I am designing a survey with a large number of nearly identical Likert scale questions. Because it would be arduous to copy in every question, especially because the survey is still in the works and the format may change before final deployment, I wrote a script that turns a spreadsheet with the questions into a .txt file in the format outlined here so that I can import them into the web builder all at once. It seems like the generated file is syntactically valid, and it imports *almost* correctly (the question types, formatting, etc. are correct). However, the question text on all of the imported questions is set to “undefined” rather than the correct text from the imported file, so I would still have to go into each one and copy the question text from my spreadsheet to make the survey useful.

There are two things that I think might be responsible, assuming my syntax is correct. First, some of the question text is in Cyrillic (the survey is for a study of Russian language acquisition) — not something I can change. Second, there are hardcoded line breaks in the questions. I tried a version with no line breaks, but it caused a parse error when I tried to import it, so it seems like that might be worse.  

Here is an excerpt from the file, illustrating the format. (The dummy text here is of the same format and length as the real text.)

xtBlock:Questions]]

tiQuestion:MC:SingleAnswer:Horizontal]]
:HID:If1i]]
Эта строка является примером.

Эта строка тоже пример.

This is an example line.

� Choices]]
1 (very bad)
2
3
4
5
6
7 (excellent)

 /Question:MC:SingleAnswer:Horizontal]]
llID:If1u]]
Эта строка является примером.

Эта строка тоже пример.

This is an example line.

оChoices]]
1 (very bad)
2
3
4
5
6
7 (excellent)

I would appreciate any direct solutions or, if the .txt import simply won’t work for this use case, different ways to approach the task of importing a large quantity of questions of this type.