Finish to clean the API not using the right encoding

[jecisc]

I fixed in #20 most of the API that was not managing encoding. To do that I added method to use binary streams, byte array and encoding instead of a string.

But there are still a few methods that should be removed and replaced by something managing the encodings. Sadly they are still used and I don't have the time to adapt them. Here is the list:

• String>>#tsLinesDo:
• String>>#positionFromTSPoint:
• String>>#tsLineIndicesDo:
• String>>#tsLines

[alesshosry]

Thank you for your contribution

Methods #tsLines, #tsLinesDo: and #tsLineIndicesDo: can be removed because they are used in #positionFromTSPoint: to do the calculations.
My suggestion would be:

• remove #positionFromTSPoint: from String; Instead I will be adding 2 methods in TSPoint:
  1. #getMyPositionInText: aString : by default the encoding will be inside the method in #utf8.
  2. #getMyPositionInText: aString usingEncoding: anEncoding : Encoding set by the user.
• I think I found an algorithm more simple to calculate the position without using #tsLines, #tsLinesDo: and #tsLineIndicesDo: .

Please @jecisc if you can confirm if this is better for you so I proceed.

[jecisc]

I see two ways to fix the users of this method:

• Use the version that do not depend on TSPoint but uses the position in bytes via the methods I added (the better solution when possible since it is the fastest)
• Reimplement #positionFromTSPoint as you suggested. To compute the position you will need the encoding. You can wrap a binary stream in a Character stream with the right encoding and use #nextLine:. Then manipulate the wrapped binary stream to read the right number of bytes.

I did not do it because I don't need this (I just opened the issue because the current implementation is wrong).
Since I don't need it, go ahead :)