Previously, mapped was following a very naive approach, which was simply
calling the field name passed as input for every record in a recordset,
sequentially.
The problem with this approach is that we will potentially recompute the
same fields multiple times for differents records, when this could be
done once per field for ALL records, and store this value in cache for
further access.
Another potential problem is that we don't take advantage of the ORM's
prefetching to fetch all the records that are not in cache at once,
instead of doing the same query for every record in the recordset.
Yet another problem is the conversion of each cache value to a record
format and then combining all of the individual records into a single
recordset, which, depending on the size of the recordset, can take an
unbelievable amount of CPU time.
With this new implementation of `mapped()` we take care of all of these
problems:
This is done by first delegating `mapped()` from the model to the field,
this mapped takes a recordset as input and it will try to batch compute
and prefetch as much as possible for the entire recordset, but it will
not keep these values for the actual output, it just stores everything
in cache and then at the end, retrieves everything from the cache to
guarantee the same order.
After the mapped, the conversion from cache format to record format is
delegated to the new `convert_to_record_multi` which will fetch all the
ids and then perform a single browse to encapsulate all of the records
into a single recordset with the least amount of overhead possible.
Part of Task 2170344
closesodoo/odoo#42611
Signed-off-by: Raphael Collet (rco) <rco@openerp.com>
Computing the `cache_key` turned out to be a big factor during the
lifespan of a `BaseModel.mapped` call and a lot of this time is spent
computing the same `cache_key` over and over.
These unnecessary computations can be easily reduced to a couple by
moving the `cache_key` method on the environment (instead of the field)
and by implementing a memo for that method. The rationale is that the
`cache_key` of a field does not change for a given environment.
The result of this patch is up to 50% faster `Field.__get__` which in
turn means a GLOBAL gain in performance, especially for methods /
functions that rely heavily on `__get__` such as `BaseModel.mapped`.
closesodoo/odoo#42674
Signed-off-by: Adrian Torres (adt) <adt@odoo.com>
Turns out we've got an operator just for the pattern of "match value
if there's one, otherwise match everything".
Also remove an example of domain in onchange doc.
closesodoo/odoo#40938
Signed-off-by: Xavier Morel (xmo) <xmo@odoo.com>
From now on, if one wants to force following operations to happen in a given company,
use with_company(company) or with_company(cid) to update the environment.
Since https://github.com/odoo/odoo/commit/a5b6f31cf28e5381e1c85f66730bcdb55998e643,
the current companies of the user are saved in the context as "allowed_company_ids"
context key.
In case of invalid context content, the api was computing the intersection
between the context content and the user company(ies) and falling back on user
company_id(s) when catching an error.
A sanity check was done, but no feedback was given to the user,
saying that the context change was falsy.
This commits changes this behavior to :
* raise an AccessError when trying to access self.env.company(ies) when
invalid or unauthorized companies are defined in the context.
* take sudo mode into consideration, allowing inter-company impacts,
even when current user doesn't have access to a given company,
if the code is done in a sudoed environment.
Co-Authored-By: Raphael Collet <rco@odoo.com>
Instead of using a filtered with `cache.get`,
use a dedicated method in the Cache class to get
the records having a different value in cache then asked.
This is mainly to avoid the creation of intermediate
`browse` of 1 record, when doing `for rec in self`
in `filtered`.
Creating browses is costly, and avoiding it leads
to performance gains.
The dedicated method `get_records_different_from`
loops on the record ids, instead of on browse records
These variable are not supposed to change within a same
environment.
The lazy property will compute these variables only
once, then store the result,
while the property were computing these variables
each time they were called.
e.g. for a 1000 iteration loop,
with `env.company`,
the company was computed 1000 times.
With a lazy property, the company will be computed one time only.
This branch is the combination of several optimizations in the ORM:
* store field values once in the cache: the cache reflects more
faithfully the database, only fields that explicitly depend on the
context have an extra indirection in the cache;
* delay recomputations by default: use method `recompute` to explicitly
flush out pending recomputations;
* delay updates in method `write`: updates are stored in a data
structure that can be flushed efficiently to the database with method
`flush` (which also flush out recomputations);
* make method `modified` take advantage of inverse fields to inverse
dependencies;
* filter records by evaluating a domain on records in Python;
* a computed field with `readonly=False` behaves like a normal field
with an onchange method;
* computed fields are computed in superuser mode by default.
