Skip to main content

Files we read

4 minute read

The kinds of files ProcuLink reads, and what to expect from each.

Files every flow reads

  • PDF files with text in them. If you can select the words on the page, we can read it.
  • CSV files.
  • Excel files (.xlsx).

Scans and photos

We also read pages that are only a picture: photos or scans saved as PNG or JPG, and PDFs that have no text in them.

For a picture, each value we read shows where on the page it was found. A value we could not find in the page's text is marked Not in the page's text. Check those with extra care.

Long files

We read a long PDF ten pages at a time, up to 100 pages. Pages after the first 100 are not read and do not count. A check tells you the document was too long to read in full.

Other files

We read Word documents (.docx), plain text (.txt), XML files and JSON files. A flow's Upload area lists exactly the types it accepts.

For a Word file, the original opens as a download instead of on screen. You can still correct values by typing them, but you cannot point at the page. See Teach a value by pointing at the page.

Limits

  • A file can be up to 10 MB.
  • A file that is password-protected or damaged cannot be opened. Remove the password, or ask the sender for a fresh copy.
  • A file with a type we do not read is refused when you upload it, with a message.

How files count towards your plan is in the billing FAQ.

What to expect

We read every value and table we find, whatever the layout. You then choose which of them the flow keeps. See Flows. A value we are not sure about is marked in amber and says how sure we are.

A required value we are not sure about is flagged for review. See Checks.

If a file does not read

Open the document to see why. Some reasons can be fixed and tried again. Others, such as a damaged file, need a new file. See Troubleshooting.