It felt weird that adding hidden fields required changing existing
flags unrelated to the field you actually want to affect, and the
upper/lower flag thing seems to work well for -s/-S sooo...
- Replaced -l/--label with -B/--hidden-by for by fields that can
be hidden from the table renderer.
- Added -F/--hidden-field as a similar thing for field fields.
- Better integrated -i/--enumerate into by fields, now these actually
maintain related order. And of course added a matching
-I/--hidden-enumerate flag.
The only downside is this is eating a lot of flag names.. But one of the
nice thing about limiting this complexity to csv.py is it avoids these
flag names cluttering up the other result scripts.
---
The -F/--hidden-fields flag I'm not so sure about, since field exprs
can't really reference each other (single pass). But it does provide
symmetry with -B/--hidden-by, and reserves the name in case hidden field
fields are more useful in the future.
Unfortunately it _is_ annoyingly inconsistent with other hidden fields
(-S/--sort, -D/--define, etc) in that it does end up in output csvs...
But this script is already feeling way over-engineered as is.
Christopher Haster
5811b11131