b0a4451c19c3ebb4a1803e6dd9ea0b554b8cc708
14
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 Gorash
|
ba1a5509fa |
[IMP] base: Remove context dependencies from get_views method
In this viewtiverse, the heroes remove the context dependencies for `get_views`, from the views and python fields (such as domain). To reduce inconsistencies and the number of rpc. Current issues: * There may be inconsistencies in views at the JavaScript level. Some overrides modify the behavior of get_views or domains on fields via context keys, therefore by changing the action, the rendering may be different. However, these views are cached. However, the cache key (Javascript) does not reflect the entire context, and requires additional post-processing from the server. * Multiple rpc for the same rendering. get_views being dependent on the context, as soon as it changes, a new rpc is performed. In most cases, when JavaScript needs the same view, there is no change depending on the context, the rpc is useless. * Inconsistency when rendering subviews, some views could be different depending on the context, this context can be modified in the view itself via the context attributes. However, the JavaScript client does not redo an rpc for each change of these sub-contexts. Therefore the result may be inconsistent. Solution: Limit as much as possible the number of context keys provided when calling get_views, and use the context provided as a cache key. The authorized keys are 'lang' and '*_view_ref'. For the cache key, options are added in the get_views method. Instead of using the context, it is inserted into python expressions. This will be evaluated by JavaScript and thus avoids inconsistencies. task-3414108 task-3414068 closes odoo/odoo#135145 Related: odoo/enterprise#47584 Signed-off-by: Raphael Collet <rco@odoo.com> |
||
|
Vincent Schippefiltandrco-odoo
|
f916298bdf |
[FIX] base: do not fetch computed field already in cache
Before this commit if a computed field is already in cache, but not its dependencies, `fetch` would fetch those dependencies. This commit ensures that fetch checks first if a computed is in cache before fetching its dependencies. This commit also follows dependencies of computed fields, if they depend on other computed fields. Finally, this commit consolidates `fetch` and `search_fetch`: they should use the same heuristics to know which fields to fetch. closes odoo/odoo#120001 X-original-commit: 6b680c463956f929db10d4c3058c36112a67e674 Signed-off-by: Vincent Schippefilt (vsc) <vsc@odoo.com> Signed-off-by: Raphael Collet <rco@odoo.com> Co-authored-by: rco-odoo <rco@odoo.com> |
||
|
Denis Ledoux
|
1744870d8d |
[IMP] base: cache back-end views
The master plan finally comes to an end. Thanks to: - odoo/odoo#87522 refactoring `load_views`, - odoo/odoo#94337 refactoring `common.Form` to prevent changing invisible fields in unit tests using `Form` instances, - odoo/odoo#95729 refactoring the behavior of `groups=` in views, - odoo/odoo#98551 removing the need of the `groups_id` many2many field on back-end views. The result returned by `get_view`/`get_views` can now finally be easily and efficiently cached, in order to cache back-end views. The goal of this revision is to cache the model views and fields already post-processed for the web client (with the modifiers, etc., already computed) without group restriction. Then, from this cached version, post-process group related features, such as removing nodes restricted with a `groups=` attribute, set the create/write button according to the user access rights to models, ... Not including the groups in the cache key allows: - to have less cached versions, (otherwise it would be one cached version per different group combination) - to not have to fetch the user groups to compute the key (with the current cache key, there is nothing to fetch from the database to compute the key) Besides, post-processing the groups features after taking the view from the cache of the view doesn't take a tremendous time: - parsing arch from/to string with etree is fast, - removing the `groups=` nodes using etree is fast, - adding the `create="False"`, `write="False"`, `delete="False"` on the view root node according to the access right of the user on the model is fast. This allows way faster calls to `get_views` by the web client, as the server no longer need, for each call, to fetch the views in database, combine the inherited views, post-process the modifiers attributes, etc. Timing tests are available on the pull request of this revision. Part-of: odoo/odoo#99417 |
||
|
Denis Ledoux
|
0501bbd62e |
[IMP] base: uniform "groups" in back-end view
This revision is to make uniform the behavior of the `groups` attribute on the Python model fields and on the node in the view architecture. In both cases, remove the node from the view completely. Before this revision, in a back-end view: - In the Python model, if a field has the `groups` attribute set and the user is not part of the groups, the field is removed, completely, from the view. - In the view architecture, if a node has the `groups` attribute set and the user is not part of the groups, the node is made invisible (not completely removed, just made invisible). in a front-end view: - if a node has a "groups" or "t-groups" set and the user is not part of the groups, the node is removed from the view. So it's 2/3 cases removing nodes restricted to a group. and 1/3 case making invisible nodes restricted to a