import json
from pathlib import Path
import pytest
from chunkhound.autodoc.generator import generate_docsite
from chunkhound.autodoc.models import CleanupConfig
from chunkhound.interfaces.llm_provider import LLMProvider, LLMResponse
class _CapturingProvider(LLMProvider):
def __init__(self) -> None:
self._model = "fake"
self.last_batch_system: str | None = None
self.last_structured_prompt: str | None = None
self.last_complete_prompt: str | None = None
@property
def name(self) -> str:
return "fake"
@property
def model(self) -> str:
return self._model
async def complete(
self,
prompt: str,
system: str | None = None,
max_completion_tokens: int = 4096,
) -> LLMResponse:
self.last_complete_prompt = prompt
return LLMResponse(
content="## Overview\nEnd-user overview.\n\n- Use case 1\n- Use case 2",
tokens_used=0,
model=self._model,
finish_reason="stop",
)
async def complete_structured(
self,
prompt: str,
json_schema: dict[str, object],
system: str | None = None,
max_completion_tokens: int = 4096,
) -> dict[str, object]:
self.last_structured_prompt = prompt
return {
"nav": {"groups": [{"title": "Group", "slugs": ["01-topic-one"]}]},
"glossary": [
{"term": "Term", "definition": "Definition.", "pages": ["01-topic-one"]}
],
}
async def batch_complete(
self,
prompts: list[str],
system: str | None = None,
max_completion_tokens: int = 4096,
) -> list[LLMResponse]:
self.last_batch_system = system
return [
LLMResponse(
content="## Overview\nCleaned.\n\n## Details\nMore.",
tokens_used=0,
model=self._model,
finish_reason="stop",
)
for _ in prompts
]
def estimate_tokens(self, text: str) -> int:
return 0
async def health_check(self) -> dict[str, object]:
return {"ok": True}
def get_usage_stats(self) -> dict[str, object]:
return {}
class _FakeLLMManager:
def __init__(self, provider: LLMProvider) -> None:
self._provider = provider
def get_synthesis_provider(self) -> LLMProvider:
return self._provider
def _write_minimal_input_dir(input_dir: Path) -> None:
(input_dir / "scope_code_mapper_index.md").write_text(
"\n".join(
[
"# AutoDoc Topics (/repo)",
"",
"1. [Topic One](topic_one.md)",
]
),
encoding="utf-8",
)
(input_dir / "topic_one.md").write_text(
"\n".join(
[
"# Topic One",
"",
"Overview body.",
"",
"## Sources",
"",
"└── repo/",
"\t└── [1] x.py (1 chunks: L1-2)",
]
),
encoding="utf-8",
)
@pytest.mark.asyncio
async def test_audience_influences_llm_cleanup_and_global_ia_prompts(
tmp_path: Path,
) -> None:
input_dir = tmp_path / "input"
input_dir.mkdir(parents=True)
_write_minimal_input_dir(input_dir)
output_dir = tmp_path / "out"
provider = _CapturingProvider()
llm_manager = _FakeLLMManager(provider)
await generate_docsite(
input_dir=input_dir,
output_dir=output_dir,
llm_manager=llm_manager, # type: ignore[arg-type]
cleanup_config=CleanupConfig(
mode="llm",
batch_size=1,
max_completion_tokens=512,
audience="technical",
),
site_title=None,
site_tagline=None,
)
assert provider.last_batch_system is not None
assert "Audience: technical" in provider.last_batch_system
assert provider.last_structured_prompt is not None
assert "Audience: technical" in provider.last_structured_prompt
assert provider.last_complete_prompt is not None
assert "Audience: technical" in provider.last_complete_prompt
site_json = json.loads(
(output_dir / "src" / "data" / "site.json").read_text(encoding="utf-8")
)
assert (
site_json["tagline"]
== "Engineering-focused documentation generated from AutoDoc output."
)
@pytest.mark.asyncio
async def test_end_user_audience_synthesizes_homepage_overview(
tmp_path: Path,
) -> None:
input_dir = tmp_path / "input"
input_dir.mkdir(parents=True)
_write_minimal_input_dir(input_dir)
output_dir = tmp_path / "out"
provider = _CapturingProvider()
llm_manager = _FakeLLMManager(provider)
await generate_docsite(
input_dir=input_dir,
output_dir=output_dir,
llm_manager=llm_manager, # type: ignore[arg-type]
cleanup_config=CleanupConfig(
mode="llm",
batch_size=1,
max_completion_tokens=512,
audience="end-user",
),
site_title=None,
site_tagline=None,
)
assert provider.last_complete_prompt is not None
assert "Audience: end-user" in provider.last_complete_prompt
index_md = (output_dir / "src" / "pages" / "index.md").read_text(encoding="utf-8")
assert "## Overview" in index_md
assert "End-user overview." in index_md
