使用C＃Word API从Word文件读取图像而不使用剪贴板

Question

I've been working on an application to read images from multiple word files and store them in one single word file using Microsoft.Office.Interop.Word in C#

EDIT : I also need to save a copy of the images on the file system, so I need the image in a Bitmap or similar object.

This is my implementation so far, which works fine:

        foreach (InlineShape shape in doc.InlineShapes)
        {
            shape.Range.Select();
            if (shape.Type == WdInlineShapeType.wdInlineShapePicture)
            {
                doc.ActiveWindow.Selection.Range.CopyAsPicture();
                ImageData = Clipboard.GetDataObject();
                object _ob1 = ImageData.GetData(DataFormats.Bitmap);
                bmp = (Bitmap)_ob1;
                images[i++] = bmp;
                /*
                bmp.Save("C:\\Users\\Akshay\\Pictures\\bitmaps\\test" + i.ToString() + ".bmp");
                */
            }
        }

I have:

• Selected the images as InlineShapes
• Copied the shape into Clipboard
• Stored the shape in the Clipboard in a DataObject
• Extracted the shape from the DataObject in Bitmap format and stored in a Bitmap object.

I've been told to refrain from using Clipboard in Word automation and use the Word APIs instead. I've read up on it and found an SO answer stating the same.

I looked up many implementations of reading images from Word files on MSDN , SO etc. but could not find any without using clipboard.

How do I read images from Word files using the Word APIs from Microsoft.Office.Interop.Word namespace alone without using Clipboard ?

Answer 1

Word documents in the Office Open XML file format store images in Base64. So it should be possible to extract that information and convert/stream it to a file. You can access the information when the document is open in the Word application using the Range.WordOpenXML property.

string shapeBase64 = shape.Range.WordOpenXML;

This will return the entire Word Open XML in the flat file OPC format. In other words, it won't contain only the picture in Base64, but the entire zip package definition as XML that surrounds it. In my quick test, the tag the contains the actual Base64 is

<pkg:binaryData>

That's a child element of

<pkg:part pkg:name="/word/media/image1.jpg" pkg:contentType="image/jpeg" pkg:compression="store">

Note that it would also be possible for you to get the entire document's WordOpenXML in one step:

document.Content.WordOpenXML

but might then need to understand the way the InlineShapes in the document body are linked to the actual information in the "media" part.

And it would be possible, of course, to work directly with the Zip Package (using the Open XML SDK, perhaps) instead of opening the document in the Word.Application.