Visio .vsdx format unzip and zip corrupts

Question

I'm attempting to modify a Visio file (Open XML format) without having to use the Windows Visio application. My first experiment is just to use 7zip to unzip a known good .vsdx file that was created using Visio. That is all good; I can view the content of the package. Without making any modifications, I use 7zip to re-zip the content and renamed to .vsdx, but when I tried to open the resulting new file using Visio, it complains that the file is corrupt. Is there a way to manually re-zip the content into something that Visio accepts as a valid Visio file? I suspect that there may be some sort of checks for the validity of the file, but can't find what that may be. Thanks for any input.

Answer 1

I would use some form of OpenXML library to get at the file's guts using some sort of "approved magic".

Understanding that you might not want to do whatever you're doing via programming, I looked for some sort of free editor.

I found this free plug-in for Visual Studio:

https://marketplace.visualstudio.com/items?itemName=bsivanov.OpenXMLPackageEditorforVisualStudio

It works in the free "Microsoft Visual Studio Community 2019" as well. I just opened the dev environment (aka: the application) and dragged a Visio .vsdx file into the app. It opened with a tree-like editor. I was able to dig down until I found the visio > pages > page1.xml "leaf". Inside there, I was able to change some text on a shape, then save the "package".

Whatever this tool does, it saves the file properly, and I was able to open the altered .vsdx file in Visio. And the text that I changed in the editor was indeed changed inside of Visio!

Answer 2

I think I've used this in the past:

"Welcome to the Open XML SDK 2.5 for Office" https://learn.microsoft.com/en-us/office/open-xml/open-xml-sdk https://github.com/OfficeDev/Open-XML-SDK

To edit Visio files without the Visio application, you'll still need to understand how Visio works, to some extent.

A simple example:

I changed the text on a shape fairly easily within one of the page.xml files. That was easy. Then I wanted to add a copy of that shape. It was simple enough to copy and paste the whole xml block for the existing shape, then change the PinX and PinY attributes to move the shape to a different location on the page.

But you won't see that shape unless you give it a unique ID within the page. I tested deleting the ID attribute (to see if Visio would figure it out on open and assign one automatically), but it didn't work. If the ID is the same as another shape, the shape is ignored when you open the file. Once I changed ID to something unused, I did see the new copy of the shape.

If you create grouped shapes, or shapes that have advanced behavior (SmartShapes, ShapeSheet formulas, etc.), then this could get complicated. As formulas need to reference other shapes by ID, so you need to manage the IDs! For simple boxes and lines, etc., it might work well (and fast) to generate these things via OpenXML. Good luck!