ObsiGate/tests/test_xlsx_viewer.py

"""Display / edit / download for .xlsx files (viewer + PUT xlsx/save)."""

from __future__ import annotations

from pathlib import Path

import pytest

openpyxl = pytest.importorskip("openpyxl")

VAULT = "TestVault"


@pytest.fixture
def xlsx_file(test_vault_dir: str) -> str:
    from openpyxl import Workbook

    path = Path(test_vault_dir) / "budget.xlsx"
    wb = Workbook()
    ws = wb.active
    ws.title = "Budget"
    ws["A1"] = "Poste"
    ws["B1"] = 100
    ws["A2"] = "Total"
    ws["B2"] = "=B1*2"
    notes = wb.create_sheet("Notes")
    notes["A1"] = "hello"
    wb.save(path)
    return str(path)


# ── Display ───────────────────────────────────────────────────────────────


class TestXlsxDisplay:
    def test_renders_all_sheets(self, client, xlsx_file):
        resp = client.get(f"/api/file/{VAULT}", params={"path": "budget.xlsx"})
        assert resp.status_code == 200
        data = resp.json()
        assert data["is_xlsx"] is True
        assert data["unsupported"] is False
        assert [s["name"] for s in data["xlsx_sheets"]] == ["Budget", "Notes"]

        first = data["xlsx_sheets"][0]["html"]
        assert "Poste" in first
        assert 'data-cell="B1"' in first
        assert "=B1*2" in first  # formula kept as text (data_only=False)
        assert 'data-cell="A1"' in data["xlsx_sheets"][1]["html"]

    def test_corrupt_xlsx_returns_500(self, client, test_vault_dir):
        bad = Path(test_vault_dir) / "corrupt.xlsx"
        bad.write_bytes(b"this is not a zip archive")
        resp = client.get(f"/api/file/{VAULT}", params={"path": "corrupt.xlsx"})
        assert resp.status_code == 500


# ── Index parity (tree visibility) ────────────────────────────────────────


class TestXlsxIndexing:
    def test_xlsx_in_supported_extensions(self):
        from backend.indexer import SUPPORTED_EXTENSIONS

        assert ".xlsx" in SUPPORTED_EXTENSIONS

    def test_xlsx_indexed_metadata_only(self, test_vault_dir, xlsx_file):
        from backend.indexer import _index_single_file_sync

        info = _index_single_file_sync(VAULT, test_vault_dir, xlsx_file)
        assert info is not None
        assert info["extension"] == ".xlsx"
        assert info["content"] == ""  # binary: never read into TF-IDF
        assert info["title"]  # filename-derived title


# ── Edit ──────────────────────────────────────────────────────────────────


class TestXlsxSave:
    def _save(self, client, body, path="budget.xlsx"):
        return client.put(
            f"/api/file/{VAULT}/xlsx/save",
            params={"path": path},
            json=body,
        )

    def test_save_updates_cell_with_number_coercion(self, client, xlsx_file):
        resp = self._save(client, {"sheet": "Budget", "cells": {"B1": "250"}})
        assert resp.status_code == 200
        assert resp.json()["status"] == "ok"

        from openpyxl import load_workbook

        wb = load_workbook(xlsx_file)
        assert wb["Budget"]["B1"].value == 250  # int, not "250"

    def test_save_leaves_other_sheets_and_formulas(self, client, xlsx_file):
        resp = self._save(client, {"sheet": "Budget", "cells": {"A1": "Titre", "B1": 42}})
        assert resp.status_code == 200

        from openpyxl import load_workbook

        wb = load_workbook(xlsx_file)
        assert wb["Budget"]["B2"].value == "=B1*2"
        assert wb["Notes"]["A1"].value == "hello"

    def test_save_empty_string_clears_cell(self, client, xlsx_file):
        resp = self._save(client, {"sheet": "Budget", "cells": {"A1": ""}})
        assert resp.status_code == 200

        from openpyxl import load_workbook

        assert load_workbook(xlsx_file)["Budget"]["A1"].value is None

    def test_save_creates_backup(self, client, xlsx_file):
        from backend.services.backups import get_backup_dir

        resp = self._save(client, {"sheet": "Budget", "cells": {"A1": "backup-me"}})
        assert resp.status_code == 200
        backup_dir = Path(get_backup_dir(VAULT, "budget.xlsx"))
        assert backup_dir.is_dir()
        assert list(backup_dir.glob("*.bak"))

    def test_unknown_sheet_400(self, client, xlsx_file):
        resp = self._save(client, {"sheet": "Nope", "cells": {"A1": "x"}})
        assert resp.status_code == 400

    def test_invalid_cell_ref_400(self, client, xlsx_file):
        resp = self._save(client, {"sheet": "Budget", "cells": {"A-1": "x"}})
        assert resp.status_code == 400

    def test_wrong_extension_400(self, client, test_vault_dir):
        (Path(test_vault_dir) / "note.md").write_text("# hi\n", encoding="utf-8")
        resp = self._save(client, {"sheet": "Sheet", "cells": {"A1": "x"}}, path="note.md")
        assert resp.status_code == 400

    def test_missing_file_404(self, client):
        resp = self._save(client, {"sheet": "Budget", "cells": {"A1": "x"}}, path="absent.xlsx")
        assert resp.status_code == 404

    def test_too_many_cells_400(self, client, xlsx_file):
        cells = {f"A{i}": i for i in range(1, 502)}
        resp = self._save(client, {"sheet": "Budget", "cells": cells})
        assert resp.status_code == 400

    def test_nested_value_400(self, client, xlsx_file):
        resp = self._save(client, {"sheet": "Budget", "cells": {"A1": {"nested": 1}}})
        assert resp.status_code == 400

    def test_missing_sheet_field_400(self, client, xlsx_file):
        resp = self._save(client, {"cells": {"A1": "x"}})
        assert resp.status_code == 400