Handle two continuous START TEXT TAG by throwing Error

[rtgdk]

• HandBuiltParser is also not giving error when two START_TEXT are encountered one after the other without the END_TEXT.
• It ignores the second START_TEXT and take the END_TEXT for the first START_TEXT

Signed-off-by: rtgdk rohit.lodhartg@gmail.com

[rtgdk]

This PR will be for identifying the line no. of the error tag/value document.
The changes made will handle the case when two <text> tags are present simultaneously without any </text> in between. This will result in proper error line no.
Previously, it used to skip the second <text> tag and take </text> for the first <text> tag.

[rtgdk]

From this, we can get the correct line and character position where the START_TEXT is starting which can be passed in the error itself. I'll commit that code soon.

[goneall]

If we break the loop, we will get an exception unterminated text block. I think it may be better to just throw an exception with a message " found within a block" - this would make the error clearer

[rtgdk]

I have added a method to retrieve the line no. of the error by storing every line of the file in an ArrayList.
As an alternative, we could also pass the original file to HandBuiltParser for searching only when error occurs.
Also, I had to add the InputStream constructor for HandBuilParser as some tests were failing.( TestBuildDocument )

[goneall]

I fixed the unit test failure. The failure was due to an invalid type cast. Changing the type cast to instantiating a new NoCommentInputStream with the previous input stream as a constructor parameter fixed the problem. @rtgdk in the future, please run the unit tests before creating the pull requests

[rtgdk]

Thanks. Will keep that in mind. 👍