Before when we do `reserved(records)`, it iterates in a reverse
order of `records`. Unfortunately, the `__reserved__` method don't
exist explicitly in BaseModel then Python fallback on
its own implementation using `__getitem__` and `__len__` (coming from Sequence): https://github.com/python/cpython/blob/3.10/Lib/_collections_abc.py#L1047-L1049
Because it uses __getitem__, it breaks the prefetch of the recordset.
Example:
-------------
```
partners = self.env['res.partner'].browse(1, 2, 3, 4, 5)
for partner in reversed(partners):
partner.name
```
will generate 5 SQL requests to fetch data (one by record)
Then create our own `__reversed__` and handle the prefetch correctly
(like `__iter__`). Now in the example it will correctly generate only
1 SQL request because of the prefetch.
task-2687953
closesodoo/odoo#79622
Signed-off-by: Raphael Collet <rco@odoo.com>
Co-authored-by: rco-odoo <rco@odoo.com>
The creation of recordset was done by class method _browse() instead of
a regular call to the model's class. As we removed the old usage of
method __init__(), we can now reuse it with a normal usage. After this
commit, we can create a recordset by simply calling its class:
registry['model_name'](env, ids, prefetch_ids)
closesodoo/odoo#79563
Signed-off-by: Raphael Collet <rco@odoo.com>
- Remove unused imports `AsIs` and `Collector`
- Remove unused logger _schema
- Remove unused function same_name
- Remove unused attribute `_needaction`
- Remove backward compatibility of `__new__` and `__init__`
- Remove backward compatibility of `__export_rows`
Also:
- remove deprecated warning in api.py
- remove the `__init__` from `ir.ui.menu`, it was useless
because the cleaning of cache (`clear_caches`) is global (a call of
`clear_caches` clean all cache method with a ormcache decoration)
Part-of: odoo/odoo#79563
The method BaseModel.update does not batch records for no reason. This
method is used in `_onchange_eval` and in few onchange in Odoo.
In _modified_triggers() avoid a useless record union.
Part-of: odoo/odoo#79563
The method refresh() is a duplicate of method invalidate_cache() and is
deprecated since version 8.0, but without any warning. Add this warning
to be able to completely remove it in the next major version.
Part-of: odoo/odoo#79563
This fixes inconsistencies when dealing with fields that are computed
and inversed by the same methods.
Consider two fields F1, F2 with the same compute and inverse methods.
Consider a record where we wrote on a dependency of the common compute
method. At this point, both fields F1 and F2 are marked to be computed.
Now let us write on F1 only. Here is what happens:
- the write discards the computation of F1, but not F2
- the inverse method of F1 is called:
- the method accesses F2
-> this calls the compute method, which assigns both F1 and F2
- the method accesses F1
-> the value of F1 has been replaced by the computation above
The issue comes from a combination of factors:
- the value of F2 must be determined by the computation;
- the computation assigns both F1 and F2;
- the computation is done while inversing F1 (and F2).
The solution is to force the computation before actually writing on the
fields and calling their inverse methods. Note that this is necessary
only when part of the fields computed by a common method are updated.
When all fields computed by a common method are updated, the computation
will automatically be cancelled.
closesodoo/odoo#81172
X-original-commit: fb4d6ee4ba8bcb5cd8030105ac57e9e02850bfc4
Signed-off-by: Raphael Collet <rco@odoo.com>
* add configuration for `flake8[flake8-rst-docstring]`
* enable docstring-related checks
* fix invalid docstrings in odoo's core & `base`
* fix a few more bits (mostly missing or incorrect `:param:` info
fields) are out of scope for the lint but my editor catches
closesodoo/odoo#74604
Signed-off-by: Xavier Morel (xmo) <xmo@odoo.com>
Until now, stored compute fields were computed after database insertion.
This meant that required fields should not be computed, for instance,
unless some hackish code was added to make it work. Another trick was
to provide some default value, but this actually prevents the field for
being computed after insertion.
This commit provides a field parameter to specify that the field should
be precomputed: adding precompute=True on the field definition force the
method create() to compute its value before inserting the new record in
the database. For the reason explained below, precomputing fields is
not always correct, and therefore the default remains to not precompute
a field.
Some stored fields must be computed after insertion, for instance:
* statistics fields computed with search/read_group/...
