cadabra | Cadabra computer algebra

I've got this code that translates a string (that starts out as a list) using a dictionary. I wanted the code to translate the string, then un-translate it back to the original.

This is the code that I've got so far:

I have a powershell build step in TeamCity:

*Sorry about my poor English. If there is anything that you don't understand, please tell me so that I can give you more information that 'make sence'.

**This is first time asking question in Stackoverflow. I've searched some rules for asking questions correctly here, but there should be something I missed. I welcome all feedback.

I'm currently solving algorithm problems to improve my skill, and I'm struggling with one question for three days. This question is from https://algospot.com/judge/problem/read/RESTORE , but since this page is in KOREAN, I tried to translate it in English.

Question
If there are 'k' pieces of partial strings given, calculate shortest string that includes all partial strings. All strings consist only lowercase alphabets. If there are more than 1 result strings that satisfy all conditions with same length, choose any string.

Input
In the first line of input, number of test case 'C'(C<=50) is given. For each test case, number of partial string 'k'(1<=k<=15) is given in the first line, and in next k lines partial strings are given. Length of partial string is between 1 to 40.

Output
For each testcase, print shortest string that includes all partial strings.

Sample Input
3
3
geo
oji
jing
2
world
hello
3
abrac
cadabra
dabr

Sample Output
geojing
helloworld
cadabrac

And here is my code. My code seems to work perfect with Sample Inputs, and when I made test inputs for my own and tested, everything worked fine. But when I submit this code, they say my code is 'wrong'.

Please tell me what is wrong with my code. You don't need to tell me whole fixed code, I just need sample inputs that causes error with my code. Added code description to make my code easier to understand.

Code Description

Saved all input partial strings in vector 'stringParts'.
Saved current shortest string result in global variable 'answer'.
Used 'cache' array for memoization - to skip repeated function call.

Algorithm I designed to solve this problem is divided into two function - restore() & eraseOverlapped().

restore() function calculates shortest string that includes all partial strings in 'stringParts'.
Result of resotre() is saved in 'answer'.

For restore(), there are three parameters - 'curString', 'selected' and 'last'.
'curString' stands for currently selected and overlapped string result.
'selected' stands for currently selected elements of 'stringParts'. Used bitmask to make my algorithm concise.
'last' stands for last selected element of 'stringParts' for making 'curString'.

eraseOverlapped() function does preprocessing - it deletes elements of 'stringParts' that can be completly included to other elements before executing restore().

I am trying to scrape the text from youtube live chat feeds using casper. I am having problems selecting the correct selector. There are many nested elements and dynamically generated elements for each new message that gets pushed out. How might one go about continually pulling the nested

some message

as they occur? I currently can't seem to grab just even one! Here's my test code: note: you can substitute any youtube url that has a live chat feed.