Less aggressive name validation

[neworderofjamie]

I realised as my microcircuit model failed to build that #443 was a little over-zealous - population names are fine to start with a number as, when they're used in generated code, they are always prefixed by a variable name (which can't start with a number)

[tnowotny]

Good point - even though I would never encourage people to start a name with a number.
Also good idea to test against use of dashes ... some people might well have that idea.

[codecov]

Codecov Report

Merging #448 (4a91262) into master (f3ebe2a) will increase coverage by 0.00%.
The diff coverage is 95.00%.

@@           Coverage Diff           @@
##           master     #448   +/-   ##
=======================================
  Coverage   88.00%   88.00%           
=======================================
  Files          78       78           
  Lines       16593    16605   +12     
=======================================
+ Hits        14602    14614   +12     
  Misses       1991     1991           

Impacted Files	Coverage Δ
include/genn/genn/currentSource.h	58.33% <0.00%> (ø)
include/genn/genn/customUpdate.h	93.87% <100.00%> (ø)
include/genn/genn/gennUtils.h	100.00% <100.00%> (ø)
include/genn/genn/neuronGroup.h	74.13% <100.00%> (ø)
src/genn/genn/gennUtils.cc	98.71% <100.00%> (+0.23%)	⬆️
src/genn/genn/synapseGroup.cc	81.38% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update f3ebe2a...4a91262. Read the comment docs.