Support commas in numeric cells

Author: Marcin Kardas

sota_extractor2/models/linking/bm25_naive.py

@@ -191,8 +191,7 @@ def annotations(r, c, type='model'):
             if type in structure[nr, c]:
                 yield Value(structure[nr, c], matrix[nr, c])
 
-
-    number_re = re.compile(r'^[± Ee /()^0-9.%±_-]{2,}$')
+    number_re = re.compile(r'^[± Ee/()^0-9.%,_+-]{2,}$')
 
     proposals = [Proposal(
         cell=Cell(cell_ext_id=f"{table_ext_id}/{r}.{c}",

sota_extractor2/models/linking/format.py

@@ -29,9 +29,9 @@ def format_to_regexp(format):
     return re.compile('^' + regexp), fn
 
 def extract_value(cell_value, format):
-    cell_value = re.sub(r"\s+%", "%", cell_value)
+    cell_value = re.sub(r"\s+%", "%", cell_value).replace(",", "").strip()
     regexp, fn = format_to_regexp(format)
-    match = regexp.match(cell_value.strip())
+    match = regexp.match(cell_value)
     if match is None or not len(match.groups()):
         return Decimal('NaN')
-    return fn(Decimal(match.group(1)))
+    return fn(Decimal(match.group(1)))