fix: cache beahviour

Dockerfile

@@ -12,6 +12,7 @@ WORKDIR $APP_CODE
 COPY ./requirements.txt $APP_CODE/requirements.txt
 COPY ./packages.txt $APP_CODE/packages.txt
 
+ENV APP_CODE $APP_CODE
 RUN apt-get update && xargs -r -a $APP_CODE/packages.txt apt-get install -y && rm -rf /var/lib/apt/lists/*
 RUN pip3 install --no-cache-dir --upgrade pip
 RUN pip3 install --no-cache-dir -r $APP_CODE/requirements.txt
@@ -33,11 +34,11 @@ RUN mkdir -p /data && chown root:user /data && chmod ug+rwX /data
 # User
 USER user
 ENV APP_DATA $APP_DATA
-ENV APP_CODE $APP_CODE
+RUN mkdir -p $APP_DATA
 ENV HOME $APP_HOME
 ENV PATH $HOME/.local/bin:$PATH
 ENV HF_HOME $HF_HOME
-RUN mkdir -p $HF_HOME $APP_DATA
+RUN mkdir -p $HF_HOME
 
 # Expose streamlit application
 WORKDIR $APP_HOME

src/app.py

@@ -5,7 +5,6 @@ import shelve
 import os
 import pathlib
 
-
 st.title("Uber pickups in NYC")
 
 DATE_COLUMN = "date/time"
@@ -16,12 +15,21 @@ DATA_URL = (
 
 DATA_PATH = pathlib.Path(os.environ.get("APP_DATA"))
 
-
 def cached_download_csv(url : str, **kwargs) -> pd.DataFrame:
-    with shelve.open(DATA_PATH.joinpath("shelf.dat").as_posix(), writeback=True, flag='c') as shelf:
-         df = shelf.get(url, pd.read_csv(url, **kwargs))
-         return df
-
+    shelf_filename = DATA_PATH.joinpath("shelf.dat").as_posix()
+    with shelve.open(shelf_filename, writeback=True, flag='c') as shelf:
+        st.info(f"Opened or created the shelf file '{shelf_filename}'")
+        maybe_cached = shelf.get(url)
+        if maybe_cached is None:
+            st.info(f"Downloawding URL '{url}'")
+            df = pd.read_csv(url, **kwargs)
+            shelf[url] = df
+            return df
+        else :  
+            st.info(f"Re-using cached URL '{url}'")
+            assert isinstance(maybe_cached, pd.DataFrame)
+            return maybe_cached
+    # The context manager 'shelf' ensure set value as written back or return 
 
 @st.cache_resource
 def load_data(nrows):