Recovering Corrupted .docx Files

Published: February 14, 2023

Files are .zips?

Did you know that files with the extension .docx are really just .zip packages? I knew that’s essentially what .apk files (Android Package) were but I never really made the connection that this could apply to other common file types as well.

But I recently came across an issue where a document I had spent a significant amount of time on was corrupted. No matter, right? Just grab the backup. Unfortunately, I had overwritten my backed up copy with the corrupted version of the file.

🤦

If I couldn’t find a way to recover it, I would have had to reproduce my changes all over again, effectively doubling my work.

An error prompt which reads: An error occurred during opening the file. This may be caused by incorrect file contents. Do you want to ignore the error and attempt to continue loading the file? There are also yes and no buttons below the prompt. — The error I was shown when attempting to open my .docx file in LibreOffice. When selecting No, LibreOffice Writer would still attempt to open but crash with another error. Selecting Yes will recover the document up to the point where the error was detected.

An error prompt which reads: File format error found at SAXParseException '[word/document.xml line 2]: Opening and ending tag mismatch' — Selecting "No" from the previous error prompt shows this error message.

Fortunately, I am very good at finding information on the Internet and so after a brief search, I came across this thread in the LibreOffice forums:

Take a copy of the ODT and open the copy using an archive manager (rename the file to a .zip extension if necessary). Extract the content.xml file. Open this file with a suitable XML editor. It will be likely that row 2 contains a very long line of XML. Scroll to character 18067. It is likely at this point there will be an obvious mistake in the XML. It may be minor or major. Once you fix it, simple reinsert the fixed version of content.xml into the ODT copy, again using an archive manager. Try and open the amended ODT using LO.
Source

If .odt files are packages of XML files, it seems likely that .docx files are as well, right? Turns out, they are! It is possible to simply change the extension of the document to .zip. From that point, the archived files can be opened and browsed like any normal directory and I can use the information provided in previous error prompts to locate the troublesome XML. The error prompt showed that issue was located in the file word/document.xml on line 2, column 19055. Sure enough, when going there, I found the following:

<w:rPr><w:del w:id="9" w:author="Dylan Hildenbrand" w:date="2023-02-13T14:15:30Z"></w:del></w:ins></w:rPr>

Removing the extra closing tag </w:ins>, saving the file, and renaming the .zip back to a .docx resolved this issue for me. My file was recovered and precious hours were not wasted. And yes dear reader, the irony of finding an extra closing tag that could have cost me several hours is not lost on me; the operator of a website titled closingtags.com.

Archived Comments

These comments have been imported from a previous commenting system, for the sake of posterity. If you left a comment using the old system and would like to have it removed, please get in touch with me using this form.