Rewrite observations rows using an iterator

eurostat_to_dbnomics.py

@@ -28,7 +28,6 @@ import argparse
 import binascii
 import hashlib
 import io
-import itertools
 import logging
 import os
 import re
@@ -255,17 +254,16 @@ def iter_git_objects_in_sdmx_series_element(series_element, sdmx_file_path, dsd_
             if column_name not in required_observations_tsv_column_names:
                 other_column_names.add(column_name)
     observations_tsv_column_names = required_observations_tsv_column_names + list(sorted(other_column_names))
-    observations_tsv_str = "\n".join([
-        "\t".join(observations_tsv_column_names),
-    ] + [
-        "\t".join(
-            observation.get(column_name, "")
-            for column_name in observations_tsv_column_names
-        )
-        for observation in observations_tsv
-    ])
 
-    observations_tsv_blob = Blob.from_string(observations_tsv_str.encode('utf-8'))
+    def iter_observations_rows():
+        yield "\t".join(observations_tsv_column_names)
+        for observation in observations_tsv:
+            yield "\t".join(
+                observation.get(column_name, "")
+                for column_name in observations_tsv_column_names
+            )
+
+    observations_tsv_blob = Blob.from_string("\n".join(iter_observations_rows()).encode('utf-8'))
     observations_tsv_blob_id = observations_tsv_blob.id
     # Some TSV observations files are exactly the same, and Git packs don't tolerate duplicate objects.
     if observations_tsv_blob_id not in yielded_git_object_ids: