*= account_bank_statement_import_csv

An exception is currently triggered when a user attempts to import a bank
transaction file containing more than 80 transaction lines (see ref file [1]).

Steps to produce an error:
- Install `Accounting (accountant)` module
- Go to `Accounting` > Click on `Bank` > Click `Upload`
- Upload  ref file [1] and click `Test/Import` >>> Error occurs

Error: `psycopg2.errors.SerializationFailure: could not serialize access
due to concurrent update`

Error from 19.0: `InvalidSavepointSpecification : savepoint "ef05b579-df3f
-11f0-bc75-74563c5c983f" does not exist`

The issue occurs because, in `model.py` code line [2] creates a `savepoint`.
Before this `savepoint` is closed, code line [3] is triggered during the
creation of the bank statement line [4] and attempts to commit the
cursor using `self.env.cr.commit()`. Because a commit is executed while
the savepoint is still active, the system fails when trying to close the
previously created savepoint.

This commit fixes the issue by avoiding cursor commits during the import
process. The `import_file=True` flag is added to the context when
`_cron_try_auto_reconcile_statement_lines` is called from `execute_import`,
allowing the method to safely skip commit/rollback logic when
`import_file` is present in the context.

[1]: https://docs.google.com/spreadsheets/d/19hKnR8pGB27xkbEHgYYXIkBaPXV8RZZE/edit?usp=sharing&ouid=111844484867458262929&rtpof=true&sd=true
[2]: https://github.com/odoo/odoo/blob/11c469086cb4d08453a70cd7bd30d7391f635ae3/odoo/orm/models.py#L971-L973
[3]: https://github.com/odoo/enterprise/blob/2683b77cd6688877c308d0733bccfc5ad84530c1/account_accountant/models/account_bank_statement.py#L218
[4]: https://github.com/odoo/enterprise/blob/2683b77cd6688877c308d0733bccfc5ad84530c1/account_accountant/models/account_bank_statement.py#L1780

sentry-6974536471
opw-5359810

Forward-Port-Of: odoo/enterprise#102760

Currently, an error occurs when opening an `accounting report` when the 
evaluated data produces a `zero denominator` during formula computation.

**Steps to reproduce:**
- Install the `account_reports` module (without demo) and enable `developer mode`.
- Navigate to Accounting > Reporting > Deferred Revenue.
- Click the `gear icon` to open the report configuration.
- Click `Add a line` to create a new report line.
- Click `Add a line` to create a new expression with:
>- Computation Engine: `Aggregate Other Formulas`
>- Formula: `0/0`
- Save the new expression and report line.
- Try to reopen the `Deferred Revenue`.

**Error:**
`ZeroDivisionError: division by zero`

**Root Cause:**
After commit [1], at [2], expressions without the `ignore_zero_division` 
subformula are evaluated in the else condition, causing an error when 
the `denominator is zero`.

**Fix:**
This commit prevents errors caused by zero denominators during formula 
evaluation and improves the clarity of the resulting error messages.

[1]:
https://github.com/odoo/enterprise/commit/2ba548564bd0bb68c36a589ed67b43251b5f45d0

[2]:
https://github.com/odoo/enterprise/blob/7adab8bfdccf5f0e97eb2894e065b63ea8200d20/account_reports/models/account_report.py#L3540-L3547

opw-5475146

Forward-Port-Of: odoo/enterprise#103773

Currently, an error occurs when generating the `IRAS Audit File` for 
a `Singapore` company

**Steps to reproduce:**
- Install the `l10n_sg_reports` and `accountant` modules (without demo data).
- Create a new company with Singapore as the `country` and set `UEN` and `GST No.`, 
  then switch to this company.
- Navigate to Accounting > Reporting > Singapore > IRAS Audit File.
- Fill in the required details and click `Generate`.

**Error:**
`TypeError: 'account.account' object is not callable`

**Root cause:**
After the refactoring in PR [1], the code at [2] mistakenly calls `browse()` 
without arguments and then attempts to call the returned recordset with 
`line_account_ids`. Since `browse()` already returns a recordset, this results 
in calling an `account.account` recordset as a `function`, causing an `error`.

**Fix:**
This commit prevents an error during IRAS Audit File generation by correcting 
the `browse()` call to pass `line_account_ids` directly as its argument.

[1]:
https://github.com/odoo/enterprise/pull/72675

[2]:
https://github.com/odoo/enterprise/blob/444f120fbd4e7158b83c609ededf1b2598fc990f/l10n_sg_reports/models/iras_audit_file.py#L210

opw-5877947

Forward-Port-Of: odoo/enterprise#106150

When creating a future appointment via the backend, the generated lead's activity deadline is incorrectly set to today instead of the appointment date.

### Steps to reproduce

- Install `appointment_crm`.
- Go to Appointments > Schedule > Staff Booking.
- Create a booking for a future date (e.g., next month) and add a customer.
- Confirm the booking.
- Open the newly created Opportunity.
- Check the "Next Activity" deadline.
- It is set to today's date instead of the appointment's date.

### Cause

The `calendar.event` model maintains two sets of fields for timing: `start`/`stop` (Datetime) and `start_date`/`stop_date` (Date). Standard logic dictates that `start_date` and `stop_date` are only populated for All Day events (`allday=True`). For regular time-specific events, these fields are computed as `False` to avoid ambiguity.

When an appointment is created, `appointment_crm` generates a linked CRM Lead and schedules an activity. The code responsible for scheduling this activity (`activity_schedule`) was explicitly passing `event.start_date` as the `date_deadline`.

Since standard backend appointments are time-specific (not all-day), `event.start_date` is `False`. The `activity_schedule` method defaults to the current date (Today) when it receives a falsy value for the deadline. Consequently, creating a future appointment resulted in an immediate deadline.

### Fix

Modify the lead creation logic to use the standard calendar helper `_get_activity_deadline_from_start`.


opw-5780578

Forward-Port-Of: odoo/enterprise#106770
Forward-Port-Of: odoo/enterprise#106162