Update Rotalumis encoding to deal with utf-8 (and more)

[mypsycho]

Rotalumis only deals with 8859-1 charset. This is not suitable for non-european or non-Windows users. To avoid user misleading, Rotalumis should have a new parameter for encoding.

This encoding should be based on encoding declared on Eclipse resource.

[mypsycho]

From meeting 09/13: Issue with encoding are:

• Number parsing
  • Rotalumis process forces local parsing of number to "default english"
• Outputting numbers
  • Because Rotalumis relies on std::string* functions.
• XML exchange
  • because of in-memory String, they relies on 8859-1 (specifically for accentuation)
• Performance: String comparison and sorting are significally slower
  • Moreover as comparison implementation must stay efficient, we want to avoid multi-byte algorithm. Unicode Collation Algorithm

Changing parsing of file and parsing of number have a major impact on several part on of the design.

[mypsycho]

As short-term decision, we will just ensured that simulated model are in 8859-1 charset on start from IDE side. Use cases are

• When user create a Poosl Project, encoding of the project must be set to 8859-1
• If user has a specific projet (for mixed technologies), on creating poosl file, if project encoding is not 8859-1, force the file with 8859-1.
• When starting simulation, checks all model files have expected charset or pop-up a confirmation.