Work done by Toufik Ben Jaa, Raphael Collet, Denis Ledoux and Fabien
Pinckaers.
closesodoo/odoo#35659
Signed-off-by: Denis Ledoux <beledouxdenis@users.noreply.github.com>
Multi is the default api for methods, it is not necessary to explicitly
decorate methods with it, adds clutter and most people use it because
they see that the rest of the code uses it.
Done with `find . -type f -name '*.py' | xargs sed -i '/@api.multi/d'`
`api.one` has been deprecated since v9 because it often makes the code
less clear and behaves in ways developers and readers may not expect
since functions decorated with it usually expect a `list` of record(s)
instead of the recordset, whereas most modern Odoo code expects `self`
to be a recordset.
The function `aggregate` was solely being used by `api.one` thus it has
also been removed since there doesn't seem to be any other use for it
thus far.
closesodoo/odoo#34555
Signed-off-by: Raphael Collet (rco) <rco@openerp.com>
The goal is to be coherent with the user property.
Actually, company_id and company_ids on the environment are no fields.
Calling env.company_id returns a browse record, not an id.
Purpose
=======
Allow the user to select the allowed companies for which he wants to see records
on top of selecting his current company.
It is confusing for users to see the records from the company he is connected to
and the records of the children companies.
Instead of using the hierarchy of companies to access records across companies,
the user can now select (from his set of allowed companies) the companies for
which he wants to access records.
/!\ This means that the user will interact with records from company A when in
company B.
Example: a SO has been created and confirmed in A. When in B, I create the
invoice from it.
Specifications
==============
1/ Deprecate the parent/children hierarchy on the res.company model. The fields are
kept on the res.company model to ensure the retro-compatibility, but won't be used
accross the standard code anymore. The only functional usage for this mechanism
was to allow to see records from several companies by creating a virtual parent
company, which will be possible with the new mechanism.
2/ By default, a user will only see the records of the company he is connected
to (or records without a company). (It is still editable by the user if needed).
For that, put this information in the user context, to allow having different
configurations on different browser tabs. Instead of having domains like
['|',
('company_id', '=', False),
('company_id', 'child_of', user.company_id.id)]
you'll have something like
['|',
('company_id', '=', False),
('company_id', 'in', company_ids)]
Note that the 'company_ids' is a value that is passed in the evaluation
context on the record rule, as we already have user, or time.
company_ids is a list of the ids of all the enabled companies in the
user's context.
3/ Out of the generic improvements brought by this task, this will illustrate
issues that could exist since several versions. For example, it should not be
possible to create a scrap order for the company A with a package of the company
B, or it should not be possible to create an invoice on the company A with
payment terms from the company B. Before the version 12.0, it was easy to
encounter this kind of issues as the admin was the SUPERUSER_ID. A positive side
effect of the fact that the SUPERUSER_ID has become an inactive user was to
make it more difficult to introduce mismatch on the records, but haven't solved
the issue, as it was still possible to do it with parent companies
configuration. Some of these issues have been fixed in this commit, but all the
business flows should be re-tested to check if an ir.rule should be introduced
(eg: a multi company rule for stock.quand.package), if the company of a record
is correctly transfered to another record created from the first record (eg:
From a SO, create an invoice and a payment, the company of the sales order
should be transfered on the invoice and the payment, even if the company of the
sales order is A and I'm logged into the company B with the company A enabled.
4/ Currently, if I click on a button on a notification email (example 'View
Task'), I face a traceback if I'm not logged into the company of the record.
Now, if you click on a button and if you have access to the record, the correct
company will be automatically set.
5/ If I display a kanban view with several records from several companies (and
an image), all the images should be displayed.
6/ Currently if you copy paste an url, this will crash if you're not in the
correct company. This won't be fixed because it's quite impossible to do it in
a clean way. This task brings a workaround. Copy/Paste -> Traceback -> Log into
the correct company, re-copy/paste -> Ok.
7/ 2 property methods have been added on the environment to retrieve the company
on which the user is logged in and the companies the user enabled, on a specific
tab.
That way, when creating a record, instead of doing
default=lambda self: self.env.user.company_id
do
default=lambda self: self.env.company_id
On the other hand, to retrieve the enabled companies, do
companies = self.env.company_ids
8/ Modify the Company Switcher widget to allow to log into another company
WITHOUT writing on the res.users (and thus bringing cache invalidation issues
and so on). Also allow to enable several companies and see records from several
companies, and independantly of the other browser's tabs.
