[FIXED] pandas: how to read html and casting all fields to string

Issue

If I import this HTML file

pd_df = pd.read_html('./output.html')
pd_df[0]

the last field becomes a float, but it's a string (in example from 05269 to 5269.0). I know I can apply pd_df = pd.read_html('./output.html',converters={'CAP': str}), but my question is: is there a way to apply globally str casting to all fields, using read_html?

Because this is an example file, this is a simple example, but often I have a lot of fields and a global option would be great.

Solution

(edited)

What you can do is read it twice. Once to get the column names, and then another time with converters to make sure all columns are read as str:

url = "https://gist.githubusercontent.com/aborruso/599153968878f452bd3c68f3de0f29c4/raw/1156d224a4290393409ceef285c238c09b6bd08e/input.html"
df = pd.read_html(url)[0]


converters = {c:lambda x: str(x) for c in df.columns}

df = pd.read_html(url, converters=converters)[0]
print(df)
# results in: 

  Beneficiario       Comune    CAP Provincia Importo
0  RNDFNC60E16  RIPACANDIDA  85020   POTENZA   09269
1  RNDFNC60E16          NaN    NaN   POTENZA   05269

print(df.dtypes) 
# results in : 
Beneficiario    object
Comune          object
CAP             object
Provincia       object
Importo         object
dtype: object

Answered By - Roy2012