DOCX needs repaired, but repair doesn't work and validator says there's no issues

Connor Shore 186

Hello,

I have a question about a docx file. It was generated using our own doc gen engine. When you open the file it says the document needs to be repaired, however when you run the file through the OOXML validator, it says there are no issues with the document. Another bizarre thing is when you open the document, repair it and then save the repaired version, when you re-open the repaired document after closing, it says it needs to be repaired again. We haven't been able to identify what is causing the document be in the "need repairs" state. Could you identify the issue that is causing this?

Link to files: https://windwardstudios.box.com/s/io6idp7l61y28ubvdtwgschoe8w7yt8e

Thank you,

Connor

Tom Jebo 1,991 Reputation points Microsoft Employee

2023-08-30T21:14:18.62+00:00

Hi @Connor Shore

I'll take a look at these documents and see if I can identify the issue with Word.

Best regards,
Tom Jebo
Microsoft Open Specifications Support
Matt Binsfeld 0 Reputation points

2023-09-07T21:59:46.6333333+00:00

Hey Tom,

Any update on this? We want to fix our software to produce valid docx in this use case. We're struggling to do so given that the OpenXML SKD Productivity tool says it's valid with zero errors or warnings.

Thanks,

Matt Binsfeld
Tom Jebo 1,991 Reputation points Microsoft Employee

2023-09-07T22:07:19.2166667+00:00

Sorry Matt, I'm still working on it.

Tom
Tom Jebo 1,991 Reputation points Microsoft Employee

2023-09-08T00:35:40.98+00:00

@Matt Binsfeld /@Connor Shore ,

Was this document based on an existing document or did your software create this programmatically from scratch? If there was an existing document, did it also trigger Word's "unreadable content" dialog and can you share that document as well?

Tom
Connor Shore 186 Reputation points

2023-09-08T14:29:55.5633333+00:00

@Tom Jebo Hey Tom, I've attached the template the output was generated from within the box.com link so you should have access to it now. The field codes in the template are essentially our indicators where we bring in data from their database to create our output just fyi. The original template does not display any unreadable content message.
Matt Binsfeld 0 Reputation points

2023-09-08T19:11:14.0266667+00:00

Hi Tom,

Just a little bit more background information. This document was generated by our product, Fluent, was previously known as Windward. I believe both myself and our old CEO Dave have worked with you on similar issues prior to our acquisition.

Thanks,

Matt Binsfeld
Matt Binsfeld 0 Reputation points

2023-09-15T20:50:17.9+00:00

Hey Tom,

Just wondering if you have any updates on this?

Thanks,

Matt
Tom Jebo 1,991 Reputation points Microsoft Employee

2023-09-15T21:52:53.8+00:00

Hi Matt,

Just wondering if you have any updates on this?

Unfortunately, no. Nothing yet.

Just a little bit more background information. This document was generated by our product, Fluent, was previously known as Windward. I believe both myself and our old CEO Dave have worked with you on similar issues prior to our acquisition.

<I wrote this a few days ago and thought I had hit "Add comment" but apparently, not. Sorry.>

yes, I kinda figured that out. Was the document created completely by code or did you start with a template and add to it? This one is a bit puzzling because all the changes made in repair mode seem to be reorganization or removal of unnecessary XML. When I test these changes individually they don't seem to bother Word. I've been stepping through Word's doc open and part parsing code and still can't seem to identify what it doesn't like that puts it into "repair" mode. That part is a slow process and I'm still working through it. Sorry about the slow progress.

I did notice that Word Online (which is quite different than desktop) opens the file but just seems to show a bunch of the images as unloadable. I've removed some of them but desktop Word still doesn't like the file.

Tom
Tom Jebo 1,991 Reputation points Microsoft Employee

2023-09-15T22:42:09.51+00:00

One observation which may just be another reorganization by-product of Word is that the v:line | o:gfxdata in the Fallback blocks in document.xml, or at least one of them, changes in content after the repair.

1 answer

Connor Shore 186 Reputation points

2023-09-21T20:39:02.4433333+00:00

Hey Tom,

I was able to figure out the source of the issue. In the document.xml there is a hyperlink with rId3 and it keeps getting repeated throughout in wrong locations. Essentially we weren't closing hyperlinks off correctly in our code and it was adding them a bunch of places they shouldn't be. This issue is now fixed and can be closed. Thank you for your help!
Please sign in to rate this answer.

1 person found this answer helpful.
Tom Jebo 1,991 Reputation points Microsoft Employee

2023-09-21T23:11:33.6666667+00:00

Hi @Connor Shore

Thanks for letting me know. That was one of the first things I saw (but I thought it was rId38 and I didn't see anything wrong with it) when I started looking at this document but after trying to replace the hyperlink several different ways, it didn't seem to help. Anyway, I'm very glad you figured it out.

Best regards, Tom Jebo Microsoft Open Specifications Support
Sign in to comment

Share via

DOCX needs repaired, but repair doesn't work and validator says there's no issues

1 answer