group. It's simpler to have a uniform behavior for the 3 cases, simpler to understandard for developers. In addition, this will help for the goal to cache back-end views. It makes possible to convert views using the `groups_id` field by moving the content of these views directly in the view to which they add content which is suppose to be completely removed when the user has not the according group. By getting rid of the `groups_id` many2many field on `ir.ui.view`, it makes possible to cache the view architecture without requiring to use the groups in the cache key. Currently, if we want to cache the view architecture, it would be required to use the intersection of the user groups with the `groups_id` groups of the view, making it costly to compute the cache key, therefore altering the performance point to cache the view architectures. closes odoo/odoo#95729 Related: odoo/enterprise#29592 Signed-off-by: Denis Ledoux (dle) <dle@odoo.com> |
||
|
Raphael Collet
|
60a1452a40 |
[REF] base: adapt code to new flush API
Part-of: odoo/odoo#87527 |
||
|
Denis Ledoux
|
b03c227e88 |
[REF] models: refactor fields_view_get, load_views
Refactor the `load_views` API so it no longer sends multiple times the same fields description. e.g. When `load_views` is called to get the kanban, tree and form views, the list of fields of the model was sent 4 times: - Once for each view, with only the fields used in the view, in `['fields_views']['kanban']['fields']` for instance - Once globally, with all the fields of the model, in `['fields']` The goal of this revision is to change that so it sends the list of all fields only once. In addition, if a view contains x2many fields, the fields description of the comodel is also sent. It was sent in the `views` key of the view fields dict. e.g. When calling `load_views` of `res.partner` to get the kanban, tree and form views, the `res.partner` fields description was actually sent 6 times: - Once for each view - Once globally - Once for each view of the many2many field `child_ids` of the form view, in - `['fields_views']['form']['fields']['child_ids']['views']['kanban']['fields']` - `['fields_views']['form']['fields']['child_ids']['views']['form']['fields']` The change suggested in this revision is to: - Remove the fields description for each view in `['fields_views']`. As it no longer contains the fields, the key becomes `['views']` instead of `['fields_views']`. - Replace the dict key `['fields']` by `['models']`, which is a dict with as key the model name and as values the model fields description. It contains the fields description for all models implied in the view: the model of the main view and the model of all one2many and many2many fields. With this change, the fields description will only be sent once by model implied in the view. In addition, the web client was getting the information about the fields sometimes in the global fields description list (e.g. `['fields']`), sometimes in the fields description list of the view type (e.g. `['fields_views']['form']['fields']`), making it a pain to try to make changes / performance gain in these field description dictionaries, because you never knew in which dict the web client was getting its info. Now, as there is only one place to get the fields description from, it's clearer and cleaner. - one2many and many2many fields views are passed directly in the main view architecture rather than being put in the `views` key of the field description. This is actually easier to treat by the web client, and this will allow in a future work to cache an entire view in one block of text rather than having to combine multiple cached blocks of text to return one view. - one2many and many2many fields which do not have directly embedded views have their views directly injected in the architecture, so the web client doesn't have to do RPC calls to `load_views` for each one2many and many2many fields not having embedded views. For instance, this allow to reduce the number of RPC calls to `load_views` from 8 to 1 when loading the form of `product.product`. Currently, this behavior is limited to 1 level deep but we consider making it go all the way down in future works. We did not do it for the moment because in certain cases it rises the processing time and the size (bytes) too much. e.g. the sale.order view can be 5 levels deep, meaning you can reach 4 dialogs on top the main view. ``` sale.order form > order_line > sale.order.line form > invoice_lines > account.move.line form > asset_ids > account.asset form > depreciation_move_ids > account.move form. ``` This will also benefit in future works to cache an entire view in one block of text rather to having to combine multiple cached block of text to get one view. - `fields_view_get` becomes `get_view`. As it no longer returns the fields description, keeping the `fields` in the name `fields_view_get` no longer makes sense. Hence removing `fields` from the method name, it becomes `view_get`. As it gets renamed anyway, we take the opportunity to rename it `get_view`, which is more in line with the general getter/setter guidelines in the model object world. - `_fields_view_get` becomes `_get_view`. For the same reasons than above. - `load_views` becomes `get_views`. This is not mandatory, there is no technical reason to rename `load_views` as it practically sends the same info as before, the view architectures and their fields description. Just in another way. We just take the opportunity