* fields referencing the current record (res.partner.commercial_partner_id)
* fields referencing another record that does not exist yet (think about
records created by one2many fields)
* fields depending on the create_date/write_date/create_uid/write_uid
Those fields shouldn't be defined with precompute=True, which triggers
their computation post record creation. This is why, by safety, we
consider the default behavior to be precompute=False.
closesodoo/odoo#80449
Signed-off-by: Raphael Collet (rco) <rco@openerp.com>
Co-authored-by: Raphael Collet <rco@odoo.com>
Co-authored-by: Yannick Tivisse <yti@odoo.com>
- The batch version of get_metadata won't work correctly for the xmlid
returns due to the limit=1 (cause by
718b0f8b63 ). Fix the
previous issue by reversing the order and without limit.
- The read should be read with the correct user.
closesodoo/odoo#79091
X-original-commit: 29a53f19a7729f486cbc6ff69893b8d150c3512d
Signed-off-by: Rémy Voet <ryv@odoo.com>
With a lot of module installed setup_models can be quite slow.
One of the reason for that are the call to resolve_mro.
This operation is fast, but called for a lot of models/fields this
become an important part of get_depends and setup_base.
- This is call once for each models in setup_base
- This is call more than once for each computed_field in get_depends.
The use of a cache shows significative performance improvements.
During the install of a database with all modules, the install times
is reduced by ~3 minutes over ~14
Part-of: odoo/odoo#78898
Co-authored-by: Raphael Collet <rco@odoo.com>
The exists method (of BaseModel) doesn't work with model
without SQL table but with a table query (see
f2ceef0e2f)
This method is call when we try to read a forbidden
record (`forbidden = missing.exists()` in `_read`)
- Fix it by using a Query object (which able the case).
- Also use a partition method instead of duplicate it in the method.
- Fix the CRM Lead Report which can return a Falsy id
task-2633558
closesodoo/odoo#77644
X-original-commit: 70a68615c722e7145b2ba8bb7ca86a8373b5aa4e
Signed-off-by: Rémy Voet <ryv-odoo@users.noreply.github.com>
Signed-off-by: Raphael Collet (rco) <rco@openerp.com>
The query count increases by one because the old code (with
hand-rolled savepoints) never released the `model_load` savepoint.
Part-of: odoo/odoo#76243
Because the field is stored it doesn't make much sense to have a
context dependency and that can cause significant issues. However
`name_get` looks pretty innocent and it's not necessarily insane to
have a context dependency *there* (or even in the average
`display_name` as they're normally non-stored computed fields).
`_compute_display_name` previously blanked (blacklisted) just a few
context values, but doing the opposite is probably the better idea.
See also: odoo/odoo#68946
OPW-2453956
Task 2500978
closesodoo/odoo#76469
Signed-off-by: Raphael Collet (rco) <rco@openerp.com>
Co-authored-by: Raphael Collet <rco@odoo.com>
For searches, Odoo uses `unaccent` if it's available. On some
technical fields this is completely unnecessary and precludes the use
of indexes.
This PR provides:
* an opt-out (`unaccent = False`) on `String` and `Text` fields
* a warning if `unaccent` is enabled on *parent_path* fields as their
performance can be rather critical and not using the index is quite
an issue (note: the check that `parent_path` fields have been moved
outside of the check for their existence as we want to check that
the field is declared and correctly configured in all cases,
probably)
Task 2627454
Part-of: odoo/odoo#76436
Assume that model A has a computed field G that depends on both fields
F1 and F2. Assume that model B inherits from A (_inherits). When G is
computed during an onchange, it must use the form values for all its
dependencies F1 and F2. Before this commit, only the field triggering
the onchange was assigned on the parent record (see (*) below.)
record parent
--- ------------+-------------+------------
initial state | F1=0, F2=0 | F1=0, F2=0
assign F1=1 | F1=1, F2=0 | F1=1, F2=0
assign F2=2 | F1=1, F2=1 | F1=0, F2=1 (*)
The commit also includes another patch: when assigning the parent
record, the ORM determines the dependent records (in this case, the main
record in the onchange). If the delegate field (many2one from B to A)
has an inverse one2many field, the ORM uses that field to determine
dependent records. However, in the case of a new record, that field is
not in cache, and its value is determined to be... empty! The patch
consists in "fixing" the value of that field when we determine the
parent record (when the delegate field is accessed.)
