Hey I have a question regarding the option in
Files - Import - Dates - Use content creation & modification dates
What am I doing. I scan documents with Devonthink and then import them into the database
What I thought this option is doing. Check the scanned document for when the doc was created and edit the content creation date of the file.
What really happens both dates are just the date when the document was scanned and imported
Am I doing something wrong or do I understand this option wrong?
The confusion might stem from what you consider “the document” to be: the physical paper, or the digital file produced by scanning?
This option is about the digital file. PDFs can store creation/modification dates as internal metadata embedded in the file itself. These dates can be different from what the file system shows—you often see that with files you download.
For example:
– the creation date embedded in the file linked in that post is 2018-04-17 19:45. But when you download it, the creation date recorded by the file system is the moment it was written to your hard drive.
When you scan a file yourself, the PDF’s internal creation date and the file system’s creation date is likely identical.
If you wish to set the creation date based on the text content of the document, you can use a smart rule.
I need to clarify the documentation on this point.
What you are referring to is the Document Date, i.e., a date in the text content of the document. The setting you are referring to examines the underlying code in the document to get the creation and modification dates of the file.
Here is a PDF on the dangers of eating laundry detergent pods.
I just downloaded it so the file has a creation and modification date of 10.25.2025. That is accurate as far as the file on my Mac goes.
However, the actual date of creation and modification is years ago. With Files > Import > Use content creation and modification dates enabled, the PDF document properly shows 1.30.2020 for those dates.
FAQ_Detergent-Pods.pdf.zip (99.0 KB)
To use the Document Date, you can use a smart rule, batch process, or a script.
For the nerdy or more curious, this information is parsed from the PDF metadata, in this case xmp metadata…
<xmp:ModifyDate>2020-01-30T14:09:24-06:00</xmp:ModifyDate>
<xmp:CreateDate>2020-01-30T14:09:21-06:00</xmp:CreateDate>
In other cases, where it’s not xmp, it would be in an Info (metadata) object like this…
11 0 obj
<< /Title (Untitled 10) /Producer (macOS Version 15.7 \(Build 24G214\) Quartz PDFContext)
/Author (HeraX) /Creator (TextEdit) /CreationDate (D:20250824211725Z00'00')
/ModDate (D:20250824211725Z00'00') >>
endobj
hm ok then it is not working for me. My process is I scann documents with Devonthink into the library with the option “use content creation & modification dates” enabled. But still the created date information for the imported file is always the day from when I imported it into the library rather than any date from within the document itself
Actually, yes it is working properly. As I just illustrated, that setting is not doing what you think it does. And if you’re scanning in DEVONthink, the creation date is logically the date you scanned it.
As @troejgaard and I have mentioned, you need to use a smart rule, batch process, or a script to use the Document Date, which is a date parsed from the content.
Here is a simple batch process example in DEVONthink 4…
There is no content creation date in the sense of DT in something that you scan. You’d have OCR the PDF and then try to get a creation date from it.
thanks for clarification
I have that already created in a smart rule but I thought with the option I could remove that smart rule. But now I also created a batch process. never used them before
Here’s a quick read for you…
