Explorer improve filtering of available runs (#693)

Bugfix to only show runs that have existing manifests in them in the
explorer.

data_explorer/app/config.py

@@ -6,6 +6,7 @@
     "selected_pipeline_path",
     "run",
     "selected_run_path",
+    "available_runs",
     "component",
     "selected_component_path",
     "partition",

data_explorer/app/interfaces/common_interface.py

@@ -97,12 +97,19 @@ def _select_run():
         """Selects a run from available runs within the chosen pipeline."""
         selected_pipeline_path = st.session_state["pipeline_path"]
 
-        available_runs = [
-            os.path.basename(item)
-            for item in os.listdir(selected_pipeline_path)
-            if os.path.isdir(os.path.join(selected_pipeline_path, item))
-            and item != "cache"
-        ]
+        def has_manifest_file(path):
+            return any("manifest.json" in files for _, _, files in os.walk(path))
+
+        available_runs = []
+        for run in os.listdir(selected_pipeline_path):
+            run_path = os.path.join(selected_pipeline_path, run)
+            if (
+                os.path.isdir(run_path)
+                and run != "cache"
+                and has_manifest_file(run_path)
+            ):
+                available_runs.append(os.path.basename(run))
+
         available_runs.sort(reverse=True)
 
         default_index = get_default_index("run", available_runs)
@@ -111,6 +118,7 @@ def _select_run():
 
         st.session_state["run"] = selected_run
         st.session_state["run_path"] = selected_run_path
+        st.session_state["available_runs"] = available_runs
 
     def create_common_interface(self):
         """Sets up the Streamlit app's main interface with common elements."""

data_explorer/app/main.py

@@ -116,13 +116,10 @@ def get_pipeline_info_df(self):
         """Get the pipeline info dataframe."""
         selected_pipeline_path = st.session_state["pipeline_path"]
 
+        available_runs = st.session_state["available_runs"]
         available_runs_path = [
-            os.path.join(selected_pipeline_path, item)
-            for item in os.listdir(selected_pipeline_path)
-            if os.path.isdir(os.path.join(selected_pipeline_path, item))
-            and item != "cache"
+            os.path.join(selected_pipeline_path, run) for run in available_runs
         ]
-        available_runs_path.sort(reverse=True)
 
         data = []
         for run_path in available_runs_path: