diff options
Diffstat (limited to 'fatcat_scholar')
| -rw-r--r-- | fatcat_scholar/schema.py | 11 |
1 files changed, 9 insertions, 2 deletions
diff --git a/fatcat_scholar/schema.py b/fatcat_scholar/schema.py index 9b2d2fb..b93962c 100644 --- a/fatcat_scholar/schema.py +++ b/fatcat_scholar/schema.py @@ -380,8 +380,15 @@ def clean_str(raw: Optional[str], strip_trailing_period: bool = False) -> Option text = ftfy.fix_text(raw) - # remove HTML - text = BeautifulSoup(text, "html.parser").get_text() + # remove HTML tags + try: + # TODO: work_h4ufpvlh3rcefacajni7sdndwa as a regression test + # TODO: consider w3clib "remove tags" as an alternative + clean_text = BeautifulSoup(text, "html.parser").get_text() + text = clean_text + except UnboundLocalError: + # TODO: passing through raw string; what should behavior actually be? + pass # TODO: for performance, compile these as globals? # replaces whitespace with single space |