Part-of: odoo/odoo#76042
Co-authored-by: William André <wan@odoo.com>
Before this commit, calling read_group on a model and grouping by a
selection field would return the groups in alphabetical order, meaning
that if your selection field's options were declared as:
('z', 'Z'), ('a', 'A')
read_group would return first group 'a' and then group 'z', instead of
groups 'z' and then 'a' which some might expect.
With this commit, it is now possible to set a fields.Selection
group_expand attribute to `True` when declaring it, which will use a
default group_expand implementation specific to Selection fields, this
means that when grouping by a Selection field with `group_expand=True`
it will always return the groups in the definition order of the
selection options.
We achieve this by leveraging the group_expand field attribute which was
designed for changing the groups returned by read_group.
Since this attribute was thought only to be implemented on a
Model-by-Model basis and here we need to use it as a generic function
for all Selection fields (explicit group_expand declarations have higher
precedence), the generic method has been implemented inside
fields.Selection and takes an extra records parameter which holds a
reference to the recordset/model on which read_group was called. This
extra parameter only applies to field implementations of group_expand
and in this case is what allows this specific feature to work with
dynamic Selection fields (function as options).
A side-effect of this implementation is that a read_group call that
groups by a Selection field with `group_expand=True` that uses the
default group_expand will always return all possible groups, even
empty ones.
Task-ID 2635052
closesodoo/odoo#75856
Signed-off-by: Raphael Collet (rco) <rco@openerp.com>
PURPOSE
Improve import wizard as followup of recent improvements.
SPECIFICATIONS
Preview lines
- make the preview lines more visible
- set font-weight: 500 on o_import_header_name
- remove the text-muted class from the data preview line
- allow the user to preview the first 5 non-empty values
- when hovering on the data preview span, display a tooltip with the first
5 non-empty values of the file column (with 'Preview' as the tooltip title)
- In the 'when a value cannot be matched' dropdown:
add a new 'skip record' option for the following field types: boolean,
many2one, many2many, selection
add a new 'set empty' for the following field types: many2one, many2many,
selection hide this option if the field is required
When a value cannot be matched:
the 'skip record' option will make sure that lines with an unmatched value
will not be imported (i.e. skipped at import)
the 'set empty' option will set the field value to False when the value
cannot be matched
- Restore colored background on alert boxes by removing some css rules
Task-2504343
closesodoo/odoo#73713
Signed-off-by: Thibault Delavallee (tde) <tde@openerp.com>
Co-authored-by: Mohammed Shekha <msh@odoo.com>
* = crm_livechat, hr, hr_holidays, im_livechat, mail_bot, purchase, sms,
snailmail, survey, test_discuss_full, test_mail, web_editor, website,
website_livechat
- Create new model `mail.guest` for guests.
- Rewrite some RPCs to target routes rather than model methods so that
guests are able to use them.
- Patch JS and python models to support guests.
- Create a stand-alone page and boot the channel in it.
task-2494829
closesodoo/odoo#75496
Related: odoo/enterprise#20417
Signed-off-by: Sébastien Theys (seb) <seb@odoo.com>
Grouping by many2many has been added with PR odoo/odoo#68958 (commit f4f570c3d6b403512d8523c74acd34d485f46af1)
However the display names are not resolved resulting in the grouping
displaying the actual id of the records.
This also breaks some features such as `read_groups` which expect the
group by data to contain a tuple (id, x) instead of just the id, this
commit adds name resolving on many2many group bys.
Task ID: 2398734
Part-of: odoo/odoo#74985
Since now, there was no possibilities to pay a Sale Order in a point of
sale without re-encoding every lines of the sale order in a pos order.
So we've now added the possibility to choose a sale order from the point
of sale, and apply a down payment or settle the selected order. You
still get the possibility to invoice it, and the stock is correctly
managed.
The SO is then updated accordingly to what have been done throught the
POS.
closesodoo/odoo#74096
Task-id: 2500902
Signed-off-by: pimodoo <pimodoo@users.noreply.github.com>
The fields dict maps field names to their description. Before this
commit, the field name wasn't specified in the description, meaning
that we could never manipulate the field description without its
key, which would often be convenient.
closesodoo/odoo#75121
Signed-off-by: Raphael Collet (rco) <rco@openerp.com>
The method is used to get progress per column in kanban view
(green-yellow-red-red lines in Project, CRM etc). There are two main
usages:
1. get statistics for ``kanban_state`` (red/green circles)
2. get statistics for ``activity_state`` (colored clock icon for overdue/today/planned)
Before this commit all cases were handled by calling search_read and then
counting records per group in a python script. This is very inefficient,
especially for ``activity_state``.
