fix(openchallenges): dumping script quick fixes

[vpchung]

Changelog

• typo fix for the word joiner unicode character (\u2060, not \u2020)
• specify that regex is used for unicode replacement

I have verified that these changes will update the CSVs as expected:

>>> challenges = df[["id", ..., "updated_at"]]
>>> challenges.loc[460, "description"]
'The Food and Drug Administration (FDA) -\u202fCenter for Devices and Radiological Health (CDRH), Sage Bionetworks, and\u202fprecisionFDA call on ...'
>>> 
>>> challenges = (
...         challenges.replace({r"\s+$": "", r"^\s+": ""}, regex=True)
...         .replace(r"\n", " ", regex=True)
...         .replace("'", "''", regex=True)
...         .replace(u"\u2019", "''", regex=True)  # replace curly right-quote 
...         .replace(u"\u202f", " ", regex=True)  # replace narrow no-break space
...         .replace(u"\u2060", "", regex=True)  # remove word joiner
...     )
>>> challenges.loc[460, "description"]
'The Food and Drug Administration (FDA) - Center for Devices and Radiological Health (CDRH), Sage Bionetworks, and precisionFDA call on ....'