Normally I wouldn't go for this, as I like to give the user a lot of control, but I think we're doing way too much work to support every potential use-case.
To that end, I want to remove some configuration. Transformers, I think, are way too convoluted and try to do too many things.
So:
[ ] Remove the ability to use multiple transformers in sequence
[ ] The only one that's really useful with others is filter, so filtering should become an option to all transformers, where applicable.
[ ] Reconfigure "input" filter in diff command to compare everything, then rely on output filters (defined in previous item) to remove cruft by default
[ ] Remove input filters
[ ] Each command will have some configuration over how it works with each transform that it works with. I recommend a "default" transformer config--where appropriate (like fields; see #39) and overrides on a per-transform basis
[ ] Only a subset of transformer configuration should be configurable by the user. fields should not be; most of the table configuration should not be. Each configuration option will need to be evaluated on a case-by-case basis.
I envision looking more closely at this when I get to documenting configuration (#66). But if someone wants to take a crack at it, I can make myself available to help.
Normally I wouldn't go for this, as I like to give the user a lot of control, but I think we're doing way too much work to support every potential use-case.
To that end, I want to remove some configuration. Transformers, I think, are way too convoluted and try to do too many things.
So:
filter
, so filtering should become an option to all transformers, where applicable.diff
command to compare everything, then rely on output filters (defined in previous item) to remove cruft by defaultfields
; see #39) and overrides on a per-transform basisfields
should not be; most of thetable
configuration should not be. Each configuration option will need to be evaluated on a case-by-case basis.I envision looking more closely at this when I get to documenting configuration (#66). But if someone wants to take a crack at it, I can make myself available to help.