This new implementation relies on ``read_group`` when possible, i.e.,
when both grouping fields (kanban column and progressbar field) are
stored (case 1). It then falls back on a naive implementation inside
``_read_progress_bar``. Cases like 2 above can be addressed by
overriding ``_read_progress_bar``.
We also added some minimal test to ensure that we don't break anything.
1. Performance test on 60 K project.task records (kanban_state):
With a filter for 6 records:
```
| measurement | before | after |
|--------------------+--------+-------|
| number of queries | 8 | 5 |
| query time, ms | 11 | 7 |
| remaining time, ms | 21 | 9 |
```
All records:
```
| measurement | before | after |
|--------------------+--------+-------|
| number of queries | 67 | 5 |
| query time, ms | 300 | 55 |
| remaining time, ms | 1780 | 12 |
```
---
opw-2346901
task-1915411
X-original-commit: 153621bdbab94a2a94a5bbfcabb4111cbc5970d8
Co-authored-by: Raphael Collet <rco@odoo.com>
Re-introduce after discussion with AL the function that allow to specify
a custom placeholder for a specific model.
It has been removed because no more used since we use avatar mixin for
res.users and res.company. But it doesn't means that each model should add
his own mixin and controller and ... Keep it simple!
Use it for product and product template to have a default placeholder more
representative of the model.
Courtesy of xlu-odoo for this design
opw-2513801
closesodoo/odoo#73576
Signed-off-by: Jérémy Kersten (jke) <jke@openerp.com>
The use-case is when a "parent" field is put on the form view, as a
readonly field, and we create a new record. Before the fix, the fields
that depend on the parent record are always False, until the record is
actually created.
closesodoo/odoo#73446
X-original-commit: ff2af932c3b73dc59620a4ce17da3a13cab12cbe
Signed-off-by: Raphael Collet (rco) <rco@openerp.com>
This allows onchange to recompute fields that depend on the inherited
field that has changed. The use-case we have is a model that inherits
from 'account.move'. On its form view, setting the (inherited) journal
field does not recompute the (inherited) company field.
When the field 'journal_id' is modified, the onchange should:
- assign the new value to move_id.journal_id (*)
- recompute move_id.company_id (related='journal_id.company_id')
- recompute company_id
This commit adds the missing part (*).
X-original-commit: 2e05d0a361aa06a34eecf64613821f2e92295298
Due to odoo/odoo#51075 it's an error to assign the `_date_name` field on
an empty recordset.
```
Traceback (most recent call last):
File "/home/odoo/src/odoo/14.0/odoo/models.py", line 1582, in _fields_view_get
arch_etree = getattr(self, '_get_default_%s_view' % view_type)()
File "/home/odoo/src/odoo/14.0/odoo/models.py", line 1495, in _get_default_calendar_view
self._date_name = dt
AttributeError: 'project.phase' object attribute '_date_name' is read-only
```
Issue observed on upgrade requests 2615 and 3121.
closesodoo/odoo#73401
X-original-commit: 1bbe88796a5b03ab1ab2faec1f0fe4f735fcb485
Signed-off-by: Christophe Simonis <chs@odoo.com>
Co-authored-by: Christophe Simonis <chs@odoo.com>
* Objectives:
- Allow the generation of empty date/datetime groups outside the bounds
defined by the existing data during a read_group, with the fill_temporal
context key.
{[Jan - empty]->[Feb]->[Mar - empty]->[Apr]->[May - empty]}
This will allow to get a guaranteed amount of groups regardless of the
existing data.
- Request contiguous date/datetime groups between custom bounds which do not
cover the entire domain of existence (existing data)
[Jan] {[Jul]->[Aug - empty]->[Sep]} [Dec]
If some records were "forgotten" before the desired "interesting" period, we
won't have to display "many" undesired empty groups to show them.
