Search code examples
pythonbase64python-docx

Unable to add base64 encoded image to word document using docx


I need to add image to a word document using python-docx module. From base64 encoded data is converted to base64 bytes, and then to image stream. Now I tried to add the image stream to word document.

import io
from docx import Document
from docx.shared import Inches

base64_img = 'iVBORw0KGgoAAAANSUhEUgAAABAAAAAQCAYAAAAf8/9hAAAACXBIWXMAAAsTAAA' \
            'LEwEAmpwYAAAB1klEQVQ4jY2TTUhUURTHf+fy/HrjhNEX2KRGiyIXg8xgSURuokX' \
            'LxFW0qDTaSQupkHirthK0qF0WQQQR0UCbwCQyw8KCiDbShEYLJQdmpsk3895p4aS' \
            'v92ass7pcfv/zP+fcc4U6kXKe2pTY3tjSUHjtnFgB0VqchC/SY8/293S23f+6VEj' \
            '9KKwCoPDNIJdmr598GOZNJKNWTic7tqb27WwNuuwGvVWrAit84fsmMzE1P1+1TiK' \
            'MVKvYUjdBvzPZXCwXzyhyWNBgVYkgrIow09VJMznpyebWE+Tdn9cEroBSc1JVPS+' \
            '6moh5Xyjj65vEgBxafGzWetTh+rr1eE/c/TMYg8hlAOvI6JP4KmwLgJ4qD0TIbli' \
            'TB+sunjkbeLekKsZ6Zc8V027aBRoBRHVoduDiSypmGFG7CrcBEyDHA0ZNfNphC0D' \
            '6amYa6ANw3YbWD4Pn3oIc+EdL36V3od0A+MaMAXmA8x2Zyn+IQeQeBDfRcUw3B+2' \
            'PxwZ/EdtTDpCPQLMh9TKx0k3pXipEVlknsf5KoNzGyOe1sz8nvYtTQT6yyvTjIax' \
            'smHGB9pFx4n3jIEfDePQvCIrnn0J4B/gA5J4XcRfu4JZuRAw3C51OtOjM3l2bMb8' \
            'Br5eXCsT/w/EAAAAASUVORK5CYII='

base64_img_bytes = base64_img.encode('utf-8')
image = io.BytesIO(base64_img_bytes)
document = Document()
document.add_picture(image, width=Inches(1.25))
document.save('./data_temp/demo.docx')

However, I got error:

UnrecognizedImageError 

Does anyone know what I did wrong?

The base64 encoded data has no problem. I tried below and it works fine. Converting the data to png file works fine.

import base64

base64_img = 'iVBORw0KGgoAAAANSUhEUgAAABAAAAAQCAYAAAAf8/9hAAAACXBIWXMAAAsTAAA' \
            'LEwEAmpwYAAAB1klEQVQ4jY2TTUhUURTHf+fy/HrjhNEX2KRGiyIXg8xgSURuokX' \
            'LxFW0qDTaSQupkHirthK0qF0WQQQR0UCbwCQyw8KCiDbShEYLJQdmpsk3895p4aS' \
            'v92ass7pcfv/zP+fcc4U6kXKe2pTY3tjSUHjtnFgB0VqchC/SY8/293S23f+6VEj' \
            '9KKwCoPDNIJdmr598GOZNJKNWTic7tqb27WwNuuwGvVWrAit84fsmMzE1P1+1TiK' \
            'MVKvYUjdBvzPZXCwXzyhyWNBgVYkgrIow09VJMznpyebWE+Tdn9cEroBSc1JVPS+' \
            '6moh5Xyjj65vEgBxafGzWetTh+rr1eE/c/TMYg8hlAOvI6JP4KmwLgJ4qD0TIbli' \
            'TB+sunjkbeLekKsZ6Zc8V027aBRoBRHVoduDiSypmGFG7CrcBEyDHA0ZNfNphC0D' \
            '6amYa6ANw3YbWD4Pn3oIc+EdL36V3od0A+MaMAXmA8x2Zyn+IQeQeBDfRcUw3B+2' \
            'PxwZ/EdtTDpCPQLMh9TKx0k3pXipEVlknsf5KoNzGyOe1sz8nvYtTQT6yyvTjIax' \
            'smHGB9pFx4n3jIEfDePQvCIrnn0J4B/gA5J4XcRfu4JZuRAw3C51OtOjM3l2bMb8' \
            'Br5eXCsT/w/EAAAAASUVORK5CYII='
base64_img_bytes = base64_img.encode('utf-8')
with open('./decoded_image.png', 'wb') as file_to_save:
    decoded_image_data = base64.decodebytes(base64_img_bytes)
    file_to_save.write(decoded_image_data)

Solution

  • add_picture expects normal image, not base64. You forgot convert base64 to normal data

     img_bytes = base64.b64decode(base64_img)
    

    or

     img_bytes = base64.decodebytes(base64_img.encode('utf-8'))
    

    import io
    from docx import Document
    from docx.shared import Inches
    import base64
    
    base64_img = 'iVBORw0KGgoAAAANSUhEUgAAABAAAAAQCAYAAAAf8/9hAAAACXBIWXMAAAsTAAA' \
                'LEwEAmpwYAAAB1klEQVQ4jY2TTUhUURTHf+fy/HrjhNEX2KRGiyIXg8xgSURuokX' \
                'LxFW0qDTaSQupkHirthK0qF0WQQQR0UCbwCQyw8KCiDbShEYLJQdmpsk3895p4aS' \
                'v92ass7pcfv/zP+fcc4U6kXKe2pTY3tjSUHjtnFgB0VqchC/SY8/293S23f+6VEj' \
                '9KKwCoPDNIJdmr598GOZNJKNWTic7tqb27WwNuuwGvVWrAit84fsmMzE1P1+1TiK' \
                'MVKvYUjdBvzPZXCwXzyhyWNBgVYkgrIow09VJMznpyebWE+Tdn9cEroBSc1JVPS+' \
                '6moh5Xyjj65vEgBxafGzWetTh+rr1eE/c/TMYg8hlAOvI6JP4KmwLgJ4qD0TIbli' \
                'TB+sunjkbeLekKsZ6Zc8V027aBRoBRHVoduDiSypmGFG7CrcBEyDHA0ZNfNphC0D' \
                '6amYa6ANw3YbWD4Pn3oIc+EdL36V3od0A+MaMAXmA8x2Zyn+IQeQeBDfRcUw3B+2' \
                'PxwZ/EdtTDpCPQLMh9TKx0k3pXipEVlknsf5KoNzGyOe1sz8nvYtTQT6yyvTjIax' \
                'smHGB9pFx4n3jIEfDePQvCIrnn0J4B/gA5J4XcRfu4JZuRAw3C51OtOjM3l2bMb8' \
                'Br5eXCsT/w/EAAAAASUVORK5CYII='
    
    img_bytes = base64.b64decode(base64_img)
    #img_bytes = base64.decodebytes(base64_img.encode('utf-8'))
    
    #print(img_bytes[:4]) # b'\x89PNG'
    
    image = io.BytesIO(img_bytes) 
    
    document = Document()
    document.add_picture(image, width=Inches(1.25))
    document.save('demo.docx')