Saturday, April 4, 2026
7 changes · saas-19.1
Enhancements to existing features
This update improves how the system handles errors when email templates fail to render. Instead of logging massive amounts of template code, the system now shows just the essential information (template name and a small snippet of the problematic code), making it much easier for support teams and administrators to identify and fix issues without cluttering server logs.
Original PR description
This is mainly a backport of an IMP done at https://github.com/odoo/odoo/pull/252455 Given it's potential to reduce server bloat and increase of QOL for sys admins and support agents, backporting (a…
This is mainly a backport of an IMP done at https://github.com/odoo/odoo/pull/252455 Given it's potential to reduce server bloat and increase of QOL for sys admins and support agents, backporting (a sligthly modified version) seemed adequate. Summary: When a QWeb template fails to render, the current logic logs the entire template source and raises a generic UserError. This leads to significant log bloat and makes it difficult for developers and support staff to identify the specific failing template or the root cause of the error. This commit improves the error handling in `mail.render.mixin` and `mail.template` by: - mail.render.mixin: Added logic to identify the failing template's name and ID if it belongs to a `mail.template` or `mail.compose.message` (mass mailing). - Log Truncation: Implemented truncation for identified templates, showing only a snippet (first and last 500 chars) in logs and UserErrors to prevent log/UI bloat while keeping full source logging as a fallback for unidentified templates. OPW-5980295 --- I confirm I have signed the CLA and read the PR guidelines at www.odoo.com/submit-pr Forward-Port-Of: odoo/odoo#257359 Forward-Port-Of: odoo/odoo#257104
Resolved issues and error corrections
This update fixes an issue where amounts with thousand separators (like 1,334.00) were incorrectly interpreted, leading to inaccurate reconciliation reports. A new function has been added to correctly parse these amounts, ensuring data integrity in financial reporting. This improves the reliability of our reconciliation processes.
Original PR description
When extracting amounts using a regex with a single capturing group, values containing thousand separators such as 1.334,00 or 1,334.00 were not correctly converted to floats. This could lead to…
When extracting amounts using a regex with a single capturing group, values containing thousand separators such as 1.334,00 or 1,334.00 were not correctly converted to floats. This could lead to incorrect amounts being interpreted in reconciliation models.
Added a new function `split_amount_str`in utils which will give
integer and decimal part for different number formats.
This is a heuristic approach, meaning it aims to provide the best
possible result for valid inputs. Invalid or ambiguous formats are
not guaranteed to be parsed correctly and may result in ('0', '0').
For the two capturing groups case, the first group is treated as the
integer part and the second as the decimal part, allowing users to
Explicitly split amounts like 9065 into 90.65 by using two groups
in their regex.
Examples:
EU format: 1.334,15 → 1334.15
US format: 1,334.15 → 1334.15
Implicit decimals: uid 01870912 0000009065 → 90.65 (using two groups)
Additional tests were added to ensure amounts with thousand separators
are correctly parsed.
Task [link](https://www.odoo.com/odoo/project/967/tasks/6026748)
task-6026748
Forward-Port-Of: odoo/enterprise#110410This update adds a specific line item to the CH balance sheet report to accurately reflect Treasury Shares (account 2980). Previously, this account was handled differently, and this change ensures correct reporting of equity balances for Swiss companies. It's a necessary correction to improve financial reporting accuracy.
Original PR description
This commit adds a dedicated report line for account 2980 (Treasury shares) to the CH balance sheet report. Account 2980 was previously included in the Legal reserves report line via the old formula:…
This commit adds a dedicated report line for account 2980 (Treasury shares) to the CH balance sheet report.
Account 2980 was previously included in the Legal reserves report line via the old formula:
```py
[('account_id.code', '>=', '290'), ('account_id.code', '<', '2991'), ('account_id.account_type', '!=', 'equity_unaffected')]
```
In recent commit https://github.com/odoo/enterprise/pull/102247/changes/81bcf433e909ce6ec56af484e9dd59a47fc87c98 the formula was narrowed down to :
```py
[('account_id.code', '>=', '290'), ('account_id.code', '<', '2970')]
```
And account 2980 was no longer considered.
