This is a performance patch.
Writing on the field `users` of `res.groups` had the side effect
to recompute the field `share` of `res.users`.
This is because the trigger of this field is
`@api.depends('groups_id')`.
The more users you had, the more time it took to
add or remove the second company of a user.
This also slowed down the creation of employee users
when the multi-company group was added
to the inherited groups of the employee group
(basically when checking "Multi-Company" in the General Settings)
as then all new employee matched the condition
`if len(user.company_ids) <= 1 and user.id in group_multi_company.users.ids:`
(Multi-Company group checked, but only one company set, by default).
closesodoo/odoo#33305
Signed-off-by: Denis Ledoux <beledouxdenis@users.noreply.github.com>
Before this patch, the ir.logging's write_uid field was a many2one which
could cause a module install/update to hang when the module is changing
the res.users model schema and when this module causes to orm to warn
through the logger. (eg: declaring two res.users fields with the same
string attribute)
In such situation the transaction cursor that is processing the
res_users table alteration will be granted an exclusive postgresql lock
hence causing the ir_logging insertion to block because of the write_uid
foreign key to res_users.
This issue has never been raised by runbot as it is using a remote
database with --log-db
Note: the write_uid conversion from m2o to int was left over in commit e6a5d82closesodoo/odoo#32015
Signed-off-by: Christophe Simonis <chs@odoo.com>
We set the default rate to be 1.0 instead of 0.0, since a rate of 0.0
doesn't make sense.
Partial backport of 298491597c
opw-1949866
closesodoo/odoo#31866
Signed-off-by: Nicolas Martinelli (nim) <nim@odoo.com>
This is related to issue #31549, fixing it in 10 since it's already a
minor issue here (importing 500 partners which all have the same parent,
on my system the import time goes from 2:30 to 2:00).
This is mostly a problem for such fields as `property_product_pricelist`
(added do the commercial fields list by `product`): since we're first
getting it then writing a field it depends on (updating the address,
which contains the country), the field gets invalidated for all records
on each record being imported, so if we're importing a bunch of partners
which all have the same parent (e.g. company employees)
`property_product_pricelist` is going to be re-computed for every
partner imported so far as well as the one parent we're interested in,
for every new partner we're creating.
The problem is much more prevalent with 12.0's batched creates as we're
first creating all the new partners then doing the updates, and thus for
each new partner we're computing `property_product_pricelist` for all
new partners and the one parent we're interested in, roughly doubling
the import time of a series of partners which all have the same parent.
closesodoo/odoo#31804
Signed-off-by: Raphael Collet (rco) <rco@openerp.com>
Werkzeug version was being checked to avoid passing quote=True to
werkzeug.utils.escape (as that parameter was changed to `True` *and
deprecated* in 0.9).
However because DeprecationWarning was made silent by default in
Python 3.2 and the way the check is implemented worked for 0.9 it
looks like nobody really noticed it's broken in the usual manner of
half-assed version checks: works for 0.9.0, doesn't work for
0.12.3 (because lexically 0.12.3 < 0.9.0).
Fix by using proper version parsing and comparing the result of that.
See also: odoo/odoo#28116closesodoo/odoo#31553
Signed-off-by: "Xavier Morel (xmo)" <xmo@openerp.com>
Before this commit, doing expression.OR() with only FALSE_LEAF would
yield [] which is equivalent to TRUE_LEAF and is therefore not correct.
The same happened (to a lesser extent) with expression.AND() within an
expression.OR(), since the former would return a [] which would be
ignored by expression.OR().
See tests for a clearer view of the use cases.
Fixes#30113, #26540closesodoo/odoo#31202
Before this, groups on non-stored inverse fields were not checked upon write.
The impact on existing fields is pretty small, since the inverse methods of
those fields are subject to access rights on the records they use.
closesodoo/odoo#30356
This is the first step to a more comprehensive handling of company-dependent
fields which are ir_properties.
With model-specific access rights, users should be able to read/update a
company-dependent field no matter their access rights on ir_property.
Before this commit, a user having access to res.partner, but not to ir.property
couldn't write on property_account_receivable/payable just because he couldn't
write the corresponding ir.property. After this commit, he can.
OPW 1923345
Previous to this commit, if one were to create an ir.model.field with a
poorly constructed domain (read: SyntaxError), the server would properly
send an error message stating that an Error occurred, however this would
be too late as the registry with the bad code would have already been
reloaded, this meant that the registry would be left in an unstable
state (read: crashed).
With this commit, a constraint on the domain is added so that we confirm
that the code in the domain field is properly constructed, thus no need
to reload the registry and therefore no crash.
closesodoo/odoo#30157
Some views have the primary mode while having an inherit_id view
In such views, if the user wants to change the inherit_id,
the mode must remain primary.
The use case behind this is a user who want to change
the inherit_id view of the view
product.product.form (product.product_normal_form_view)
to another view.
This view is in primary mode while having an inherit_id
(product.product_template_form_view).
In such a case, the primary mode must remain,
otherwise the view will no longer be opened by default
when opening a product.
Indeed, when searching the default form view of a model,
it only searches for primary mode views for this model.
The `all` is there because this is an `api.multi` method.
