FMSLogo Code

Add test coverage for CRLF -> LF conversion when reading files without a BOM.

Add test coverage for reading UTF-8 files with variations on EOL sequences

Add test converage for more variations on UTF-8 encodings

Refactoring: Centralize the definition of the UTF8_BOM.

Fix a crashing bug in Test->Run Selection.
The code was not updated to handle Unicode correctly.

Add warnings for PCH problems.
This prints diagnostics that are useful for understanding PCH errors.

Test that the encoding of the loadstream is saved and restored by lload.

Add tests that LOAD honors the BOM within the file it loads

Always open files in binary mode and handle CRLF <-> LF conversion ourselves
This fixes some problems with READPOS and with reading UTF-16LE.

Merge in change from trunk, mostly to get new tests for LOAD