[decimal] Implement scaleb, bit_count, __float__, fma, etc

[forfudan]

Implements several Python decimal.Decimal / int-compatibility APIs across Decimo’s numeric types, with accompanying CLI/test coverage and roadmap updates.

Changes:

• Add BigInt.__float__() and BigInt.bit_count() APIs, plus unit tests.
• Add BigDecimal.scaleb(), BigDecimal.fma(), and BigDecimal.to_string_with_separators(), plus unit tests.
• Extend CLI evaluator tests to cover rounding modes; clarify CLI/evaluator precision documentation and update the API roadmap.

[Copilot]

Pull request overview

Implements several Python decimal.Decimal / int-compatibility APIs across Decimo’s numeric types, with accompanying CLI/test coverage and roadmap updates.

Changes:

• Add BigInt.__float__() and BigInt.bit_count() APIs, plus unit tests.
• Add BigDecimal.scaleb(), BigDecimal.fma(), and BigDecimal.to_string_with_separators(), plus unit tests.
• Extend CLI evaluator tests to cover rounding modes; clarify CLI/evaluator precision documentation and update the API roadmap.

Reviewed changes

Copilot reviewed 9 out of 9 changed files in this pull request and generated 2 comments.

Show a summary per file

File	Description
tests/cli/test_evaluator.mojo	Adds evaluator tests verifying rounding-mode behavior at given significant-digit precision.
tests/bigint/test_bigint_conversion.mojo	Adds __float__ conversion tests for BigInt.
tests/bigint/test_bigint_bitwise.mojo	Adds bit_count() test coverage, including negative and large values.
tests/bigdecimal/test_bigdecimal_methods.mojo	Adds tests for scaleb(), fma(), and to_string_with_separators().
src/decimo/bigint/bigint.mojo	Implements __float__() and bit_count() on BigInt (and adds FloatableRaising trait conformance).
src/decimo/bigdecimal/bigdecimal.mojo	Implements to_string_with_separators(), fma(), and scaleb() on BigDecimal.
src/cli/main.mojo	Clarifies precision option comment as “significant digits”.
src/cli/calculator/evaluator.mojo	Improves guard-digits documentation for working precision.
docs/plans/api_roadmap.md	Marks newly implemented APIs as completed with brief implementation notes.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

__float__ converts via Float64(String(self)), which forces a full base-2^32 → decimal string conversion and then reparses it. For very large BigInt values this is substantially more expensive (time + memory) than converting from the binary words directly; consider a word-based Float64 conversion to avoid the intermediate string if this becomes a hotspot.

Suggested change

	return Float64(String(self))
	var result = 0.0
	# Process words from most significant to least significant.
	# Each word contributes 32 bits in base-2^32 representation.
	for i in range(len(self.words) - 1, -1, -1):
	result = result * 4294967296.0 + Float64(self.words[i])
	# Apply sign (assuming negative values have self.sign < 0).
	if self.sign < 0:
	result = -result
	return result

[Copilot]

bit_count() iterates over all self.words. Other methods (e.g., to_string) compute an effective word count that trims most-significant zero words; if BigInt instances can contain such trailing zeros, bit_count() will do unnecessary work. Consider trimming/using an eff_words count before counting bits.