background and color snippet options
The background and color snippet options communicate so that if a
color is set while a background is already set, the backgrond is
removed. The problem is that the event propagation was not stop
so if a snippet had a background and one of its child set a color,
the background was removed.
Xpath has been altered and logos displayed in right column or portal.
The pupose of these logos is to know when you arrive on the website (ecommerce)
with which payment method you can paid.
category was changed to "Website" so it'd appear under the website
category in the store, but it needs to be "Theme" otherwise it doesn't
appear in the "Choose your theme" menu after installing website.
Since 5a9e1af64 our guess_mimetype accepted a `default` named argument.
This commit fix the use of library used instead of our implementation, so
we don't get an error from an unexpected argument when having python-magic
installed on a given server.
Since python-magic functions always return a default mimetype:
- empty content: application/x-empty
- text content: text/plain
- binary content: application/octet-stream
the guess_mimetype default argument is not used with python-magic
implementations.
closes#13343
opw-687354
When an owner is defined on a quant, the current behavior is to to allow
only the owner to use it. This is restrictive for internal transfers:
when a quant is owned by Partner A, only a picking with owner Partner A
can select it.
The fix is to make things a bit less restrictive: when a quant is owned
by Partner A, a picking without owner or with Partner A as owner can
select it.
opw-687207
2 rules were defined with the same id thus only one was loaded in
database, resulting in missing read access on model
'project.task.history.cumulative'.
This patch fixes the ids of these rules and clarify their names
In the new API, default calls are supposed to return recordset
(even empty recordset), not bool values for relationnal fields.
If the table for this model was already present and if no l10n_x
was installed, no default value could be computed correctly for
existing entries. This caused a crash on install.
opw-683470
When we were trying to delete a stage in project where there were
some tasks, an error was raised because of the 'project.task.history.cumulative'
records that inherit 'project.task.history' which have a compute method on
on end_date.
To fix this, we set the 'compute' attribute to 'None' and the readonly
'attribute' to 'True'.
If you create an issue when `default_project_id` is set, the stage_id
is initialized at the lowest sequence of the stages of the project,
which is normal. However, when you change the project of the task,
the stage_id was not correctly reset. If you save at this time, you
add the selected stage in the project.
Also, the stage_id was not correctly cleared if you remove the project.
If you create an issue directly, the key `default_project_id` is not set
in the context. This key is used to compute the default stage of the
created issue. It is done by creating a domain. Unfortunately, the
domain is always filled with `[('fold', '=', False)]` and thus it
always returns a result that makes no sense when `default_project_id`
is not set. Worse, if you actually create the issue, the found stage
that was not in your project is now a part of it.
The fix is to return no result if the `default_project_id` is not set
in the context.
If you create a task when `default_project_id` is set, the stage_id
is initialized at the lowest sequence of the stages of the project,
which is normal. However, when you change the project of the task,
the stage_id was not correctly reset. If you save at this time, you
add the selected stage in the project.
Also, the stage_id was not correctly cleared if you remove the project.
Make bill_to generic, and callable
Use Json synthax and not Dict synthax (even if it was compatible) fot t-field-option
Replace deprecated t-field-options by t-options
Before this patch, the order of a project's task type (its columns)
were defined by a sequence. However, when you quick-create columns,
they have the same sequence value: 1 by default. This results in
weird behavior when you edit a task type (for example, by changing its
tooltips), the column is moved at the end.
By ordering by sequence and id, the behavior is more intuitive
as the colum doesn't move when you edit it.
Now having
* sanitize: run the sanitizer to clean the html (removing javascripts,
unwanted tags, ...)
* sanitize_tags: only a subset of tags is allowed in html content.
Unwelcomed tags are remove dand their content stripped.
* sanitize_attributes: only a subset of attributes is allowed.
* sanitize_style: only a subset of style attributes is allowed. Style
attributes are parsed to keep only a white list.
* strip_style: all style is removed. It bypasses sanitize_style as there
is no need to sanitize something that is removed.
* strip_classes: remove class attributes
Fields parameters have also been updated to match the sanitize options. Html
fields by default are sanitized with sanitize_tags activated but without any
further options. All addons have been updated to match the new options
according to their previous behavior.
As we will soon improve the sanitizer we will be able to sanitize email
templates body. However this implies some cleaning in the templates to
be sure mako is not considered as invalid html / xml and therefore removed
from the template body.
If you create a task directly, the key `default_project_id` is not set
in the context. This key is used to compute the default stage of the
created task. It is done by creating a domain. Unfortunately, the
domain is always filled with `[('fold', '=', False)]` and thus it
always returns a result that makes no sense when `default_project_id`
is not set. Worse, if you actually create the task, the found stage
that was not in your project is now a part of it.
The fix is to return no result if the `default_project_id` is not set
in the context.
When the routing of the phantom bom is the same as the parent BoM,
the phantom BoM should not recreate the work orders, but use the same
work orders as the parent BoM. That way, you can create one master
routing and the "consumed in operation" of the phantom bom can be used
linking to the master routing.
If the phantom bom has no routing or the same, it can also use
the "consumed in operation" of his parent bom line if none is defined.
The web modules defines a "layout" template which is only used by the
login default page. Now the backend template "web.webclient_bootstrap"
and the frontend template "website.layout" both use the "web.layout"
template as base.
Some parameters can be passed to customize the default "web.layout"
template:
- "head": content will be placed at the end of the <head> tag of the
page
- "title": to set the title of the page (the <title> tag should not be
placed "by hand")
- "x_icon": the url for the page icon
- "html_data": dictionnary whose key-value pairs will be placed as
attribute-value for the <html> tag
- "body_classname": the classname to set on the <body> tag
Note: xpath expression on the website.layout template can now only find
the #wrapwrap element and its children. An expression "//body" will not
have any result. If something has to be added in the body element but
not in the #wrapwrap element, simply change the xpath
<xpath expr="//body" position="inside"> with
<xpath expr="//div[@id='wrapwrap']" position="after">
qsm-odoo