Avoid duplicating web addon in enterprise by extracting a common basis.
Enterprise features stay in enterprise, but use that common basis.
Mainly:
- JS refactoring and linting
- Conversion of .sass into .less split into multiple files
- Templates cleaning and DOM simplification
- Re-generation of web.pot, and update of .po files
When providing an args of ``null`` (or ``undefined``) and a non-empty
kwargs, the kwargs would be removed/ignored.
While explicitly providing a null args is not necessary, it's perfectly
valid.
* use new support to colorize reconciled bank statement lines
* fixes tokenization of ``is not`` to not blow up on ``not foo``
* improves calling code to get truthiness of evaluated expression
instead of truthiness of javascript-equivalent objects: an empty list
is falsy in Python but truthy in javascript, since the evaluated
expression is in Python one would expect Python truthiness rules
without having to manually convert to boolean
closes#5070
* Accesses in contexts & domains are object derefs, so using dicts was
dumb
* But objects still need to round-trip through in case of e.g. o2m
commands in contexts, so py.object needs a toJSON (or a special
object kind needs to be added, specifically for round-tripping
objects through)
bzr revid: xmo@openerp.com-20121203122312-gc499mujf4l0nuz7
should probably go through a testing phase (checking its result
against that of eval_domain_and_context) for a while, but apart from
that it's supposed to work.
'stdlib' stuff (datetime.date.today, datetime.timedelta,
time.localtime, time.time, time.strftime) still need to be
written. Isn't there also a dateutil.relativedeleta somewhere?
Anyway that needs to be done and the various evaluation contexts need
to be defined, but the valuator itself should mostly work.
bzr revid: xmo@openerp.com-20120227073721-nkgeiqacbzch8xev