[MIG] queue_job_cron: Migration to 16.0

[MIG] queue_job: Migration to 16.0

Revert "[IMP] queue_job: requeue zombie jobs after hard shutdown"

This reverts commit db883b4.

[IMP] queue_job: requeue zombie jobs after hard shutdown

fixup! [FIX] use named cursor to iterate jobs

[13.0][IMP] queue_job: use context to execute the job (optional)

[13.0][FIX] base_import_async: keep context during import.

In some case, the lang set into the context is important to do the import. For example when the file contain a translatable name as a key (ex: product.attribute) and currently as there is no context, the search is executed in English (default lang) instead of user's lang

[14.0][IMP] queue_job current company

Use the current company to trigger the job (+ add related tests)

[14.0][FIX] queu_job: allowed_company_ids => use with_company(...)
Fill allowed_company_ids from context with the job's company instead of every allowed companies of the user.
Because most of the time, a job is related to only one company. And adding every allowed companies of the user into the context may load some unexpected records (during search for example). Because standards ir.rule use ['|',('company_id','=',False),('company_id', 'in', company_ids)] and this 'company_ids' is filled with every allowed companies from the context.

Merge commit 'refs/pull/365/head' of https://github.com/oca/queue int…

…o 13.0-wpu_master

Merge commit 'refs/pull/365/head' of https://github.com/oca/queue int…

…o 13.0-wpu_master