of this pull request to suggest a cleaner API: `_get_view`, `get_view` and `get_views`. - Arguments `toolbar=False, submenu=False` fo the methods `_fields_view_get` and `fields_view_get` are converted to a kwargs `**options` in `_get_view` and `get_view`. The rationale is that submenu was already no longer used (deprecated) and the mobile options is introduced. The mobile options is necessary to tell the server to send the mobile views for x2many fields (kanban instead of tree). Instead of adding a new argument each time we add a new option to `fields_view_get`, it seems wiser to have a kwargs `**options` to avoid to re-write all overrides each time a new option is introduced. - `_fields_view_get` returned a dict containing the arch in text and some of the view information. Now, `get_view` returns a tuple with the view architecture as an `etree` node, and the view as a browse record. The rationale is that all overrides of `_fields_view_get` were about modifying the arch only (e.g. changing the address format/re-organizing the address related field nodes of the partner according to the company country). To do so, all these overrides were doing `etree.fromstring` to parse the arch which was sent in text to convert it to an `etree`, then operations were done on the `etree`, and then `etree.tostring` was called to convert back the arch to string. With this change of signature to send the arch as an `etree`, all these back and forth `etree.fromstring` -> `etree.tostring` are avoided, allowing some performance gain and less code in the end. - A cleanup of the keys returned in the dict of `fields_view_get` has been performed in `get_view`: - `fields` is removed, as explained above, - `view_id` is renamed `id`, - `name` is removed, it was unused by the web client, - `type` is removed, it was unused by the web client, - `field_parent` is removed, it was unused by the web client, - `base_model` is removed, it was unused by the web client. - `filters` is moved from the global dict returned by `load_views` (now `get_views`) to the dict returned by `fields_view_get` (now `get_view`) as it applies only to the `search` view type. - Retro-compatible methods for the 3 methods `fields_view_get`, `_fields_view_get` and `load_views` are provided, with deprecation warnings in them. - The web client could cache the model fields description (as it already caches the views), so it doesn't need to fetch them again if it asks for another view of a model for which he already has the fields description. If we do so, `get_views` could return only the list of models used by the views, without the fields description as of now, and the web client would then call `fields_get` independently only for the models for which it doesn't have yet the fields description. This would avoid the server to return the fields description and to call `fields_get`, which is costly, for each `get_views`, therefore gaining performances. - Inject the views of the one2many and many2many fields all the way down, unlimited depth level, as explained above. - Cache with `ormcache` the architecture of back-end views. This is already done for qweb views, it's not done for back-end views. Therefore the postprocessing of the views is performed for each `get_views`, which is costly, while the view architecture doesn't change for users belonging to the same groups, according to the groups implied by the view. This pull request is co-authored by Aaron Bohy (aab) for the web client part and Denis Ledoux (dle) for the server part. Part-of: odoo/odoo#87522 |
||
|
Julien Castiaux
|
90b52d144a |
[REF] base: Use Command helper for x2many
Task: 2366606 |
||
|
Yannick Tivisse
|
04dc074804 | [IMP] base: Adapt tests to work with/without demo data | ||
|
fw-bot
|
ed1236275f |
[FIX] base: Remove label from views if user groups forbid the field
Without this patch, if a view defines a `<label for="forbidden_field"/>` anywhere, the web client tries to fetch the field data, but since the user has no access, it gets an error. Now, if the label targets a forbidden field, it gets removed from the view, as happens with the related field. The client doesn't know of that field's existence and no errors happen. closes odoo/odoo#38876 Signed-off-by: Thibault Delavallee (tde) <tde@openerp.com> |
||
|
Julien Castiaux
|
4f03a5f136 |
[FIX] *: remove old deprecated modules/functions
PEP-594 is deprecating a bunch of modules. As part of the cleanup, we are also dealing with long deprecated modules, functions and aliases. * `assert_` -> `assertTrue` * `assertEquals` -> `assertEqual` * `assertNotEquals` -> `assertNotEqual` * `assertAlmostEquals` -> `assertAlmostEqual` * `assertRaisesRegexp` -> `assertRaisesRegex` * `assertRegexpMatches` -> `assertRegex` * `base64.encodestring` -> `base64.encodebytes` * `base64.decodestring` -> `base64.decodebytes` * `inspect.getargspec` -> `inspect.signature` * `inspect.formatargspec` -> `inspect.signature` * `logging.warn` -> `logging.warning` closes odoo/odoo#36863 Task: 2003936 Signed-off-by: Raphael Collet (rco) <rco@openerp.com> |
||
|
Raphael Collet
|
b7fd679a6c | [FIX] *: sudo() -> with_user() | ||
|
Yannick Tivisse
|
a5b6f31cf2 |
[IMP] base: Contextualize the multi company
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
closes odoo/odoo#32341
Signed-off-by: Yannick Tivisse (yti) <yti@odoo.com>
|
||
|
Raphael Collet
|
0783318f97 | [FIX] adapt logger names | ||
|
Raphael Collet
|
9e64f9f951 |
[REF] openerp: move openerp to odoo
|