AgentSkillsCN

refactor:functional

声明式函数式编程。

SKILL.md
--- frontmatter
name: refactor:functional
description: Declarative Functional Programming

Refactor this Swift codebase to follow functional programming principles:

Replace Mutability with Immutability

Convert var to let wherever possible Replace mutable collections with immutable transformations Use value types (structs/enums) instead of reference types (classes) where appropriate Eliminate property mutations in favor of creating new instances

Extract Pure Functions

Identify functions with side effects and refactor them to be pure Move I/O operations (network calls, file access, database queries) to the edges of the system Replace stateful methods with functions that take input and return output Ensure functions don't modify external state or rely on mutable global variables

Use Higher-Order Functions

Replace for-loops with map, filter, reduce, compactMap, flatMap Convert imperative conditional logic to functional chains Use function composition to build complex operations from simple ones Leverage lazy sequences for performance where appropriate

Functional Error Handling

Replace try-catch blocks with Result types where suitable Use Optional chaining and nil-coalescing operators Implement functional error propagation with map and flatMap on Result Avoid force unwrapping (!); use guard let, if let, or functional alternatives

Eliminate Shared Mutable State

Replace singletons with dependency injection of immutable services Convert stateful classes to stateless functions or immutable structs Use actors or other concurrency-safe patterns when mutation is necessary Pass data explicitly rather than relying on shared state

Apply Functional Patterns

Implement custom operators for common transformations Use currying for partial function application Create small, composable functions following the single responsibility principle Model domain concepts as algebraic data types (enums with associated values)

Specific Transformations

Convert delegate patterns to closure-based APIs Replace inheritance hierarchies with protocol composition Transform imperative algorithms to recursive or fold-based implementations Use KeyPaths for type-safe property access

Example transformation patterns to look for:

Loops that build arrays → map/filter/reduce Nested if-else → pattern matching with switch Mutable accumulator patterns → fold/reduce Callback hell → flatMap chains or async/await Stateful iteration → sequence/generator patterns

Maintain Swift idioms while applying functional principles. The goal is readable, maintainable code that leverages Swift's functional capabilities without sacrificing clarity.