"""Every third-party import must be declared in requirements.txt or pyproject.toml.

`requests` — the HTTP client behind EVERY COSMOS call, imported at module level in
`cosmos/client.py` — was in neither file. A clean `pip install -r requirements.txt` then
failed on import, not at the first request. `openpyxl`, which reads the competitor workbook on
the default path, was missing too. Meanwhile `openai` was declared and never imported: the
real import is `langchain-openai`, so declaring `openai` alone did not enable the LLM path.

These tests parse the actual imports instead of trusting a hand-maintained list, so adding a
dependency without declaring it fails here rather than on someone else's machine.
"""
from __future__ import annotations

import ast
import pathlib
import re
import sys

ROOT = pathlib.Path(__file__).resolve().parents[1]
SKIP_DIRS = {"__pycache__", ".venv", "new_archtetcure_and frontend", ".mojibake_backup"}
LOCAL = {"pricing_agent", "dashboard", "config", "app", "legacy_app", "backend", "tests"}

# import name -> distribution name, where they differ.
DIST = {
    "yaml": "pyyaml",
    "dotenv": "python-dotenv",
    "apify_client": "apify-client",
    "langchain_openai": "langchain-openai",
    "pydantic_settings": "pydantic-settings",
    "google": "google-auth",
    "sp_api": "python-amazon-sp-api",
    "PIL": "pillow",
}
# Test-only, and declared under [dev].
TEST_ONLY = {"pytest"}


def _sources():
    for f in ROOT.rglob("*.py"):
        if not SKIP_DIRS.intersection(f.parts):
            yield f


def _third_party_imports() -> dict[str, set[str]]:
    found: dict[str, set[str]] = {}
    for f in _sources():
        try:
            tree = ast.parse(f.read_text(encoding="utf-8"))
        except SyntaxError:                       # not our problem here
            continue
        for node in ast.walk(tree):
            mods: list[str] = []
            if isinstance(node, ast.Import):
                mods = [a.name.split(".")[0] for a in node.names]
            elif isinstance(node, ast.ImportFrom) and node.module and node.level == 0:
                mods = [node.module.split(".")[0]]
            for m in mods:
                if m and m not in sys.stdlib_module_names and m not in LOCAL:
                    found.setdefault(m, set()).add(str(f.relative_to(ROOT)))
    return found


def _declared() -> set[str]:
    text = (ROOT / "requirements.txt").read_text(encoding="utf-8")
    text += (ROOT / "pyproject.toml").read_text(encoding="utf-8")
    names = re.findall(r"^\s*[\"']?([A-Za-z][A-Za-z0-9._-]*)\s*(?:[><=!~]|[\"']|$)",
                       text, re.M)
    return {n.lower().replace("_", "-") for n in names}


def test_every_import_is_declared_somewhere():
    declared = _declared()
    missing = {}
    for mod, files in _third_party_imports().items():
        if mod in TEST_ONLY:
            continue
        dist = DIST.get(mod, mod).lower().replace("_", "-")
        if dist not in declared and mod.lower().replace("_", "-") not in declared:
            missing[dist] = sorted(files)[:3]
    assert not missing, f"undeclared dependencies: {missing}"


def test_requests_is_in_requirements_txt():
    """Module-level in cosmos/client.py — the dashboard cannot import without it."""
    req = (ROOT / "requirements.txt").read_text(encoding="utf-8")
    assert re.search(r"^requests\b", req, re.M), "requests missing from requirements.txt"


def test_openpyxl_is_in_requirements_txt():
    """The default competitor path reads a workbook; without this it raises at runtime."""
    req = (ROOT / "requirements.txt").read_text(encoding="utf-8")
    assert re.search(r"^openpyxl\b", req, re.M), "openpyxl missing from requirements.txt"


def test_the_llm_extra_names_the_package_actually_imported():
    """`openai` alone does not enable the narrative — the import is `langchain_openai`."""
    req = (ROOT / "requirements.txt").read_text(encoding="utf-8")
    assert re.search(r"^langchain-openai\b", req, re.M)


def test_dashboard_requirements_stay_a_subset_of_the_package():
    """Anything the dashboard needs must also be installable via `pip install -e .`."""
    req_names = {
        m.group(1).lower()
        for m in re.finditer(r"^([A-Za-z][A-Za-z0-9._-]*)",
                             (ROOT / "requirements.txt").read_text(encoding="utf-8"), re.M)
    }
    proj = (ROOT / "pyproject.toml").read_text(encoding="utf-8").lower()
    # streamlit/plotly/pandas/numpy live under the [ui] extra; everything else is a core dep.
    for name in req_names:
        assert name in proj, f"{name} is in requirements.txt but not in pyproject.toml"