tidylinq

A set of utilities I've found useful when developing LLM tooling.

Usage:

from pydantic import BaseModel, Field
from textwrap import dedent
from tidylinq import from_iterable, completion_with_schema, retry

class Translation(BaseModel):
    term_source: str = Field(description="The source language term")
    term_target: str = Field(description="The term translated to the target language.")
    example_source: str = Field(description="Example of using the term in the source language.")
    example_target: str = Field(description="Example sentence target to the target language.")

messages = [
    {
        "role": "system",
        "content": dedent("""
            You are an expert translator. Output a structured Japanese response in JSON format for the provided English words.
        """).strip(),
    },
]

translations = (
    from_iterable(["cat", "dog", "hyena"])
    .select(
        lambda w: retry(completion_with_schema, backoff=1.0)(
            model="gemini/gemini-2.5-flash",
            messages=messages + [{"role": "user", "content": w}],
            response_schema=Translation,
        ),
        parallelism=4,
    )
    .with_progress("Translating words")
)

results = translations.to_list()
for translation in results:
    print(f"English: {translation.term_source}")
    print(f"Japanese: {translation.term_target}")
    print(f"Example (EN): {translation.example_source}")
    print(f"Example (JP): {translation.example_target}")