtype)  # object, expected float64

Snowflake returns NUMBER(38, 8) as double in Arrow format, so the data is already float64 after _fetch_from_cursor. However __pandas_result__ was defaulting to plain PandasData as the data_mapper, which runs normalize_decimal on every value — wrapping each float64 in a Python decimal.Decimal object and producing object dtype.

Fix

1. Override execute() in the Snowflake backend to pass data_mapper=SnowflakePandasData to __pandas_result__
2. Add convert_Decimal to SnowflakePandasData that casts to float64 instead of running normalize_decimal

Issues closed

• Resolves bug: Snowflake: NUMBER columns return as object dtype via Arrow (Decimal128 not cast to float64) #11990

[NickCrews]

see the linked issue, I'm not yet convinced this is a bug