small fixes

app.py

@@ -1,3 +1,7 @@
+import os
+from functools import lru_cache, wraps
+from typing import Any, Callable
+
 import arxiv
 import gradio as gr
 import pandas as pd
@@ -5,12 +9,26 @@ from apscheduler.schedulers.background import BackgroundScheduler
 from cachetools import TTLCache, cached
 from setfit import SetFitModel
 from tqdm.auto import tqdm
-import os
 
 os.environ["HF_HUB_ENABLE_HF_TRANSFER"] = "1"
 
 CACHE_TIME = 60 * 60 * 12  # 12 hours
-MAX_RESULTS = 1_000
+MAX_RESULTS = 1000
+
+
+# def list_cacheable(func: Callable[..., Any]) -> Callable[..., Any]:
+#     @lru_cache(maxsize=100)
+#     def cacheable_func(*args: Any, **kwargs: Any) -> Any:
+#         return func(*args, **kwargs)
+
+#     @wraps(func)
+#     def wrapper(*args: Any, **kwargs: Any) -> Any:
+#         # Convert lists to tuples to make them hashable
+#         args = tuple(tuple(arg) if isinstance(arg, list) else arg for arg in args)
+#         kwargs = {k: tuple(v) if isinstance(v, list) else v for k, v in kwargs.items()}
+#         return cacheable_func(*args, **kwargs)
+
+#     return wrapper
 
 
 @cached(cache=TTLCache(maxsize=10, ttl=CACHE_TIME))
@@ -94,13 +112,14 @@ all_possible_arxiv_categories = sorted(prepare_data().category.unique().tolist()
 broad_categories = sorted(prepare_data().broad_category.unique().tolist())
 
 
+# @list_cacheable
 def create_markdown_summary(categories=None, new_only=True, narrow_categories=None):
     df = prepare_data()
     if new_only:
         df = df[df["prediction"] == "new_dataset"]
     if narrow_categories is not None:
         df = df[df["category"].isin(narrow_categories)]
-    if categories is not None:
+    if categories is not None and not narrow_categories:
         df = prepare_data()
         if new_only:
             df = df[df["prediction"] == "new_dataset"]