Rather than adding it to the Legal reserves formula, a dedicated Treasury shares report line (CH_290_C) has been added under report line CH_290. New line as account 2980 represents a correction of equity (negative item) and is conceptually distinct from legal reserves. The parent line aggregation formula has been updated accordingly:
```py
CH_290_A.balance + CH_290_B.balance + CH_290_C.balance
```
see affected account: https://github.com/odoo/odoo/blob/19.0/addons/l10n_ch/data/template/account.account-ch.csv#L108
Forward-Port-Of: odoo/enterprise#111302This fix prevents files from becoming orphaned when an error occurs during file upload. Previously, if a file write operation failed (such as due to disk space issues), the system wouldn't clean up the incomplete file, causing storage waste and blocking future uploads of the same file. Now files are properly marked for cleanup before the write operation begins, ensuring they're removed if something goes wrong.
Original PR description
If an error occurs during the file write operation, the file will not be marked for garbage collection, which can lead to orphaned files taking up disk space or blocking other same file to be written. Step to reproduce the issue: 1. Create an attachment with a large file (e.g., 100MB) and save 2. During the file write operation, simulate an IOError (e.g., by filling up the disk space or changing file permissions) 3. The file will not be marked for garbage collection, and it will remain 4. Further attempts to create this same attachment will result in error: "The attachment collides with an existing file." opw-6055037 opw-5907025 Forward-Port-Of: odoo/odoo#257384 Forward-Port-Of: odoo/odoo#257274
This fix resolves a problem where copying and pasting text in certain website builder elements (like buttons and footer text) was incorrectly adding paragraph formatting. The editor now properly handles inline elements without wrapping them in unwanted paragraph tags, making the editing experience more intuitive and preventing unintended formatting changes.
Original PR description
Since [website builder refactor], the editor wraps inline elements at the root of an editable boundary. This is causing issues in website builder when the root element is not supposed to contain `p`…
Since [website builder refactor], the editor wraps inline elements at the root of an editable boundary. This is causing issues in website builder when the root element is not supposed to contain `p` or `div` elements (added to wrap the inline elements). This behavior can be configured globally with `allowInlineAtRoot`. And since 6bc5946796a364634cb5eb217e61a5b882a1b6b2, it can be configured per-element with a "predicate" resource. This commit adds a predicate to allow inline at root of editable when the root is (according to its tag) not supposed to contain `p` or `div` elements. Steps to reproduce: - Open website builder on a product page - Select text in "Add to cart" button - Copy - Paste - Bug: a `p` element is created (on master) - Open website builder - Select "Company" below footer - Copy - Paste - Bug: a `p` element is created [website builder refactor]: 9fe45e2b7ddbbfd0445ffe25a859e67a316d02b2 task-5868460 Forward-Port-Of: odoo/odoo#257512 Forward-Port-Of: odoo/odoo#247231
This update improves how the "My Team" filter works in Live Chat agent reports. Instead of grouping by department, the system now uses a manager-based hierarchy, showing each agent their own records plus records of employees they directly manage. This provides a more accurate and intuitive view of team performance data.
Original PR description
Replace the department-based domain with a hierarchy-based domain. The new filter includes the current user's records and the records of employees whose manager is the current user. follow-up of https://github.com/odoo/odoo/pull/253567 --- I confirm I have signed the CLA and read the PR guidelines at www.odoo.com/submit-pr Forward-Port-Of: odoo/odoo#257548
This fix resolves an issue where POS orders created in Jordan before installing the electronic invoicing module could not be submitted to JoFotara due to missing unique identifiers. The system now automatically generates these identifiers when needed, ensuring all orders can be successfully submitted to the tax authority.
Original PR description
Currently, POS orders in Jordan do not have a UUID if they were created before installing the module. This causes submission failures to JoFotara. Steps to reproduce: - Create a POS order in Jordan without the module l10n_jo_edi_pos installed - Install l10n_jo_edi_pos - Try to submit the order to JoFotara, it fails due to missing UUID with the error "Invoice UUID is required" This fix force a computation of the order UUID when we add the header node, to ensure that this will not cause submission failures to JoFotara. opw-6041688 Forward-Port-Of: odoo/odoo#255996