These data should be in no update. Else at next update, it can change the configuration
of the production.

This commit closes #13404
Courtesy of @bealdav

Since Chrome 53.0.2785, the input select is hover the existing tag.
With this fix, the existing tag are displayed at the right side of
the input to create a new one.

This commit closes #13655

Courtesy of @8cells

When a user registers for an event on the front-end, the confirmation of
the associated SO will also confirm the registration. However, if a SO
is created manually and a ticket event is added to it, the confirmation
of the SO will not trigger the confirmation of the registration.

In the second case, the registration simply doesn't exist yet at
confirmation and is created by the method `default_get` of
`registration.editor`.

We make sure to confirm these registration when we make the final
reservation.

opw-688665

In a multi-company environment, we need to make sure to copy the company
of the PO to the company of the picking. Otherwise, it might happen than
the PO as a different company than the picking.

Closes #13129

If a model uses nested sets to maintain a hierarchical record
ordering, then bulk inserts (e.g. via a CSV import) are likely to
suffer from an O(n^2) performance penalty, since there is a high
chance that each new record will require many of the existing records'
parent_left and parent_right fields to be updated.

The context parameter 'defer_parent_store_computation' may be passed
to the load() method, in order to defer updating the parent_left and
parent_right fields until after the bulk operation is complete.  This
appears to have last been fully supported in version 7 (via the old
import_data() method).

It is still possible to specify 'defer_parent_store_computation',
e.g. via an XML-RPC call such as

  m.execute_kw(dbname, uid, password, 'stock.location', 'load', [],
               {'fields': fields, 'data': data,
                'context': {'defer_parent_store_computation': True}})

This will currently defer the recalculation of parent_left and
parent_right, but does not correctly call _parent_store_compute()
after the bulk operation is complete.

Add the missing call to _parent_store_compute() to perform a one-time
update of the parent_left and parent_right fields if needed after the
import operation is complete.

Also fix the child_of_domain() and parent_of_domain() expression
methods, so that the 'child_of' and 'parent_of' operators give correct
answers when updates to parent_left and parent_right have been
deferred.

Signed-off-by: Michael Brown <mbrown@fensystems.co.uk>

Closes #13688 (part 2/2)

The expression context is currently passed to to_ids() but is missing
from all calls to child_of_domain() or parent_of_domain().  These
methods are effectively always passed a context of None.

Signed-off-by: Michael Brown <mbrown@fensystems.co.uk>

Closes #13688 (part 1/2)

Bulk inserts of product definitions (e.g. via a CSV import) will
suffer from an O(n^2) performance penalty due to an unfortunate cache
invalidation bug.  The underlying cause is difficult to determine due
to the labyrinthine nature of the relationship between product.product
and product.template, but some key elements are:

- The "product_variant_count" field is a non-stored computed field,
  and so gets invalidated by the call to recs.modified(upd_todo) in
  BaseModel.create()

- Field.determine_value() will call _prefetch_field() to repopulate
  "product_variant_count" by refetching all records from the database,
  despite the fact that "product_variant_count" is not stored in the
  database.

Using the length of "product_variant_ids" directly (instead of going
via "product_variant_count") seems to avoid this problem, possibly
because the definition of "product_variant_ids" conveys more
information about the circumstances in which its value must be
invalidated.

An alternative solution of making product_variant_count a stored field
(using the same trigger as for other variant-derived fields such as
volume and weight) works but still results in some unnecessary fetches
and invalidations.

Signed-off-by: Michael Brown <mbrown@fensystems.co.uk>

Closes #13689

[IMP] account: It is better to order also by 'id', esp. when 'sequence' could match among multiple records.

In saas-12 and up, upstream_dependencies returns:

- when called from old API : a list of ir.module.module ids
- when called from new API : a recordset of ir.module.module

From 9.0 to saas-11, when called from new API it would also returns a
list of ids which can bring needless inconsistencies.

With this commit when called from new API a recordset is returned.

opw-690015

Neither the super-user nor the sudo() mode should need
to use read_group() on any of the private user fields.
So we can simplify the code and remove this special case.

The same is not true for read(), though.

Also restrict XML data attribute evaluation context
even for real module data files. This will prevent
accidentally depending on context parameters that
would not be available inside base_import_module.

Move code related to redirection in its own class method. This way all
controllers perform their own actions and checks before calling the
redirection itself. Moreover it is necessary for some controllers to be
able to call the redirection independently outside of the mail/view
controller scope.

This commit also fixes some links generated for mail notification email
buttons. Assign and (un)follow buttons now effectively use the link helper
when generating the associated URL instead of manually crafting it.

Custom routes for notification buttons used in HR and CRM are added.
Indeed in order to be able to fine-tune the behavior called by those
buttons, using dedicated controllers is simpler than trying to use
a generic one. As there are few custom buttons in notification emails
using dedicated controllers does not lead to a code overhead.

Also restrict XML data attribute evaluation context
even for real module data files. This will prevent
accidentally depending on context parameters that
would not be available inside base_import_module.

When receiving new mail (not replies) to an alias we should not take
case into account.

This also homogenize the treatment of local parts.  For instance, lines
967 and 980 convert the local part to lower case to avoid
case-sensitivity issues.

Also `mail_alias` normalizes alias names by lowering case and finding,
if necessary to make it unique, a suffix to alias name provided.

From RFC 5321, section 2.4:

  > Exploiting the case sensitivity of mailbox local-parts impedes
  > interoperability and is discouraged

Closes #334
Closes #13037

The analytic account of the write-off move is not copied on the line
when doing a reconciliation.

opw-688635

On a given instance with 600 000 stock.history, the query changed by
this commit takes:

- before this change => 32 minutes
- after this change => 20 seconds

opw-685747

[FIX] sale: fix context update wrongly migrated / updated at 94716a3f

Purpose of the code is to automatically add the recipients of quotation emails
as followers. It is done using the mail_post_autofollow context key. However
current code add the context key in all cases and not just when composing
a message on sale order.

Following commit 6006c0d7, the shipping address can be changed after
the call to `checkout_form_save`. This is the case, for example, when
the user has the same shipping and invoicing address. We need to make
sure to recompute the fiscal position after the shipping address is
finally set.

opw-689281