Hi, I'm a Prefect noob, so please bear with me. I seem to be having a strange problem with decoding from byte string. I am using Prefect to perform a query from my database, which has text in RTF format (this is an old database!) in which the text is stored as latin-1 encoded byte strings. The query is performed using

pd.read_sql()

and so the text is then stored in a column in the pandas dataframe. I then decode the strings and use

striprtf

to convert to plain text. I have never had a problem performing this step in Jupyter notebooks or on multiple machines, but when I run this in Prefect, for a portion of the text I get a

UnicodeDecodeError

despite using

text.decode(encoding='latin-1', errors='replace')

. I've tried using

chardet

but have had no luck. Thanks in advance for the help.