Oh, sorry, I forgot to explain why links on the actual texts is not enough.
There is quite a lot of white space in the cell, and when using the pdf or flipping pages solution (or whatever), we want the user to be able to click anywhere in the area that concerns a certain article.
Here is an illustration, with the clickable areas marked out. All white space would be "dead" and unclickable with no links.
Selecting the whole cell, makes the hyperlink panel grey, and you can not write a new address in it.
The way hyperlinks work is, assuming you're doing this for an interactive PDF, is defined by the PDF standard. What you want to do is not supported.
In my question I wrote that "Obviously hyperlinks from cells are not supported".
And I asked what people do to deal with this, that is "what's the best workaround".
I suggested using empty textframes on top of each cell. That would work, but those text frames could be in the way for editing, if the document is to be used after the interactive pdf is exported. We might do it this way, but have not decided. The other idea we have is to split the table into a fixed number of textframes, which would also work, but accidents could happen more easily with a lot of text frames (frames being moved, and not getting in the right position etc) than in a table.
I'm using scripting, so the work with assigning url:s will be done automatically, but I'd like more ideas than the ones I presented (which I'm not especially fond of).
This is outside my workflow, but would making a button in each cell work?
Maybe just adding a layer with the buttons would be the simplest thing to do.. Just hide the layer when you need to edit the text...
Best suggestion yet is yours, about adding new layer!
Empty textframes with the hyperlinks on a separate layer, is the best combination of workarounds yet. I know nothing about buttons (what they could do better than the hyperlink on a textframe).
Thank's for all creative suggestions! (If there are more of them, feel free to add them as well!)
@Andreas – to choose buttons over text based hyperlinks would mean you can define the area that is responsive like you want (the size is the size of the button) and you could design different button states in different ways (color, images, what not…).
For the scripter in you: Hyperlink objects have no geometric bounds, Button objects have…
Just an idea:
Imagine a script that is scanning all cells of a selected table and if there are hyperlinks detected uses position, width and height of that cell to generate a button just on top of the cell (it could be well on a different layer above) using the information the hyperlink provides. Every time the table is changed by the user and the script is fired again, the old buttons are removed and new buttons will be built…
If there are more hyperlinks than one in a specific cell, the buttons will be built around the area the hyperlinked text will occupy…