Imported text (.pdf)

[Curious7]

Hi,

again I thank you for your hard job in doing A. P to a great application.

My question is the following:

When I import a pdf, let's say a 16 page brochure with text and pictures into the Affinity Publisher, it opens very nice.All elements are there.

However, it shows every text line as a separate text frame and I want the text imported in one textframe, as it is in the pdf.
Is there a way of getting all the imported text in one frame, at least paragraf wise or even better one frame per page?

 

Best wishes 

 

[Alfred]

When you open a PDF file in any of the Affinity desktop apps you’re presented with the ‘PDF Options’ dialog. Among other things, you can choose to have separate lines of text treated as a single text frame to aid text flow.

[thomaso]

11 minutes ago, Alfred said:

When you open a PDF file

Great feature, hadn't even thought about to import a PDF in AfPub in this way.

Now I wonder why they do not allow this option window to appear when placing a PDF file, too? (Or have I just not discovered such a preset yet?)

 

btw: little, apparently coincidental issues with wrapped text:

[garrettm30]

It may be helpful to understand that "Group lines of text into text frames" can only estimate which lines should go together, as the PDF file does not include that information. It looks like it works for basic blocks of text, but it fails on more complicated issues. In cases like your screenshot, I am not sure how it could possibly tell the difference between two separate columns with an image between versus a single column with an image in the middle.

Importing a PDF is pretty amazing as it is, but we can only expect so much from it, as it is in fact a conversion process. Think of PDF as a digital printout, and to import it, Publisher must try to reconstruct it as an editable document.

[thomaso]

Yes of course. But is not it interesting that AfPub recognizes / interprets blocks of text in the green areas of my screenshot, whereas it does not work in the red areas - even though they are wrapping around same curves.

[garrettm30]

Interesting for sure. But some things that a human can easily recognize are a huge challenge to get a computer to recognize. Surely it could be improved, but at what cost of development time? And at best, is will still fail at times.

The most curious aspect to me is that it kept individual lines on the far left beginning with "sapien. augue" rather than treating that as a single column.

[thomaso]

These single lines didn't make me wonder that much: since it has to work horizontally different than vertically it appears to recognize a similar text to the right of the "sapien. augue" lines. To avoid making two wrong vertically blocks (one left, one right of the circle shape) it prefers horizontally and keeps single lines instead.

This also might explain the red areas: they, too, have similar text in their close horizontal surrounding. Which the green areas don't have.