9/ When focusing on a tab, save the current company configuration on the local
storage. That way, when doing 'CTRL+T' or a middle click, the context is
propagated to the new tab.
10/ Improve the error message in case of multi company access errors. Now, when
the user is in debug mode, display the related names of the records and the name
of the user who brings the issue.
11/ Remove the context erasing when writing on a res.users
This is probably coming from the migration to new API of the base module.
The context was not propagated at this moment, which was a common mistake at
that time. When migrating the module, probably by using the 'black box' method,
as the context was not propagated, it was erased on the new version. This is
now an issue because the context (i.e. the enabled companies) was erased when
writing on a res.users, leading to tracebacks.
See: https://github.com/odoo/odoo/commit/7eab8e26d3d46c53f4be924d6a34e80a66e74960#diff-4c2e738ee8f64f11806c889ea097b5e7R624
12/ Fix the crash manager on redirect warnings. The issue is the following
- Create an invoice on a company without a configured CoA.
- Set a partner
- On the onchange_partner_id, a redirect warning is raised to propose you
to configure a CoA
- Click on 'Go to the configuration panel'
- A generic warning says something like 'Do you want to discard your changes?'
- Click on yes, the page refreshes, but not on the redirect action.
Now, set correctly the action on the hash, and reload instead. The breadcrumb is
lost for example, but you reach the correct action at least.
13/ Introduce a res.group to enable/disable the multi company per tab
feature.
14/ To help the users to know which tab is in which company, add the
possibility to have a favicon per company. When creating a company,
the classical 'O' icon is colored by default in a random color.
15/ Remove the company switcher on the frontend. This was mainly there
to allow a user to swicth to the company linked to the website.
This behavior is now transparent to the user. If the website A is
activated, then the company set on the context is the company of the
website.
16/ Deprecated the _company_default_get method on the res.company
model. Remove the method _get_company on the res.users model.
17/ Add 'allowed_company_ids' and 'current_company_id' on the pyeval
context. You can now use those variables on domains in the views to
access directly to the activated company.ies on the current tab.
TaskID: 1960971
closesodoo/odoo#32341
Signed-off-by: Yannick Tivisse (yti) <yti@odoo.com>
This revision moves the `cache_key` to the first level
dict of the cache, instead of the last one.
Doing so, we reduce the number of times the reference
to the cache key is stored in the dict.
For instance,
for 100.000 records, 20 fields and 2 env (e.g. with and without sudo)
formerly, there were 100.000 * 20 * 2 occurences of cache key references
now, there is only 2 references.
Storing references to an object consumes memory.
Therefore, by reducing the number of object references
in the cache, we reduce the memory consumed by the cache.
Also, we reduce the time to access a value in the cache
as the cache size is smaller.
The time and memory consumption are therefore improved,
while keeping the advantages of revision
d7190a3fd0
which was about sharing the cache of fields
which do not depends on the context, but
only on the cursor and user id.
This revision relies on the fact there are less different references
to the cache key then references to fields/records.
Indeed, this is more likely to have 100.000 different records stored
in the cache rather than 100.000 different environments.
Here is the Python proof of concept that was used
to make the conclusion that setting the cache_key
in the first level dict of the cache is more efficient.
