feat: réutiliser le DuckDB existant si le bootstrap échoue (#78)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

src/db.py

@@ -118,13 +118,24 @@ def _ensure_database() -> Path:
     db_path = Path(os.getenv("DUCKDB_PATH", "./decp.duckdb"))
     parquet_path = os.getenv("DATA_FILE_PARQUET_PATH", "")
     lock_path = db_path.with_suffix(".duckdb.lock")
+    db_exists = db_path.exists()
 
     with open(lock_path, "w") as lock_fd:
         fcntl.flock(lock_fd, fcntl.LOCK_EX)
-        if should_rebuild(db_path, parquet_path):
+        try:
-            build_database(db_path)
+            if should_rebuild(db_path, parquet_path):
-        else:
+                build_database(db_path)
-            logger.debug("Base de données déjà disponible et à jour.")
+            else:
+                logger.debug("Base de données déjà disponible et à jour.")
+        except Exception as e:
+            if db_exists:
+                logger.error(
+                    f"Bootstrap données KO ({e}). "
+                    f"Réutilisation du DuckDB existant : {db_path}"
+                )
+            else:
+                logger.critical("Aucune base DuckDB et reconstruction impossible.")
+                raise
     return db_path
 
 

tests/test_db.py

@@ -312,3 +312,59 @@ def test_concurrent_build_serialized(tmp_path, monkeypatch):
     assert errors == []
     assert db_path.exists()
     assert not tmp_path_artifact.exists()
+
+
+def _raise(*args, **kwargs):
+    raise RuntimeError("boom")
+
+
+def test_ensure_database_reuses_db_when_should_rebuild_raises(tmp_path, monkeypatch):
+    import src.db as db
+
+    dbf = tmp_path / "decp.duckdb"
+    dbf.write_bytes(b"existing")
+    monkeypatch.setenv("DUCKDB_PATH", str(dbf))
+    monkeypatch.setenv("DATA_FILE_PARQUET_PATH", "http://unreachable")
+    monkeypatch.setattr(db, "should_rebuild", _raise)
+    result = db._ensure_database()  # ne doit pas lever
+    assert result == dbf
+    assert dbf.read_bytes() == b"existing"
+
+
+def test_ensure_database_reuses_db_when_build_raises(tmp_path, monkeypatch):
+    import src.db as db
+
+    dbf = tmp_path / "decp.duckdb"
+    dbf.write_bytes(b"existing")
+    monkeypatch.setenv("DUCKDB_PATH", str(dbf))
+    monkeypatch.setenv("DATA_FILE_PARQUET_PATH", "http://unreachable")
+    monkeypatch.setattr(db, "should_rebuild", lambda *a, **k: True)
+    monkeypatch.setattr(db, "build_database", _raise)
+    db._ensure_database()  # ne doit pas lever
+    assert dbf.read_bytes() == b"existing"
+
+
+def test_ensure_database_raises_on_cold_start(tmp_path, monkeypatch):
+    import src.db as db
+
+    dbf = tmp_path / "decp.duckdb"  # n'existe pas
+    monkeypatch.setenv("DUCKDB_PATH", str(dbf))
+    monkeypatch.setenv("DATA_FILE_PARQUET_PATH", "http://unreachable")
+    monkeypatch.setattr(db, "should_rebuild", lambda *a, **k: True)
+    monkeypatch.setattr(db, "build_database", _raise)
+    with pytest.raises(RuntimeError):
+        db._ensure_database()
+
+
+def test_ensure_database_builds_when_needed(tmp_path, monkeypatch):
+    import src.db as db
+
+    dbf = tmp_path / "decp.duckdb"
+    dbf.write_bytes(b"old")
+    monkeypatch.setenv("DUCKDB_PATH", str(dbf))
+    monkeypatch.setenv("DATA_FILE_PARQUET_PATH", "http://x")
+    called = {}
+    monkeypatch.setattr(db, "should_rebuild", lambda *a, **k: True)
+    monkeypatch.setattr(db, "build_database", lambda p: called.setdefault("built", p))
+    db._ensure_database()
+    assert called.get("built") == dbf