Related and function manual fields may not be loadable during initial
setup (partial=True) due to dependencies being not loaded. Ignore
theses fields during partial loading.
When loading the registry, the parent store are defered at the end of
the registry initialisation. However, when modules should be removed, a
brand new registry is build without deleted modules, losing the list of
models on which parent store should be recomputed.
According to PEP302, the signature of `Finder.find_module` should be
`find_module(fullname, path=None)`.
Ever since it was introduced in 64ec5f36df the addons import hook
defines the second parameter as mandatory, which is an issue for
systems relying on the specified behaviour (and not needing to
provide a path) like the stdlib's `pkgutil.find_loader`.
fixes#10670
During database creation via the database manager,
or when using the startup option `--load-language`,
the selected language(s) will be installed as soon
as either:
- the base module is installed/updated (because
base_data.xml includes a call to res.lang.install_lang()
- the registry is loaded (after loading `base`,
the system installs the requested languages, even
if the server is not in update/install mode)
This is implemented by passing a global config
option `load_lang
This behavior was modified as of saas-7 by PR
for the command-line and for the database manager.
In both cases, we don't want the installation
to be repeated the next time either of these
event occur. Essentially the `load_language`
During database creation via the database manager,
or when using the startup option `--load-language`,
the selected language(s) will be installed as soon
as either:
- the base module is installed/updated (because
base_data.xml includes a call to res.lang.install_lang()
- the registry is loaded (after loading `base`,
the system installs the requested languages, even
if the server is not in update/install mode)
This is implemented by passing a global config
option: `load_lang`.
This behavior was modified as of saas-7 by PR
for the command-line and for the database manager.
In both cases, we don't want the installation
to be repeated the next time either of these
event occur. Essentially the `load_language`
config option is a single-use flag, so we
have to pop it from the config after using it.
A more complete refactoring is needed to
remove all uses or tools.config as a
global "context" for passing parameters.
This reverts commit ca7eeb0ed0, which
introduced a regression where, in some situations, an attempt to serve
an asset could be done before the routing map could be generated,
resulting in a 404 and a failure to load the web client.
The function invokes methods `_auto_init`, `init` and `_auto_end` on models,
and now uses the new API. This enables to define those methods in the new API.
The goal is to avoid recomputing field several times. Consider, for instance,
two fields F and G, such that G depends on F. Suppose that G is recomputed
before F. Saving G to database proceeds well, but when F is saved to database,
G is invalidated and marked for recomputation. Field G is possibly recomputed
twice on some records.
Avoid this situation by chosing a field such that none of its dependencies must
be recomputed; use a topological sort based on field dependencies for that
purpose. In the example above, G will never be recomputed before F.
Allows importing the core ``openerp`` as ``odoo`` (e.g. ``from
odoo.tools import cache``) and importing modules as
e.g. ``odoo.addons.purchase`` **from modules**. From the openerp core,
imports *must* remain to ``openerp``, as those run before import hooks
can be installed.
* AddonsHook allows Odoo imports to be imported as both odoo.addons.* or
openerp.addons.* and will load addons to both names simultanously
* OdooHook makes the Odoo core accessible as odoo as well as openerp
In test mode, only one cursor is available, and longpolling will take
it, and not give it back before 60s, causing a phantomjs timeout. This
commit simply return an error when the server is in test mode.
Sadly, I had to patch the web client to prevent logging the error in
this case, because that's the way phantomjs detect if there is a
problem.
When a new field has to be computed for the first time on existing records,
mark that field as todo, invalidate it in cache, and perform all recomputations
at once at the very end.
The `unittest2` package is simply a backport of `unittest` from the
Standard Library of Python 2.7 to previous versions.
There is no reason to use it any longer.
Closes#6941
When a model is set up, it can retrieve fields from a similar model (with the
same base classes). Only proper (not inherited), regular (non-related) fields
can be shared that way, since other fields may depend on other models.
Raphael Collet