def extract_csv(head, separator):
"""
csv file => data frame => html
Args:
file_ - file-like object opened in binary mode, pointing to .csv
Returns:
html - html version of *first sheet only* in workbook
info - metadata
"""
warnings_ = io.StringIO()
# this shouldn't balloon memory because head is limited in size by get_preview_lines
try:
data = pandas.read_csv(io.StringIO('\n'.join(head)), sep=separator)
except pandas.errors.ParserError:
# temporarily redirect stderr to capture warnings (usually errors)
with redirect_stderr(warnings_):
data = pandas.read_csv(
io.StringIO('\n'.join(head)),
error_bad_lines=False,
warn_bad_lines=True,
# sep=None is slower (doesn't use C), deduces the separator
sep=None)
html = remove_pandas_footer(data._repr_html_()) # pylint: disable=protected-access
return html, {'note': TRUNCATED, 'warnings': warnings_.getvalue()}
def extract_excel(file_):
"""
excel file => data frame => html
Args:
file_ - file-like object opened in binary mode, pointing to XLS or XLSX
Returns:
html - html version of *first sheet only* in workbook
info - metadata
"""
first_sheet = pandas.read_excel(file_, sheet_name=0)
html = remove_pandas_footer(first_sheet._repr_html_()) # pylint: disable=protected-access
return html, {}