your OOXML file (“.docx”, “.xlsx”, “.vsdx”, …) and its “modified” timestamp

• https://en.wikipedia.org/wiki/Office_Open_XML_file_formats
• https://github.com/JochenHayek/misc/blob/master/using_timestamps_in_filenames/create_snapshot_from_OOXML.sh
• https://en.wikipedia.org/wiki/XMLStarlet

VSDX does not get listed as an OOXML conform file format, but for this purpose (here) we can treat it like one.

Your “.docx” (or “.xlsx”) file is a ZIP file with a docProps/core.xml inside:

$ unzip -l YOUR.docx
…
… docProps/core.xml
…

This is a convenient way to extract docProps/core.xml to STDOUT:

$ unzip -p YOUR.docx docProps/core.xml
…

This is how to get the XML reformatted using xmlstartlet:

$ unzip -p YOUR.docx docProps/core.xml | xml fo

This command line shows you the possible XPath expressions:

$ unzip -p YOUR.docx docProps/core.xml | xml el
…
cp:coreProperties/dcterms:modified
…

How to extract “modified” to STDOUT?

$ unzip -p YOUR.docx docProps/core.xml | xml sel --template --value-of cp:coreProperties/dcterms:modified

And how to extract the timestamp w/o anything but decimal digits?

$ unzip -p YOUR.docx docProps/core.xml | xml sel --template --value-ofcp:coreProperties/dcterms:modified | tr -d ':TZ-'

…