Create a pandas dataframe from a qrc resource file

问题:

I would like to save a CSV file into a qrc file and than read it putting its contents in a pandas dataframe, but I have some problems.

I created a qrc file called res.qrc:

<!DOCTYPE RCC><RCC version="1.0">
  <qresource>
    <file>dataset.csv</file>
  </qresource>
</RCC>

I compiled it obtaining the res_rc.py file.

To read it I created a python script called resource.py:

import pandas as pd
import res_rc
from PySide.QtCore import *

file = QFile(":/dataset.csv")
df = pd.read_csv(file.fileName())
print(df)

But I obtain the error: IOError: File :/dataset.csv does not exist

All the files (resource.py, res.qrs, res_rc.py, dataset.csv) are in the same folder.

If I do res_rc.qt_resource_data I can see the contents.

How can I create the pandas dataframe?

回答1:

The qresource is a virtual path that only Qt knows how to obtain it and can change internally without warnings, in these cases what must be done is to read all the data and convert it into a stream with io.BytesIO

import io
import pandas as pd
from PySide import QtCore
import res_rc


file = QtCore.QFile(":/dataset.csv")
if file.open(QtCore.QIODevice.ReadOnly):
    f = io.BytesIO(file.readAll().data())
    df = pd.read_csv(f)
    print(df)