codenames-2-proper-nouns.py

# Read the file into a dataframe.
with open('dictionary/instance_types_en.ttl', 'r', encoding="utf8") as f:
    f.readline()
    for l in f.readlines():
        try:
            split = l.split(' ')
            rows.append((split[0].split('/')[-1], split[2]))
        except:
            print(l)
            
df = pd.DataFrame.from_records(rows[:-1], columns=['object', 'type'])

# Clean up the formatting and filter it.
df = df[df['type'].str.contains('dbpedia.org/ontology')].copy()
df['object'] = df['object'].str.strip(
    r'[<>]'
).str.split(
    '__', expand=True
)[0].str.replace(
    '_', ' '
).str.replace(
    r'\(.*\)', ''
).str.strip().drop_duplicates()

df['object'] = df['object'].str.replace(r'\s+', ' ')
df['object'] = df['object'].str.replace('%22', '"')
df['object'] = df['object'].str.replace('%3F', '?')

df['type'] = df['type'].str.strip(r'[<>]').str.split('/', expand=True)[4]

df = df[~(df['object'] == '')].copy()