gendesign/backend/tests/services/chat/test_tools.py
Light1YT fceaaf9a2c
All checks were successful
CI / changes (push) Successful in 6s
CI / frontend-tests (push) Has been skipped
CI / changes (pull_request) Successful in 5s
CI / frontend-tests (pull_request) Has been skipped
Deploy / changes (push) Successful in 5s
Deploy / build-frontend (push) Has been skipped
Deploy / build-backend (push) Successful in 2m41s
CI / backend-tests (push) Successful in 6m24s
CI / backend-tests (pull_request) Successful in 6m23s
Deploy / build-worker (push) Successful in 3m26s
Deploy / deploy (push) Successful in 1m12s
feat(chat): LLM tool-loop + §19 redaction wiring for #957 (Step 2+3)
Add the LLM prose-composition path for the parcel-forecast chat, layered
over the deterministic Step-1 fallback which stays the safety net.

- chat/tools.py: 5 read-only section tools (exec_summary, product_recommendation,
  forecast, risks, scenarios) — pure slices of the loaded report dict, no DB/
  recompute, graceful on missing sections. market_now (raw analyze blob) and meta
  are deliberately NOT exposed -> highest-PII data cannot reach the LLM.
- chat/safe_payload.py: the §19 gate — single place that builds the outbound
  SafePayload from a section-aggregate allowlist; honors is_confidential hard-block.
- chat/orchestrator.py: manual tool-call loop with call-cap/termination, real
  grounded_in provenance; any LLMResult.ok=False (disabled/timeout/rate_limited/
  redaction_refused/call_cap/provider_error/empty) degrades to the deterministic answer.
- llm/prompts.py: versioned chat_system@v1 — answer only from sections, never
  fabricate numbers, advisory tone, decline out-of-scope.
- api/v1/chat.py: branch on settings.llm_enabled; sync complete bridged via
  run_in_threadpool. Default-off -> deterministic path, no provider built.
- Tests: fake provider only (no network), planted-secret redaction-boundary +
  per-reason fallback + call-cap + numbers-from-report coverage.

Refs #957
2026-06-08 17:45:01 +05:00

151 lines
5.9 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

"""Тесты секционных tool'ов чата (#957, Step 2) — PURE срезы report_dict, без БД/recompute.
Покрываем:
• tool_specs: 5 спек, валидная OpenAI-форма, без параметров.
• executors: достают нужную секцию ВЕРБАТИМ; числа не пересчитываются.
• robust: отсутствует секция → маркер «недоступно», НЕ KeyError.
• execute_tool/sections_for_tool: реестр имя→executor + provenance-секции.
• неизвестное имя tool'а: безопасный маркер + пустые секции (не падаем).
"""
from __future__ import annotations
from typing import Any
from app.services.chat import tools
_EXPECTED_TOOLS = {
"get_exec_summary",
"get_product_recommendation",
"get_forecast",
"get_risks",
"get_scenarios",
}
def _report() -> dict[str, Any]:
"""Узнаваемый отчёт (форма SiteFinderReport.as_dict())."""
return {
"exec_summary": {
"headline": "Комфорт-класс",
"key_numbers": {"цена_руб_м2": 250000},
"overall_confidence": "medium",
},
"future_market": {
"forecasts_by_horizon": [{"horizon": 12}, {"horizon": 24}],
"future_supply": {"pressure": 0.4},
},
"product_tz": {"obj_class": "комфорт", "mix": [{"fmt": "2k"}]},
"scoring": {"special_indices": {"indices": {"cannibalization": {"value": 0.31}}}},
"confidence": {"level": "medium", "factors": {"deals": "ok"}},
"scenarios": {"by_scenario": {"base": {}, "conservative": {}}},
}
# ── Спеки ───────────────────────────────────────────────────────────────────────
def test_tool_specs_cover_all_five() -> None:
specs = tools.tool_specs()
assert len(specs) == 5
assert {s["function"]["name"] for s in specs} == _EXPECTED_TOOLS
def test_tool_specs_are_valid_openai_function_shape() -> None:
for spec in tools.tool_specs():
assert spec["type"] == "function"
fn = spec["function"]
assert isinstance(fn["name"], str) and fn["name"]
assert isinstance(fn["description"], str) and fn["description"]
params = fn["parameters"]
assert params["type"] == "object"
# Секционные tool'ы без аргументов (модель не управляет вычислениями).
assert params["properties"] == {}
assert params["additionalProperties"] is False
# ── Executors: верные срезы, числа вербатим ──────────────────────────────────────
def test_exec_summary_slice_is_verbatim() -> None:
out = tools.get_exec_summary(_report())
assert out["headline"] == "Комфорт-класс"
# Число берётся как есть, не пересчитывается.
assert out["key_numbers"]["цена_руб_м2"] == 250000
def test_product_recommendation_slice() -> None:
out = tools.get_product_recommendation(_report())
assert out["obj_class"] == "комфорт"
assert out["mix"] == [{"fmt": "2k"}]
def test_forecast_slice() -> None:
out = tools.get_forecast(_report())
assert len(out["forecasts_by_horizon"]) == 2
assert out["future_supply"]["pressure"] == 0.4
def test_risks_merges_scoring_and_confidence() -> None:
out = tools.get_risks(_report())
assert set(out) == {"scoring", "confidence"}
indices = out["scoring"]["special_indices"]["indices"]
assert indices["cannibalization"]["value"] == 0.31
assert out["confidence"]["level"] == "medium"
def test_scenarios_slice() -> None:
out = tools.get_scenarios(_report())
assert set(out["by_scenario"]) == {"base", "conservative"}
# ── Robust: отсутствующая/пустая секция → маркер, НЕ KeyError ─────────────────────
def test_missing_section_returns_not_available_marker() -> None:
assert tools.get_exec_summary({}) == {"available": False}
assert tools.get_forecast({}) == {"available": False}
assert tools.get_scenarios({}) == {"available": False}
def test_risks_robust_to_missing_subsections() -> None:
out = tools.get_risks({})
assert out == {
"scoring": {"available": False},
"confidence": {"available": False},
}
def test_empty_section_dict_treated_as_unavailable() -> None:
# Пустой dict секции → маркер (нечего показывать), а не «{}» как валидные данные.
assert tools.get_product_recommendation({"product_tz": {}}) == {"available": False}
# ── Реестр: execute_tool / sections_for_tool / is_known_tool ─────────────────────
def test_execute_tool_dispatches_by_name() -> None:
out = tools.execute_tool("get_exec_summary", _report())
assert out["headline"] == "Комфорт-класс"
def test_execute_unknown_tool_returns_marker_not_raise() -> None:
assert tools.execute_tool("get_nonexistent", _report()) == {"available": False}
def test_sections_for_tool_provenance() -> None:
assert tools.sections_for_tool("get_exec_summary") == ("exec_summary",)
assert tools.sections_for_tool("get_risks") == ("scoring", "confidence")
assert tools.sections_for_tool("get_nonexistent") == ()
def test_is_known_tool() -> None:
assert tools.is_known_tool("get_forecast") is True
assert tools.is_known_tool("nope") is False
def test_executors_do_not_mutate_report() -> None:
report = _report()
before = dict(report["exec_summary"])
tools.get_exec_summary(report)
assert report["exec_summary"] == before