| Age | Commit message (Collapse) | Author |
|---|
|
Be sure that we have a beautiful error page and not a black/white error
Be sure that slug_matching with 308 redirect works as expected and don't
raise a RequestUID exception.
closes odoo/odoo#64889
Signed-off-by: Jérémy Kersten (jke) <jke@openerp.com>
|
|
Basic tests & tests of the 3 fixes of #64328
|
|
preparation for cleaning of v13
task-2065018
closes odoo/odoo#39051
X-original-commit: 8a0fc6476c70a4126043de3b3efbd1096e41f968
Signed-off-by: Jérémy Kersten (jke) <jke@openerp.com>
|
|
* = stock, test_website, web, website_forum, website_slides, base
Replace KarmaError with AccessError and remove the related override made
on crash_manager and ir_http.
task-2069890
closes odoo/odoo#36655
Signed-off-by: Jérémy Kersten (jke) <jke@openerp.com>
|
|
The `website_published` field from the website's mixins is basically a readonly
from `is_published` field.
On read, this field will simply read `is_published` and check if the record's
website_id is accessible (only for the multi mixin).
On write, it will always write on `is_published`.
This commit improves a few things:
- A lot of code was writting on website_published which was just then writting
on is_published. Writting directly on is_published makes more sense.
- Some backend fields would still reference `website_published` instead of
`is_published` which would just go through the related for no reason.
Plus, using `is_published` will make the field tooltip more accurate as we
are not in a website context ('Visible on current website' to 'Is Published')
- Filter and search on tree view were still using the `website_published`
related field, which is just a readonly when we are not in a frontend
context.
- Some create and write function would have security check on
`website_published` value but that was wrong as the user could bypass that by
simply writting on `is_published`. For the write method, check `is_published`
is more accurate as it will cover both case since `website_published` will
then call the write method on `is_published`
|
|
The purpose of these tests is to verify that errors informations
are displayed correctly on the website.
Part of https://github.com/odoo/odoo/pull/32132
task-1894820
|
|
Before this commit, when a view was removed during a module update (eg: the
record was deleted from the file), its COW views would not be deleted.
This could lead to unwanted behaviors including tracebacks(1).
This commit is somehow related/an extension of b5fe23055d that handle COW view
write during module update and 2e32cc5aa3 that remove COW views during module
uninstall.
task-1931683
Note: Some existing tests had to be run post install
(1) A view doing a t-call is COW'd, then that view and its t-called view are
removed. If the cow view is not removed, the t-call will crash (see tests
in this commit for detailed case).
closes odoo/odoo#31295
Signed-off-by: Jérémy Kersten (jke) <jke@openerp.com>
|
|
Since AST was introduced to compile Qweb templates, the reset templates
functionality is not working anymore.
Actually, it does not even appear anymore as the conditions to trigger and show
the reset view behavior can't be True anymore.
Indeed, the code is still importing the old QWebException from odoo.exceptions
which is now just an empty class.
The QWebException that should now be used is the one from odoo.base.models.qweb
More than just a wrong import, the QWebException properties have completely
changed.
Since its last working version (Odoo 9.0), a lot of changes occured, mostly:
- Odoo 11.0: new model website.page, which completely changed the pages
behavior and changed the models.
- Odoo 12.0: multi-website, which introduced the Copy On Write (COW) that
redefined how we write on view in a website context. We now need to reset
copied/specific views that have no model_data_id.
This fix is for a stable version. There is some flows that can't be fixed
without a refactoring or new fields (mainly because of the arch_fs being
erased when writing on `arch`, plus we can't introduce a module reload in
stable).
The goal of this commit is to fix the basic and standard case of broken views,
which is when non-technical users are breaking the views through HTML editor.
To reset the views, we will read the original arch in the XML file, as the
dev mode is doing.
To be eligible for the reset, the broken view need an `arch_fs` set. If the
view arch has been modified anywhere else than the frontend, the `arch_fs`
will be erased and the view wont be able to be reset.
As the broken views will be specific views, we will simply read the generic
view in the XML file and write it on the specific. We can't simply remove the
specific view as we might be on a specific tree.
Specific case, in case of custom views create dropping a snippet in an oe_struc
we just delete this inehrited view to allow end user to have their page back.
In some case of bad compilation in sub template, we try to guess which inherited
view, t-called view is broken by string matching based on 'last_path_node' xpath.
Add new module test_website to add tests, it will be usefull for futur test of
install/uninstall and ...
Todo in master:
- try to track on QWebException the real template that is broken and not only
the last node from the main template.
Co-authored-by: "Romain Derie <rde@odoo.com>"
Co-authored-by: "Jérémy Kersten <jke@openerp.com>"
closes odoo/odoo#29957
|
