ExcelFile.parse() does not return TextFileReader object, but rather a DataFrame

[rkyleg]

When specifying a chunksize during the ExcelFile.parse() method, I should be back an iterable object, but instead a DataFrame is returned.

If you return just the TextParser (change parser.read() to just parser) in the code block below, then iterating over the parsed ExcelFile works and returns the specified number of chunks.

The code block is taken from pandas.io.excel.py line 334

Hopefully I didn't make that confusing.

parser = TextParser(data, header=header, index_col=index_col,
                            has_index_names=has_index_names,
                            na_values=na_values,
                            thousands=thousands,
                            parse_dates=parse_dates,
                            date_parser=date_parser,
                            skiprows=skiprows,
                            skip_footer=skip_footer,
                            chunksize=chunksize,
                            **kwds)
return parser.read()

[jreback]

can you show an example of usage

[rkyleg]

Sure, code below is taken from my script that is using the workbook.parse() method to return the iterable object. Before I changed 'return parser.read()' to just 'return parser' in the code above, this did not work as it would return each 'cell' of data in the dataframe in the 'for chunk' loop rather than each row.

df = workbook.parse(sheetname, chunksize=2000, iterator=True)
for chunk in df:
       df = chunk.where(pandas.notnull(chunk), None)
       df.to_sql(name='table_name', con=self.engine, if_exists='append', index=False)

[mroeschke]

Looks like TextParser doesn't exist anymore. And looks like there there isn't a complete, reproducible example. Happy to reopen if this issue gains more context