```Python
import os
import psutil
import time
from collections import defaultdict
cr = object()
uid = 1
fields = [object() for i in range(20)]
number_items = 500000
p = psutil.Process(os.getpid())
m = p.memory_info().rss
s = time.time()
cache_key = (cr, uid)
cache = defaultdict(lambda: defaultdict(dict))
for field in fields:
for i in range(number_items):
cache[field][i][cache_key] = 5.0
# cache[cache_key][field][i] = 5.0
print('Memory: %s' % (p.memory_info().rss - m,))
print('Time: %s' % (time.time() - s,))
```
- Using `cache[field][i][cache_key]`:
- Time: 3.17s
- Memory: 3138MB
- Using `cache[cache_key][field][i]`:
- Time: 1.43s
- Memory: 756MB
Even worse, when the cache key tuple is instantiated inside the loop,
for the former cache structure (e.g. `cache[field][i][(cr, uid)]`),
the time goes from 3.17s to 25.63s and the memory from 3138MB to 3773MB
Here is the same proof of concept, but using the Odoo API and Cache:
```Python
import os
import psutil
import time
from odoo.api import Cache
model = env['res.users']
records = [model.new() for i in range(100000)]
p = psutil.Process(os.getpid())
m = p.memory_info().rss
s = time.time()
cache = Cache()
char_fields = [field for field in model._fields.values() if field.type == 'char']
for field in char_fields:
for record in records:
cache.set(record, field, 'test')
print('Memory: %s' % (p.memory_info().rss - m,))
print('Time: %s' % (time.time() - s,))
```
- Before (`cache[field][record_id][cache_key]` and cache_key tuple instantiated in the loop):
- Time: 4.12s
- Memory: 810MB
- After (`cache[cache_key][field][record_id]` and cache_key tuple stored in the env and re-used):
- Time: 1.63s
- Memory: 125MB
This can be played in an Odoo shell, for instance
by storing it in `/tmp/test.py`, and then
piping it to the Odoo shell:
`cat /tmp/test.py | ./odoo-bin shell -d 12.0`
closesodoo/odoo#29676
instead of updating the cache by records.
We use `cr.fetchall()` and `zip` to get the values by field, which is more
convenient to be stored in cache.
Result of `cr.fetchall()`:
```
[
(3, 'Marc Demo', 'demo@example.com', '032'),
(1, 'Mitchell Admin', 'admin@example.com', '001'),
]
```
After being grouped by field:
```
[
[3, 1],
['Marc Demo', 'Mitchell Admin],
['demo@example.com', 'admin@example.com'],
['032', '001'],
]
```
That way we can update the cache of a field for all records at once, by calling
the builtin `update` method of `dict` with the `zip` of record ids and
corresponding values.
This significantly speeds up the storage of field values in the cache.
closesodoo/odoo#30817
The cache is quite time critical, as it is accessed millions times when reading
multiple fields on hundred of thousands of records.
Avoiding indirections and `if` statement when possible speeds up the cache
access time.
This revision moves the `cache_key` to the first level
dict of the cache, instead of the last one.
Doing so, we reduce the number of times the reference
to the cache key is stored in the dict.
For instance,
for 100.000 records, 20 fields and 2 env (e.g. with and without sudo)
formerly, there were 100.000 * 20 * 2 occurences of cache key references
now, there is only 2 references.
Storing references to an object consumes memory.
Therefore, by reducing the number of object references
in the cache, we reduce the memory consumed by the cache.
Also, we reduce the time to access a value in the cache
as the cache size is smaller.
The time and memory consumption are therefore improved,
while keeping the advantages of revision
d7190a3fd0
which was about sharing the cache of fields
which do not depends on the context, but
only on the cursor and user id.
This revision relies on the fact there are less different references
to the cache key then references to fields/records.
Indeed, this is more likely to have 100.000 different records stored
in the cache rather than 100.000 different environments.
Here is the Python proof of concept that was used
to make the conclusion that setting the cache_key
in the first level dict of the cache is more efficient.
```Python
import os
import psutil
import time
from collections import defaultdict
cr = object()
uid = 1
fields = [object() for i in range(20)]
number_items = 500000
p = psutil.Process(os.getpid())
m = p.memory_info().rss
s = time.time()
cache_key = (cr, uid)
cache = defaultdict(lambda: defaultdict(dict))
for field in fields:
for i in range(number_items):
cache[field][i][cache_key] = 5.0
# cache[cache_key][field][i] = 5.0
print('Memory: %s' % (p.memory_info().rss - m,))
print('Time: %s' % (time.time() - s,))
```
- Using `cache[field][i][cache_key]`:
- Time: 3.17s
- Memory: 3138MB
- Using `cache[cache_key][field][i]`:
- Time: 1.43s
- Memory: 756MB
Even worse, when the cache key tuple is instantiated inside the loop,
for the former cache structure (e.g. `cache[field][i][(cr, uid)]`),
the time goes from 3.17s to 25.63s and the memory from 3138MB to 3773MB
Here is the same proof of concept, but using the Odoo API and Cache:
```Python
import os
import psutil
import time
from odoo.api import Cache
model = env['res.users']
records = [model.new() for i in range(100000)]
p = psutil.Process(os.getpid())
m = p.memory_info().rss
s = time.time()
cache = Cache()
char_fields = [field for field in model._fields.values() if field.type == 'char']
for field in char_fields:
for record in records:
cache.set(record, field, 'test')
print('Memory: %s' % (p.memory_info().rss - m,))
print('Time: %s' % (time.time() - s,))
```
- Before (`cache[field][record_id][cache_key]` and cache_key tuple instantiated in the loop):
- Time: 4.12s
- Memory: 810MB
- After (`cache[cache_key][field][record_id]` and cache_key tuple stored in the env and re-used):
- Time: 1.63s
- Memory: 125MB
This can be played in an Odoo shell, for instance
by storing it in `/tmp/test.py`, and then
piping it to the Odoo shell:
`cat /tmp/test.py | ./odoo-bin shell -d 12.0`
closesodoo/odoo#29676closesodoo/odoo#30554
Consider a many2one field `foo_id` on model `bar`, with an inverse one2many
field `bar_ids` on model `foo`. During an onchange, the statement
bar.foo_id = foo
puts a special value in cache to add `bar` to the value of `foo.bar_ids`
without explicitly reading `foo.bar_ids`.
