feat(expression): cast numeric literals to decimal type

[huan233usc]

What

Add casting of numeric literals (int, long, float, double) to a decimal target type in Literal::CastTo, so a numeric value can be used as a default for a decimal column.

Previously CastFromInt / CastFromLong / CastFromFloat / CastFromDouble had no kDecimal case and fell through to NotSupported, so a default like Literal::Int(12) or Literal::Double(9.99) for a decimal(9, 2) column was rejected. Java allows these (IntegerLiteral.to, DoubleLiteral.to, etc. scale the value to the target scale), so this brings the C++ literal cast layer to parity for numeric sources.

How

• Integer → decimal: CastIntegerToDecimal scales the integer (scale 0) up to the target scale via Decimal::Rescale(0, scale), then verifies the result fits the target precision (FitsInPrecision). Example: 12 → decimal(9,2) yields unscaled 1200 (12.00).
• Float/double → decimal: CastRealToDecimal parses the value's shortest round-tripping decimal representation (matching Java's BigDecimal.valueOf(double) via Double.toString), then rounds to the target scale with HALF_UP rounding (round half away from zero, as Java does — 2.5 → 3, -2.5 → -3), and checks precision. Non-finite values are rejected.
• Both paths reject an out-of-range decimal scale before indexing the powers-of-ten table (DecimalType does not bound its scale on construction, so decimal(9, 40) would otherwise read past the table).

Scope

Follow-up split out from the v3 default-value work (see the CastDefaultToType discussion on #793). It only extends the shared Literal::CastTo layer for numeric sources.

Testing

LiteralTest.IntegerCastToDecimal (int/long scaling, out-of-precision rejection, out-of-range scale rejection) and LiteralTest.RealCastToDecimal (float/double scaling, HALF_UP rounding incl. negative, round-down, out-of-precision and non-finite rejection), all verified fail-without / pass-with. Full expression_test passes (495 tests).

[wgtmac]

Please guard drop before calling Rescale. For values like 1e-100 cast to decimal(9,2), drop is 98 and indexes past the powers-of-ten table; Java rounds this to 0.00.

[huan233usc]

Fixed. Rounding now goes through a shared RescaleHalfUp helper that returns 0 when the drop exceeds the max scale (so 1e-100 -> 0.00 instead of indexing past the table). Added a regression test.

[wgtmac]

Please allow negative decimal scales here and in the real-value path. Iceberg decimal(P,S) supports negative S, and Java uses setScale(scale, HALF_UP), so 149 -> decimal(9,-2) should round to 1E+2 instead of failing.

[huan233usc]

Done — both the integer and real paths now allow negative scales and round HALF_UP, so 149 -> decimal(9,-2) gives 1E+2. Tested for both int and double sources.