Пофиксил баг с таймаутом

app/main.py

@@ -1,7 +1,5 @@
-import asyncio
 import base64
 import io
-import multiprocessing
 import os
 from contextlib import asynccontextmanager
 
@@ -170,62 +168,20 @@ async def get_index(request: Request):
     return templates.TemplateResponse("index.html", {"request": request})
 
 
-def _eval_worker(formula_str, mode, return_dict):
-    """Изолированный рабочий процесс для опасных вычислений SymPy"""
-    try:
-        if mode == "render":
-            x_dense = np.linspace(-10, 10, 500)
-            formula_str_clean = (
-                formula_str.replace("^", "**").replace("np.", "").replace("math.", "")
-            )
-            t = standard_transformations + (implicit_multiplication_application,)
-            expr = parse_expr(formula_str_clean, transformations=t)
-            y_dense = evaluate_expr_to_points(expr, x_dense)
-            return_dict["result"] = (x_dense.tolist(), y_dense)
-
-        elif mode == "ideal":
-            x, y, mask = get_ideal_math(formula_str)
-            return_dict["result"] = (x, y, mask)
-    except Exception as e:
-        return_dict["error"] = str(e)
-
-
-async def safe_eval_math(formula_str, mode="render", timeout=2.0):
-
-    loop = asyncio.get_running_loop()
-
-    def _run_process():
-        manager = multiprocessing.Manager()
-        return_dict = manager.dict()
-        p = multiprocessing.Process(
-            target=_eval_worker, args=(formula_str, mode, return_dict)
-        )
-        p.start()
-        p.join(timeout)
-
-        if p.is_alive():
-            p.terminate()
-            p.join()
-            return {"error": f"Таймаут: слишком сложная формула: {formula_str}"}
-
-        if "error" in return_dict:
-            return {"error": return_dict["error"]}
-
-        return {"result": return_dict.get("result", (None, None, None))}
-
-    return await loop.run_in_executor(None, _run_process)
-
-
 @app.post("/render_formula")
 async def render_formula(req: FormulaRequest):
-    res = await safe_eval_math(req.formula, mode="render", timeout=2.0)
-
-    if "error" in res:
+    x_dense = np.linspace(-10, 10, 500)
+    try:
+        formula_str = (
+            req.formula.replace("^", "**").replace("np.", "").replace("math.", "")
+        )
+        t = standard_transformations + (implicit_multiplication_application,)
+        expr = parse_expr(formula_str, transformations=t)
+        y_dense = evaluate_expr_to_points(expr, x_dense)
+        return {"x": x_dense.tolist(), "y": y_dense}
+    except Exception:
         return {"x": [], "y": []}
 
-    x_dense, y_dense = res["result"]  # type: ignore
-    return {"x": x_dense, "y": y_dense}
-
 
 @app.post("/predict")
 async def predict(data: PredictRequest):
@@ -239,12 +195,7 @@ async def predict(data: PredictRequest):
     else:
         if not data.formula:
             return {"error": "Введите формулу."}
-
-        res = await safe_eval_math(data.formula, mode="ideal", timeout=2.0)
-        if "error" in res:
-            return {"error": res["error"]}
-
-        x_math, y_math, mask = res["result"]
+        x_math, y_math, mask = get_ideal_math(data.formula)
 
     if mask is None or not np.any(mask):
         return {"error": "Не удалось распознать математику."}