MTLGenerationStrategy

protocol MTLGenerationStrategy : Sendable

Overview

Generation strategies abstract the destination and lifecycle of generated text, enabling MTL templates to generate output to files, memory, or other targets without changing the template execution logic.

Overview

MTL supports different generation strategies:

• File system: Write directly to files on disk

• In-memory: Accumulate text in memory for testing or programmatic access

• Custom: Implement custom strategies for databases, network, etc.

Strategy Lifecycle

Each file block in an MTL template follows this lifecycle:

1. Create writer: createWriter(url:mode:charset:indentation:) is called

2. Template writes: Template statements write to the returned writer

3. Finalize: finalizeWriter(_:) is called to commit/save the output

Example Usage

// File system strategy
let fileStrategy = MTLFileSystemStrategy(basePath: "/output")
let writer1 = try await fileStrategy.createWriter(
    url: "models/Person.swift",
    mode: .overwrite,
    charset: "UTF-8",
    indentation: MTLIndentation()
)
await writer1.writeLine("class Person {}")
try await fileStrategy.finalizeWriter(writer1)
// File written to /output/models/Person.swift

// In-memory strategy (for testing)
let memoryStrategy = MTLInMemoryStrategy()
let writer2 = try await memoryStrategy.createWriter(
    url: "test.txt",
    mode: .create,
    charset: "UTF-8",
    indentation: MTLIndentation()
)
await writer2.writeLine("Test output")
try await memoryStrategy.finalizeWriter(writer2)
let files = await memoryStrategy.getGeneratedFiles()
// files["test.txt"] == "Test output\n"