Strange behavior of PDF form texts in Okular that I built

[Anıl Özbek]

Hi, I'm trying to do something with Okular to fill PDF forms more
natively. To do this, I started by adding a transparent background
instead of a white background.

I came across to a problem not directly related to the changes I made.
When I make a single change, the text in the form changes differently,
and the text doesn't respond to my inputs after that.

I applied the following document to make changes in Okular and install
it to my system:

https://community.kde.org/Get_Involved/development

I made some additions to the script to fix the problem but it didn't work:

export LANG=tr_TR.UTF-8
export LC_CTYPE="tr_TR.UTF-8"
export LC_NUMERIC="tr_TR.UTF-8"
export LC_TIME="tr_TR.UTF-8"
export LC_COLLATE="tr_TR.UTF-8"
export LC_MONETARY="tr_TR.UTF-8"
export LC_MESSAGES="tr_TR.UTF-8"
export LC_PAPER="tr_TR.UTF-8"
export LC_NAME="tr_TR.UTF-8"
export LC_ADDRESS="tr_TR.UTF-8"
export LC_TELEPHONE="tr_TR.UTF-8"
export LC_MEASUREMENT="tr_TR.UTF-8"
export LC_IDENTIFICATION="tr_TR.UTF-8"

I'm using Fedora 28 KDE spin with Turkish locale. Do you have any idea
how to solve the problem?

https://www.dropbox.com/s/2sxs0wtqw0v7vn7/okular1.png?dl=0 (form with
transparent background)
https://www.dropbox.com/s/8vlh7jb238iizty/okular2.png?dl=0 (problem)
https://www.dropbox.com/s/2joey8hca1wm4bw/foxit.png?dl=0 (the result I
would like to achieve in the long run)

-- 
Anıl Özbek