nits 2

align_data/sources/articles/google_cloud.py

@@ -93,8 +93,8 @@ def iterate_rows(sheet: Worksheet) -> Iterator[SheetRow]:
     SheetRow.set_sheet(sheet)
 
     # we start the enumeration at 2 to avoid the header row
-    for i, row_data in enumerate(sheet.get_all_records(), 2):
-        yield SheetRow(i, row_data)
+    for row_id, row_data in enumerate(sheet.get_all_records(), 2):
+        yield SheetRow(row_id, row_data)
 
 
 def upload_file(filename, bytes_contents, mimetype, parent_id=None):

align_data/sources/articles/indices.py

@@ -25,8 +25,9 @@ def fetcher():
         if contents := fetch_element(url, main_selector):
             return list(filter(None, map(formatter, contents.select(item_selector))))
         return []
-    fetcher.__name__ = formatter.__name__.replace("format_", "") + '_fetcher' 
+    fetcher.__name__ = formatter.__name__.replace("format_", "") + '_fetcher'
     # formatter called "format_anthropic" -> fetcher called "anthropic_fetcher"
+    #TODO: Make this more explicit
     return fetcher
 
 

align_data/sources/arxiv_papers.py

@@ -52,12 +52,6 @@ def get_version(id: str) -> str | None:
         return res.group(1)
 
 
-def is_withdrawn(url: str):
-    if elem := fetch_element(canonical_url(url), ".extra-services .full-text ul"):
-        return elem.text.strip().lower() == "withdrawn"
-    return None
-
-
 def is_withdrawn(url: str) -> bool:
     if elem := fetch_element(canonical_url(url), '.extra-services .full-text ul'):
         return elem.text.strip().lower() == 'withdrawn'

align_data/sources/greaterwrong/greaterwrong.py

@@ -63,11 +63,11 @@ class GreaterWrong(AlignmentDataset):
     start_year: int
     min_karma: int
     """Posts must have at least this much karma to be returned."""
-    af: bool = False
+    af: bool
     """Whether alignment forum posts should be returned"""
 
     limit = 50
-    COOLDOWN_TIME : float = 0.5
+    COOLDOWN_TIME: float = 0.5
     done_key = "url"
     lazy_eval = True
     source_type = 'GreaterWrong'