We could consider adding a `self.ensure_one`,
but it breaks the API if someones calls this method with multiple records.
This is likely to happen, as this method is used
in `write` which is likely to be called with multiple records.
In my opinion, in master, this should be replaced by an
onchange so the user can see the change of mode
when he adds or remove an inherit_id for a view.
opw-1916324
closesodoo/odoo#30241
The method `read()` may be very slow when reading relational fields and
computed fields, because the computed fields can be computed on a recordset
that is larger than expected.
The issue occurs on model 'res.partner' when reading fields 'child_ids' and
'purchase_order_count', for instance. Suppose we read those two fields on a
partner with 1000 contacts. First, the one2many field is read from the
database and stored to the cache; the latter adds the value ids to the
prefetching of 'res.partner'. Then, the fields are fetched from the cache.
When 'purchase_order_count' is accessed on the partner, the field is computed
on all its children as well...
closesodoo/odoo#29867
Currently name_search on res_partner holds code to be done by a customized SQL
query allowing to speedup the search [1]. However when dealing with given args
there may be a crash if there is a domain based on user_ids fields.
Indeed this field is defined as auto_join [2]. It means the result of get_sql
returns where parameters based on joined tables. Those tables are not available
in the from clause in the original query.
This commit fixes that issue by correctly taking the from_clause from get_sql.
That way joined tables are available. Small update of the query is necessary
as fields are now res_partner.{id/email/vat/reference} as there may be several
tables linked in the query.
A test has been added to avoid regression.
[1] See https://github.com/odoo/odoo/commit/05ec12692f99b69924765fd0ce384632547f03f9 for a recent modification
and implementation of this query, even if it exists since a looooong time
[2] See https://github.com/odoo/odoo/commit/db8203c27a21acdbcad2cf1c394b6fea3cf13688 for the auto_join addition
closesodoo/odoo#29827
Modifying a source term in an XML/HTML translated field can lose translations
if the same term is translated in several languages.
closesodoo/odoo#29078
The other methods do test password length before verifying it, so even
if check_credentials() is not meant to be called directly, it's better
to keep it consistent with the alternatives.
Closes#29023
Seems like something similar to odoo/odoo#17111 can happen on Python 2.
The same change as #17111 is a bit involved for -stable, but this looks
pretty harmless.
Probably fixes#23781.
When two transactions conflicts (eg. deleting same data, see [1]) Odoo
will retry the whole transaction several times hoping for the best.
But when rendering template, the error handling would prevent this
feature. For example a recurring issue was:
- loading quickly two times the /web route
- for each recompute the assets
=> this could lead to 2 concurrents transactions that would delete
previous same attachment (ie. DELETE FROM ir_attachment where id=3).
With this changeset, when getting a template fails because of a
transaction rollback, we let the issue bubble up so our retry system is
used.
So instead of a "500 Internal Server Error" page and in log:
bad query: b'DELETE FROM ir_attachment WHERE id IN (3)'
ERROR: could not serialize access due to concurrent update
"GET /web HTTP/1.1" 200 -
"GET /web HTTP/1.1" 500 -
... big traceback ...
load could not load template
we would get the requested page without error and:
bad query: b'DELETE FROM ir_attachment WHERE id IN (3)'
ERROR: could not serialize access due to concurrent update
"GET /web HTTP/1.1" 200 -
SERIALIZATION_FAILURE, retry 1/5 in 0.8920 sec..
"GET /web HTTP/1.1" 200 -
As a side node, the issue was exacerbated in some instances:
- when running a database on another server: assets are recomputed
- when a module was installed/uninstalled: assets may are recomputed
- when updating the source code: assets may be recomputed
- when starting server: requests could be stacked waiting for readiness
- when using google chrome: the "Use a prediction service to load pages
more quickly" option may load a page two times. for example:
-> an URL is entered in address bar
-> a prediction request to it is started URL
-> go to this page (Enter) when that request is not already resolved
-> the prediction request is cancelled and a new request is started
[1] https://www.postgresql.org/docs/9.6/static/transaction-iso.html#XACT-REPEATABLE-READ
note: 10.0 backport of 11.0 #26778
opw-1849167
closes#26785
The override of fields_get adds "virtual" fields corresponding to
groups.
If for example we make "res.users" have inherit "mail.thread", we get
these "virtual" field as if they had "track_visibility", since we do a
"fields_get" with fields having track_visibility expecting to only get
back "track_visibility" ones.
So the system would then fail trying to track visibility on fields like
"in_group_5" and for example a res.users could not be created anymore.
With this fix, we fix the fields_get so it respect the fields we ask of
it.
fixes#22332
opw-1878654
closes#22338closes#26705
Co-authored-by: Wolfgang Pichler <wpichler@callino.at>
On a new DB with demo data, as Demo User:
- Create a partner
- Unlink the partner
An `AccessError` is raised.
The error is raised on `ir.property`, for the field
`property_product_pricelist`. An entry was indeed automatically created
at partner creation. However, a regular user is not allowed to delete an
`ir.property`.
The property can be safely deleted as SUPERUSER as long as the proper
unlink access rights have been tested beforehand.
opw-1870737
Denis Ledoux