TestExcel

Notes: the spec is unclear on a few points that are required by the tests.

• fullCellText:
  • What is returned when you inspect a cell directly.
  • This is the original string that was used to create the cell, unevaluated.
  • For example, strings must have double quotes around them
  • If you set a cell to a number without a trailing ".0", fullCellText should return it without a trailing ".0"
  • fullCellText will also be used the spreadsheet is saved/serialized
• abbreviatedCellText.
  • What is returned when the cell is seen in the grid
  • That means it is evaluated
  • So strings do not have double quotes
  • Even if a cell has an integer value, it should be printed with ".0" (Though the extra credit tests seem to be inconsistent about this)
  • The string returned here must also be exactly 10 characters wide

I disagree about some of the design

• I like the fact that the Cell has a method for the unevaluated text and a method for the evaluated text, but I don't think it should worry about formatting, like making things 10 characters wide, that should be handled by the Spreadsheet class.
• The tests should allow double values to appear either with or without the ".0"
• Cells shouldn't be comparable to Object, they should just be comparable to other Cells. And it's a bit weird that all Text cells come before Real cells, why not just compare the output as text?
• All cells share the characteristic that they save the original text used to create the cell, and that's what is returned by fullCellText. So I created a common abstract base class for all cells (RootCell), even though that's not what the guide suggests.
• The solution also uses instanceof for serialization and deserialization, I prefer adding a cell type to Cells.
• The project avoids dealing with the nuances of floating point arithmetic, which is probably good.