This is the simple code I've been testing with:

importdatetimefromdocximportDocumentSVG="""<svg version="1.1"     width="300" height="200"     xmlns="http://www.w3.org/2000/svg">  <rect width="100%" height="100%" fill="red" />  <circle cx="150" cy="100" r="80" fill="green" />  <text x="150" y="125" font-size="60" text-anchor="middle" fill="white">SVG</text></svg>"""svg_filename="drawing.svg"withopen(svg_filename,"w")asf:f.write(SVG)document=Document()now=datetime.datetime.now()document.add_heading("Document Title",0)document.add_heading(f"{now}",0)document.add_picture(svg_filename)document.save("demo.docx")

Hi@takis thanks for the improvement - for my SVG files there are relative scalings in the width and height attributes. I had to take the data from the viewbox attribute. This is a complete 'hack' and probably does not respect SVG specification regarding pt/pixel/ppi etc.

Example svg header:
<svg baseProfile='full' height='100.000000%' version='1.1' viewBox='0.000000 0.000000 1240.000000 1753.000000' width='100.000000%' xmlns='http://www.w3.org/2000/svg' xmlns:xlink='http://www.w3.org/1999/xlink'>

in svg.py:

    @classmethod    def _dimensions_from_stream(cls, stream):        stream.seek(0)        data = stream.read()        root = ET.fromstring(data)        # FIXME: The width could be expressed as '4cm'        # See https://www.w3.org/TR/SVG11/struct.html#NewDocument        width_str = root.attrib["width"]        height_str = root.attrib["height"]        if width_str.endswith(r'%') and height_str.endswith(r'%'):            vb_lims = str(root.attrib["viewBox"]).split(' ')            width_scale = float(width_str[:-1]) * 0.01 * 72/96            height_scale = float(height_str[:-1]) * 0.01 * 72/96            width = int(width_scale * float(vb_lims[2]))            height = int(height_scale * float(vb_lims[3]))        elif width_str.endswith('pt') and height_str.endswith('pt'):            width = int(float(width_str[:-2]))            height = int(float(height_str[:-2]))        else:            width = int(width_str)            height = int(height_str)        return width, height

Hi@takis, I'm new using docx (and python), and I'm tryng to load some SVG images in a word document, looking for that I found this, but try to replicate the code you were testing with and it doesn't work for me, I get the docx.image.exceptions.UnrecognizedImageError. So it isn't "included" in docx now? if not, is there a way for using your code? thanks in advance, and sorry if it's not a place for asking this, i'll remove the comment if needed.

This would be amazing if it could be merged into a new release. In the meantime, if somebody wants to play around with SVG files, I created a monkey patch just as was made in the case for EMF files:#196 (comment). Just add the code from the gist below to a file calleddocx_svg_patch.py, and at the top of your own python script you haveimport docx_svg_patch. I haven't tested this extensively, and I am not sure how well this plays with other image types now, but it's at least a start.

Link to gist:https://gist.github.com/Kladdy/d3bdb9bbf2c38d4f194ea9a7904fc3f2

I've created a new pull request for this, as there were many conflicts with the updated code base:
#1343

Hi@takis, I'm new using docx (and python), and I'm tryng to load some SVG images in a word document, looking for that I found this, but try to replicate the code you were testing with and it doesn't work for me, I get the docx.image.exceptions.UnrecognizedImageError. So it isn't "included" in docx now? if not, is there a way for using your code? thanks in advance, and sorry if it's not a place for asking this, i'll remove the comment if needed.

Hi@JaverHumberto, no, unfortunately, the code was not merged. I've created a new pull request and updated the code though. If you want to try it, you can use my fork for now:
https://github.com/takis/python-docx/

Hi@Kladdy

I've updated the pull request and I'm also still hoping that it will get merged one day.

Do you have any updates on this?
The SVG files are still not compatible with python-docx?

Hi@goomesthiago

Unfortunately not, but I do try to keep my fork in sync with the main repository.