[decimal] Implement tan(), cot(), csc(), sec() for BigDecimal

[forfudan]

This pull request enhances the BigDecimal module by adding new trigonometric functions, improving precision handling, and updating test cases to ensure correctness. The most important changes include the addition of tangent, cotangent, cosecant, and secant functions, updates to the sin and cos functions for zero handling, and the inclusion of comprehensive test cases for the new functionality.

Enhancements to BigDecimal functionality:

• Added new trigonometric functions: tan, cot, csc, and sec to the BigDecimal struct, enabling calculations for tangent, cotangent, cosecant, and secant with arbitrary precision. [1] [2]
• Improved handling of zero values in sin and cos functions to use BigDecimal(BigUInt.ZERO) and BigDecimal(BigUInt.ONE) for better precision and consistency. [1] [2] [3]

Updates to test cases:

• Added test cases for new trigonometric functions (tan, cot, csc, and sec) to the TOML test data file, covering edge cases and typical values.
• Updated the run_test function to directly compare BigDecimal objects instead of converting them to strings for improved accuracy in assertions.
• Expanded the test_bigdecimal_trigonometric function to include tests for cos, tan, and cot functions alongside existing tests for sin and arctan.

Codebase improvements:

• Refactored import statements in trigonometric.mojo to include decimojo.bigdecimal.constants, ensuring access to constants like π for trigonometric calculations.