How to complete pdf from website form data

Question

I've got data that customers are entering into a website form that I need to transfer to pdf form here http://www.revenue.ie/en/tax/it/forms/form11.pdf. Is there a way to automate this?

Answer 1

I have bad news, good news, bad news and finally some good news.

The bad news (part 1) is that the PDF document you refer to is not a form.

It looks like a form to humans, but it's not a form when a machine looks at it:

The above screen shot is a view inside the PDF using iText RUPS. RUPS takes a look at the root dictionary of the PDF (that's where a PDF viewer starts its search for objects in the file after reading the trailer and the cross-reference table).

To a machine, the rectangles and the labels preceding these rectangles are nothing but a bunch of arbitrary lines and characters drawn on a page.

The good news (part 1) is that you can add content to those pages.

You can do this the hard way by taking a ruler and measuring the position of every "field" and by adding content at those absolute positions. That's trivial nor efficient. Should the form change (for instance: the IRS moves things around in the next fiscal year), then you have to start coding all over again.

You can also do this in a better way: just turn the "flat" PDF into a "smart" PDF by making it an interactive form.

I've opened the document in Acrobat and I asked Acrobat to Create a form automatically:

You then let Acrobat do its magic (it can take a while) and this will result in a file that looks like this one: form11.pdf.

The result is rather bad because I've asked a machine to do the work:

As you can see, plenty of fields have names such as "undefined_x", Acrobat thinks that the page number is a field, check box fields are treated as text fields, fields that expect amounts are seen as different fields because of the , and . signs, and so on. Turning the flat PDF into a real form requires human intervention. I made one change manually (I'm not going to do your work in your place). There are two fields on page 2 asking you if you're single that, to the human eye, look like check boxes; one under 2. and one under 3.

I removed the text field under 2. and made it a check box named "manually_edited_single":

You can see that it is now indeed a check box with possible values Off and Yes. This is very different from what we see under 3:

This field is named "Single", and whoever is familiar with PDF clearly sees that the form field / widget annotation (*) shown in the screen shot above is not a check box.

(*) the field dictionary and annotation dictionary are joined in this case.

The bad news (part 2) is that "fixing" the form is a manual process.

It's going to take you an hour or so to join different fields that belong together, to replace text fields by check boxes (or radio fields) where necessary, to remove fields that aren't actually fields and add fields that were missed by Adobe Acrobat.

The good news (part 2) is that filling out the resulting form is a no-brainer.

That question has been answered many times before, see for instance: How to fill out a pdf file programatically?