* Implementation:
Modify the fill_temporal context key to accept a dictionary format,
with 3 keys, to be used in _read_group_fill_temporal:
1. fill_from
2. fill_to
3. min_groups
- (1., 2.) offer a way to guarantee that at least a certain date range will
be returned as groups. They are also the bounds to fill. Any group outside
those bounds will not be removed, but there will be no empty group added. If
any of those keys is omitted, existing bounds are used, if applicable. If
fill_from and fill_to are not specified, and there is no data, no group is
returned.
edge case example:
existing dates: [Jan, Sep]
fill_from: [Apr]
fill_to: [Jul]
interval: 1 month
result: [Jan, |Apr, May, Jun, Jul|, Sep]
general usage examples:
a. We have a Forecast in a Kanban view. We are in April and we want the
forecast to cover the [Apr - Jul] period.
- domain: [(>= Apr), (<= Jul)] (limit the read_group to
the forecast)
- fill_from: Apr (current month)
- fill_to: Jul
The result would be [Apr, May, Jun, Jul]
b. We want the same Forecast [Apr - Jul], but there is a chance that some
records were forgotten in Jan, and we want to be able to reschedule
them, but we don't need the fill_temporal to take effect from Jan.
- domain: [(<= Jul)] (we removed the first bound)
- fill_from: Apr (current month)
- fill_to: Jul
The result would be [Jan, |Apr, May, Jun, Jul|]. We have the forecast
with the fill_temporal, and every month preceding the forecast as long
as there is at least one record in it (no empty month before the
forecast).
- (3.) Will be used as a way to guarantee a certain amount of returned groups
for a read_group grouped on a {date, datetime} field. This will only work
if there is at least one group to be used as reference (starting point) to
create supplementary groups if necessary.
example :
existing dates: [Mar, Apr]
min_groups: 4
interval: 1 month
result: [Mar, Apr, May, Jun]
- fill_temporal: {} will be considered as True (backend & frontend)
- The new key options are independant from each other (do not require the
others to be set in order to function properly)
Task-ID: 2243913
PR odoo/odoo#69380
1. Objective
- We are in a Kanban view, and we group by a "date/datetime" field.
- In order to "quick create" a record in a group, or to "drag & drop" a
record between 2 groups, we need a default value for this date field.
- In this case, a group represents a "range" of dates. A default value could
be the last date of this range.
- Prior to this commit, the view had only access to a display string for this
date range. We want to have access to the concrete bounds (start/end dates)
- Since some date/datetime fields have backend constraints, it would be
difficult to generate a compliant global default value. As such, the use of
the default value should be disabled by default and activated on a field by
field basis.
2. Usage
- Use the last day (for dates), or second (for datetimes) of the range to set
a default value in kanban "quick create" and "drag & drop"
- Use the end of the range of the last group from a read_group to be able to
request the next group (chronological order) in subsequent read_group calls
3. What are the changes in this commit
- read_group from the "core" "models.py", with the added __range for
date(time) fields, a dictionary for each group. The keys are the field
names and the values are a dictionary with the following keys: value:
- from: starting date(time) of the range (inclusive)
- to: ending date(time) of the range (exclusive)
- XML changes:
- with a new xml attribute on <field> tag allow_group_range_value (for the
kanban view):
- we allow (or disallow) supported non-readonly fields to be:
- draggable
- quick created
- this attribute can be used only on date(time) fields, and if not set,
the default behavior is false
- the naming is subject to contention because it relates to
different features. The relation comes from the constraint:
To perform a drag&drop or a quickCreate, we must be able to get the
value from the group containing the record.
- JS changes:
- we add "group.range" after a read_group for date fields, which is a
dictionary: {field_name: {from, to}}, for each date(time) groupby field
- since date(time) fields can use the format "date_field:granularity" when
used in a groupby, we have to properly split out the granularity to ensure
compatibility with the code previously in place (drag&drop procedures)
- update mock_server and sample_server to make use of the date range.
Adding support for datetime grouping to both mock and sample servers, for
tests and previews.
- the default value for kanban drag&drop and quickCreate features is the
last day/second of the range for date/datetime fields
4. Why can't this range be computed from the original display value with
moment.js
The Babel python library that we use for date(time) formatting (2.6.0) is kept
at the same version as the package in stable Debian Linux. This version has
some issues with displaying weeks consistently around the new year in certain
locales.
We cannot reverse compute a date(time) label with moment.js to get a range since
nothing guarantees that the range will be the same one that Babel used to
output it.
5. read_group return format discussion
Prior to this commit, the format of the value for the groupBy field was
either :
- String : used as a display value for the group.
