Feat s3 loading (#177)

* refactor: improved txt reader

* remove print

* feat: load swcs from s3

* feat: read json

* bug: read local txt

* bug: swc loading, merge sites

---------

Co-authored-by: anna-grim <anna.grim@alleninstitute.org>

Author: anna-grim

src/segmentation_skeleton_metrics/data_handling/swc_loading.py

@@ -329,6 +329,7 @@ def read_from_gcs_swcs(self, bucket_name, swc_paths):
                 threads.append(
                     executor.submit(self.read_from_gcs_swc, bucket_name, path)
                 )
+                break  # TEMP
 
             # Store results
             swc_dicts = deque()
@@ -475,11 +476,6 @@ def read_from_s3_swcs(self, bucket_name, swc_paths):
             Dictionaries whose keys and values are the attribute names and
             values from an SWC file.
         """
-        # List filenames
-        bucket_name, prefix = util.parse_cloud_path(s3_path)
-        swc_paths = util.list_s3_paths(bucket_name, prefix, extension=".swc")
-
-        # Parse SWC files
         swc_dicts = deque()
         for path in swc_paths:
             content = util.read_txt(bucket_name, path).splitlines()
@@ -518,7 +514,6 @@ def read_from_s3_zips(self, bucket_name, zip_paths):
                 )
 
             # Store results
-            pbar = tqdm(total=len(processes), desc="Read SWCs")
             swc_dicts = deque()
             for process in as_completed(processes):
                 result = process.result()

src/segmentation_skeleton_metrics/skeleton_metrics.py

@@ -525,7 +525,7 @@ def verify_site(self, gt_graph, fragment_graph, gt_node, fragment_node):
         self.fragments_with_merge.add(fragment_graph.name)
         self.merge_sites.append(
             {
-                "Segment_ID": fragment_graph.name,
+                "Segment_ID": fragment_graph.segment_id,
                 "GroundTruth_ID": gt_graph.name,
                 "Voxel": tuple(map(int, voxel)),
                 "World": tuple([float(round(t, 2)) for t in xyz]),
@@ -924,7 +924,7 @@ def __call__(self, gt_graphs, fragment_graphs, merge_sites):
         """
         # Check if merge sites is non-empty
         if len(merge_sites) == 0:
-            return _
+            return None
 
         # Compute metric
         pbar = self.get_pbar(len(merge_sites))