Executing the above statement a second time, the cache of `foo.bar_ids` is no
longer empty. This causes the actual value of `foo.bar_ids` to be read and
updated. The issue is that this can be slow for large values of `foo.bar_ids`.
Avoid reading the value of the one2many field by handling the case where the
cache contains the special value: simply update the special value to take into
account the second assignment.
closesodoo/odoo#28982
Followup from the previous commit: before this, _write takes 32% of
total runtime, of which 13.7% is ultimately assignable to add_todo.
Turns out for the test case we mostly keep adding records to recorsets
where they're already present, so the 13.7% of runtime in add_todo are
mostly spent creating orderedsets (11.17) then converting those back
into recordsets (1.75%) with some time spent creating lists and
appending records (already present) in them.
Checking if the records are already present before merging them in
decreases add_todo's runtime cost to 0.5%, and ultimately _write to
20%.
As many of the ignorable calls were merging a singleton or an empty
recorset into the parent recordset, optimising for these cases was
added to BaseModel's <= and >=.
Today, Odoo is really tricky to use without seeing the screen, it must be improved to be usable.
This PR forbid to use labels without a "for" attribute, add some title, rule and aria attributes in HTML. With that, Odoo will be fully usable with a screen reader.
* [IMP] Labels must have a for attribute. Improve accessibility.
* [IMP] Better error message when trying to read a missing cached value
* [FIX] Add some aria-label and title attributes for screen readers.
* [FIX] Template name is not included in the error message in case of SyntaxError in QWeb
* [FIX] Improve the Tour failed at step error message to be more explicit.
* [IMP] Add aria-labels
* [FIX] Add missing aria-label on failing test
* [IMP] aria-hidden means hidden. Fix all bad aria-hidden and hide aria-hidden for all.
* [IMP] Color names on kanban views and many2many tags
* [IMP] Add some checks on views for accessibility.
* [IMP] Add `alt` attribute on `img` tags.
* [IMP] Add aria-label and title on non-described icons
* [IMP] Add button role to widgets with btn class
* [IMP] Translate aria and formatted attributes.
* [IMP] Remove wrong aria-labelledby
* [IMP] Add menu role on dropdowns
* [IMP] Buttons must be focusable
* [IMP] Add aria attributes on progress bars
* [IMP] Improve accessibility of basic widgets
* [IMP] Change main layout to more semantic tags
* [IMP] Add menuitem role when missing
* [IMP] Remove wrong role='presentation'
* [IMP] Improve accessibility of tab panels
* [IMP] Add aria-invalid on invalid fields
* [IMP] Add aria-sort on ordered columns
* [IMP] Add role on alerts
* [IMP] Use dialog role, header, main and footer tags for modals
* [IMP] Add labels on o_status
* [IMP] Improve accessibility of kanban view with feeds and articles
* [IMP] Add alerts in case of new messages
* [IMP] Add widget, navigation or img role to aria-labelled items
Instead of searching for ids in the whole cache, search for the ones we will potentially prefetch.
Example: imagine you have 1K records in cache, and only 3 records in your prefetch set. Instead of retrieving the 1K records that have a value in cache, retrieve which of the 3 records that have no value in cache.
Adrian Torres