test_cache_name_formats.py

#!/usr/bin/env python3
"""Smoke tests for cache filename/slug generation.

This validates that the cache manager produces the desired filename pattern
for different processor/model combinations (CPU model-1, CPU model-2, GPU models,
MarkPDFdown, Sparrow).
"""

import os
from contextlib import contextmanager

from src.utils import CacheManager


@contextmanager
def _temp_env(**entries):
    """Temporarily patch environment variables."""

    original = {key: os.environ.get(key) for key in entries}
    try:
        os.environ.update({k: v for k, v in entries.items() if v is not None})
        yield
    finally:
        for key, value in original.items():
            if value is None and key in os.environ:
                os.environ.pop(key, None)
            elif value is not None:
                os.environ[key] = value


def expect(cache_key: str, expected_prefix: str):
    if not cache_key.startswith(expected_prefix):
        raise AssertionError(
            f"Cache key '{cache_key}' does not start with expected prefix '{expected_prefix}'"
        )


def main() -> int:
    cm = CacheManager()
    filename = "Example File.pdf"

    # CPU Model-1 default
    with _temp_env(CPU_MODEL="cpu-model-1"):
        key = cm._get_cache_key(filename, "CPU")
        print("CPU Model-1 key:", key)
        expect(key, "cpu_model-1_")

    # CPU Model-2
    with _temp_env(CPU_MODEL="cpu-model-2"):
        key = cm._get_cache_key(filename, "CPU")
        print("CPU Model-2 key:", key)
        expect(key, "cpu_model-2_")

    # GPU model from environment (use first configured model)
    from src.config import app_config  # Imported lazily to avoid circular issues

    gpu_models = app_config.get_available_gpu_models()
    first_display, first_model = next(iter(gpu_models.items()))

    with _temp_env(OPENROUTER_MODEL=first_model):
        key = cm._get_cache_key(filename, "GPU")
        print(f"GPU ({first_display}) key:", key)
        expected_prefix = f"gpu_{first_display.lower().replace(' ', '_').replace('(', '').replace(')', '').replace('-', '_')}"
        expect(key, f"{expected_prefix}_")

    # MarkPDFdown
    key = cm._get_cache_key(filename, "PDF to Markdown Converter")
    print("MarkPDFdown key:", key)
    expect(key, "markpdfdown_")

    # Sparrow
    key = cm._get_cache_key(filename, "Schema Support")
    print("Sparrow key:", key)
    expect(key, "sparrow_")

    print("All cache key naming checks passed ✅")
    return 0


if __name__ == "__main__":
    raise SystemExit(main())