Struct Inference

The simplest demonstration of Turn's Cognitive Type Safety. Define a struct, then use infer to have the LLM populate it from a natural language prompt. Turn compiles the struct schema at build time, sends it to the LLM as a JSON Schema constraint, and validates the response before binding it to the typed variable.

If the LLM returns malformed JSON or the wrong field types, Turn's coercion engine retries automatically — the caller always receives a correctly typed value or an explicit Result::Err.

struct_infer.tn

// Turn Language: Struct Inference
// Demonstrates Cognitive Type Safety: inferring a structured type.

// 1. Define the shape of thought
struct Sentiment {
  score: Num,
  reasoning: Str
};

call("echo", "Analyzing sentiment...");

// 2. Infer — the LLM populates the struct; Turn validates the output
let result = infer Sentiment {
  "I love this new language! It makes building agents so much easier.";
};

// 3. Type-safe access — no casting, no null checks
call("echo", "Score: " + result.score);
call("echo", "Reasoning: " + result.reasoning);

if result.score > 0.8 {
  call("echo", "Verdict: POSITIVE");
} else {
  call("echo", "Verdict: NEGATIVE/NEUTRAL");
}

return result;

Run it:

export TURN_INFER_PROVIDER=~/.turn/providers/turn_provider_openai.wasm
turn run impl/examples/struct_infer.tn

The full implementation is in impl/examples/struct_infer.tn.