Most notably, for date and datetime, this value cannot be used as a field
value in most cases. It would be more practical to also return the date
range along with the display value, so that Drag&Drop and QuickCreate
functionalities can be properly enabled in views
example :
"March 2021" has the range ['&', (field, '>=', 2021-03-01),
(field, '<', 2021-04-01)]
This information is lost at the end of the read_group.
- Array : used to indicate a res_id in case of a many2one field
res_id = array[0]
field_diplay_value = array[1]
Since we cannot use the array format to send the date(time)s of the period range
because it could be confused with the many2one format, this commit introduces
the use of a new `__range` key which is a dictionary with specific related
keys to be used to defer usable data to the web client.
Task-ID: 2243913
PR odoo/odoo#69380
The import logging (ish) assumes that if an exception has at least 2
args the second arg is metadata added by the callee.
As it turns out, `UnicodeEncodeError` has *five* arguments, none of
which is added by us. So if encoding something fails during the
process (e.g. because the file contains a lone surrogate, which leads
to the database insert failing when psycopg2 tries to encode the query
to UTF8), then the `_log` function itself will fail, yielding a very
unhelpful error of:
dictionary update sequence element #0 has length 1; 2 is required
(because we tried to update a dict using a string).
This issue occurs only during *field conversion* and most fields have
no need to interact with the database (so don't need to encode the
value, which is what fails), however it is a problem when the invalid
string is used as a record name to look for (e.g. an m2o).
Further improve the experience by converting the UnicodeEncodeError to
a ValueError using the stringified UEE: `_log` assumes the first
argument to the exception is an error message of some sort, but for
UnicodeError subclasses it's just the encoding involved in the
error (here `utf-8`), which doesn't really serve as an error message.
Stringifying the exception generates a complete error message which is
quite a bit more helpful.
Specific update notes:
* avoid modifying the exception in-place, doesn't seem useful
* not sure why `field_name` was added as part of the augmentation
rather than up-front when `record` is created
Issue 2480064
closesodoo/odoo#72517
X-original-commit: 6c3c500929cd463cd3a1749f4ede8f3a8afd5748
Signed-off-by: Xavier Morel (xmo) <xmo@odoo.com>
The method read_combined() has been deprecated in favor of explicit
calls to read() and get_combined_arch().
There are places where we call _get_combined_arch() instead, in order to
avoid parsing the XML that has just been serialized. This saves useless
serialization-deserialization.
Task: 2541577
Meta task: 2463632
Co-authored-by: Fabien Pinckaers <fp@odoo.com>
Co-authored-by: Raphael Collet <rco@odoo.com>
Before this commit, `get_base_url()` could not be called on an empty recordset.
It might be called on an empty recordset for instance when getting the paypal
payment endpoint URLs.
This commit allows that and returns the ICP value in that case.
The goal is to always use the util method `get_base_url()` instead of directly
accessing the ICP.
closesodoo/odoo#68201
Related: odoo/enterprise#17538
Signed-off-by: Jérémy Kersten (jke) <jke@openerp.com>
Description of the issue/feature this PR addresses:
It is currently quite difficult to differentiate users. Most of the time, people
don't take the time to upload an actual avatar so everybody looks the same. This
PR generates a custom avatar with the users initials and random color to
differentiate them. For res.users, res.partner and hr.employee, image fields now
hold the binary image and avatar are used to show the image or svg.
Current behavior before PR:
Avatar had only random colors and was being saved in database, being inefficient
Desired behavior after PR is merged:
A new mixin defines image fields and in case no image is set, it generates an
SVG image with the user's initials and random color.
closesodoo/odoo#69819
Task: 2404630
Related: odoo/enterprise#18199
Signed-off-by: Sébastien Theys (seb) <seb@odoo.com>
The method filtered_domain() is broken for domains with hierarchical
terms ('child_of'/'parent_of').
To see *one* of the ways the implementation is broken, let `A` be a
model with `parent_id` pointing to `A`, and `a1` a record of model `A`
without parent (`a1.parent_id` is `False`), then this fails:
assert a1 in a1.filtered_domain([("parent_id", "child_of", a1.id)])
The reason it fails is that on
https://github.com/odoo/odoo/blob/f5519586d214a9b34ad24683a7f97c47802a3bad/odoo/models.py#L5377-L5380
`data` is empty since `a1` has no parent, thus
https://github.com/odoo/odoo/blob/f5519586d214a9b34ad24683a7f97c47802a3bad/odoo/models.py#L5403-L5404
fails, therefore the result of `filtered_domain` is empty.
