Trees in a dataset should be named

[Jon Hill]

Each tree element should have a name, generated automatically, that is immutable. Currently, they are numbered and hence change when added to or deleted from.

This makes automatic processing (and the overlap) easier and more understandable.

To do:
 - write function to automatically name a tree (based on what is there when imported)
 - add the name to the schema - make it non-editable?
 - add info to the manual
 - check the tutorial

[Jon Hill (jon-hill)]

Other issues here:
 - existing tests check tree names, etc, but these are not in the data right now. Do we add the names to each Phyml in the dataset?
 - Do we hack a function in that checks if tree names are set and if not, set's them? This covers old data
 - _clean_data needs to add the names - we *should* load all existing datasets in, clean them (maybe not those designed to test clean_data...) and add the names. A user can do this.
 - Also, a name_all_trees function would be useful for a user to run without cleaning data. Jaime will find it useful when generating Phyml's from the database too

[Jon Hill (jon-hill)]

Current branch has:
 - updated manual
 - function on STK menu to name all unamed trees
 - standardise source name also does the tree name
 - all tests pass
 - new tests for naming trees