riccardobucco / wikipedia_downloader Goto Github PK

Description: Some columns might be incompatible with utf-8 (for example: pagelinks.cl_sortkey).
Current behavior: All the columns whose values are strings are decoded with utf-8.
Expected behavior: Some values can't be decoded with utf-8, because they represent a sequence of bytes.

Current behavior: When you install the module through pip in a clean virtual environment, and then you try to import it, an error occurs due to a missing dependency (pandas).
Expected behavior: All the required dependencies should be automatically installed.

Description: It would be nice to have the possibility to specify the dtypes of each column of the dataframe returned by the get_dataframe function.
Current behavior: The get_dataframe function automatically infers the dtype of each column.
Expected behavior: The user should be able to specify the dtypes. There could also be some predefined dtypes, based on the table and column names.

Description: The get_dataframe function doesn't work when the where parameter is not explicitly specified.
Current behavior: When you try to call the function without specifying the where parameter (for example wpd.get_dataframe("en", "page"), you get the error
TypeError: 'NoneType' object is not iterable
Expected behavior: The absence of the where parameter means that no filters have to be applied.