Note: the implementation of the hierarchical operators is full of quirks
that are hard to emulate otherwise than by reusing the original code.
As a consequence, the current implementation may be broken in more than
one way.
Let's see another way the implementation is broken: let `B` be a model
without a `parent_id` field and with a `friend_id` field pointing
to `B`, and let `b1` be a record of model `B`. Then
b1.filtered_domain([("friend_id", "child_of", b1.id)])
throws an exception of the form shown below:
ValueError: Invalid field 'parent_id' in leaf "<osv.ExtendedLeaf: ('parent_id', 'child_of', 1) ...
Meanwhile the following code is still valid and returs b1:
B.search([("friend_id", "child_of", b1.id)])
closesodoo/odoo#71237
X-original-commit: e7a5ba95d8b7df5bbf545ef8afe0a1f5d0f70272
Signed-off-by: Raphael Collet (rco) <rco@openerp.com>
Use `frozendict` for dicts that often remain empty on models, like
_inherits and _depends, and also make `frozendict` more compact in
memory.
On a registry with 296 modules, this saves 635 kilobytes of memory,
which is about 6% of the registry's memory footprint.
closesodoo/odoo#70402
Related: odoo/enterprise#18142
Signed-off-by: Raphael Collet (rco) <rco@openerp.com>
This makes the union of many dicts into a single dict.
On a registry with 296 modules, this saves 300 kilobytes of memory,
which is about 3% of the registry's memory footprint.
This mainly keeps field.related's type consistent.
On a registry with 296 modules, this saves 325 kilobytes of memory,
which is about 3% of the registry's memory footprint.
This avoids a call to __init__() which no longer has access to the
model's class' parent classes, and was not even correct...
closesodoo/odoo#70400
Related: odoo/enterprise#18141
Signed-off-by: Raphael Collet (rco) <rco@openerp.com>
The base classes of models in the registry are modified when models are
extended by inheritance. But modifying cls.__bases__ is a costly
operation, so we manage to do it once per model when loading a registry.
Our benchmark consists in loading a registry with 296 modules. The net
time to load the registry is 25% smaller. In other words, the registry
loads 25% faster. Note that the time does not include the time to load
modules themselves, which does not change.
Have in a purchase order a first row with unit price 0
Export the PO adding unit price in the list of fields to export
The unit price is not reported
This occur because the first line of the order lines is meld with the
purchase order line but in the process the 0.0 values is discarded
opw-2510917
closesodoo/odoo#70368
X-original-commit: 242bf5024ce6ae9c32176cd1a65a4f57a242dd45
Signed-off-by: agr-odoo <agr-odoo@users.noreply.github.com>
Also speed up the basic setup of fields that are always duplicated
top-level (on the model's registry class). This saves time and memory,
as we also discard field.args and field._base_fields on toplevel fields
(those values are no longer useful after setup).
The attributes field.depends and field.depends_context are problematic
for sharing fields across registries, because they depend on the model's
registry class, which may vary from one registry to another. In order
to make computed fields shareable, we have to move those values away
from fields.
For the same reason, field.recursive should not be inferred, because its
value may depend on the registry, although it is generally not the case.
Moreover, the flag recursive=True is set on a field when field triggers
are determined (on the registry). A compute method may be called before
the flag is set (if no update has been done yet), and that can lead to
incorrect computations.
This happened in test TestUsers2.test_reified_groups in module 'base'.
The user groups view was apparently determined without the flag being
set, and the view depends on the recursive field 'trans_implied_ids',
which was not correctly computed.
We thus force developers to be explicit about recursive computed fields.
The code now logs a warning when the flag is not set up properly.
The basic setup of fields is now made in Field.__set_name__(), which
automatically makes this information shared across registries, because
it is done when importing Odoo modules.
Fields are no longer retrieved with inspect.getmembers(), which is
time-consuming. Instead, the classes defining models automatically
collect their own fields (via Field.__set_name__()), and the field
retrieval is done by using those collections. Also, a field no longer
needs to introspect its model class to find its field definitions.
A class that defines a model automatically determines its '_name',
'_inherit' and '_module' upon creation, and the fields determine their
'name', 'model_name' and '_module' via Field.__set_name__().
Loading a registry with 290 modules from scratch is now 25% faster.
