[19.0][FIX] edi_core_oca: use savepoint to prevent state corruption on errors

[guewen]

Forward #284

[OCA-git-bot]

Hi @etobella, @simahawk,
some modules you are maintaining are being modified, check this out!

[simahawk]

/ocabot merge patch

[OCA-git-bot]

This PR looks fantastic, let's merge it!
Prepared branch 19.0-ocabot-merge-pr-316-by-simahawk-bump-patch, awaiting test results.

[OCA-git-bot]

Congratulations, your PR was merged at 4bce3cf. Thanks a lot for contributing to OCA. ❤️

[HviorForgeFlow]

@simahawk @etobella @guewen this change brokes edi_queue_oca module, because doubling the savepoint (queue one and now edi_core_oca one) makes lose environment context :(

[guewen]

this change brokes edi_queue_oca module, because doubling the savepoint (queue one and now edi_core_oca one) makes lose environment context :(

@HviorForgeFlow do you have more details about this? Are you sure it is related? As far as I know, successful savepoints do not alter the environment or the context. BTW we are using this patch in production (in 18.0 though) since May 21th alongside edi_queue_oca.

[HviorForgeFlow]

@guewen I need investigate further but take a look on the error on the process:

Expected singleton: res.users()
Traceback (most recent call last):
  File "/opt/odoo/custom/src/odoo/odoo/orm/models.py", line 5939, in ensure_one
    _id, = self._ids
    ^^^^
ValueError: not enough values to unpack (expected 1, got 0)

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/opt/odoo/auto/addons/edi_core_oca/models/edi_backend.py", line 480, in exchange_process
    with self.env.cr.savepoint():
         ^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/odoo/custom/src/odoo/odoo/sql_db.py", line 116, in __exit__
    self.close(rollback=exc_type is not None)
  File "/opt/odoo/custom/src/odoo/odoo/sql_db.py", line 120, in close
    self._close(rollback)
  File "/opt/odoo/custom/src/odoo/odoo/sql_db.py", line 146, in _close
    self._cr.flush()
  File "/opt/odoo/custom/src/odoo/odoo/sql_db.py", line 182, in flush
    self.transaction.flush()
  File "/opt/odoo/custom/src/odoo/odoo/orm/environments.py", line 592, in flush
    self.default_env.flush_all()
  File "/opt/odoo/custom/src/odoo/odoo/orm/environments.py", line 383, in flush_all
    self._recompute_all()
  File "/opt/odoo/custom/src/odoo/odoo/orm/environments.py", line 376, in _recompute_all
    self[field.model_name]._recompute_field(field)
  File "/opt/odoo/custom/src/odoo/odoo/orm/models.py", line 6960, in _recompute_field
    field.recompute(records)
  File "/opt/odoo/custom/src/odoo/odoo/orm/fields.py", line 1891, in recompute
    apply_except_missing(self.compute_value, recs)
  File "/opt/odoo/custom/src/odoo/odoo/orm/fields.py", line 1861, in apply_except_missing
    func(records)
  File "/opt/odoo/custom/src/odoo/odoo/orm/fields.py", line 1915, in compute_value
    records._compute_field_value(self)
  File "/opt/odoo/auto/addons/mail/models/mail_thread.py", line 484, in _compute_field_value
    return super()._compute_field_value(field)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/odoo/custom/src/odoo/odoo/orm/models.py", line 4954, in _compute_field_value
    determine(field.compute, self)
  File "/opt/odoo/custom/src/odoo/odoo/orm/fields.py", line 81, in determine
    return needle(*args)
           ^^^^^^^^^^^^^
  File "/opt/odoo/auto/addons/account_accountant/models/account_move.py", line 76, in _compute_signing_user
    is_backend_user = self.env.user.has_group('base.group_user')
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/odoo/custom/src/odoo/odoo/addons/base/models/res_users.py", line 1074, in has_group
    self.ensure_one()
  File "/opt/odoo/custom/src/odoo/odoo/orm/models.py", line 5942, in ensure_one
    raise ValueError("Expected singleton: %s" % self)
ValueError: Expected singleton: res.users()

[guewen]

That's weird indeed. The savepoint does a flush, perhaps at this point in the code execution any flush would crash like this, so the fact that it happens on this savepoint might be only a symptom?

[HviorForgeFlow]

By default queue job do a savepoint, but when doing the second by edi_core_oca, it crashes...need further investigation yes...maybe in v18 it is not happening, or just the code don't check the user or other context which don't crash anything.

[guewen]

@HviorForgeFlow it may be related to this OCA/queue#922 actually?

[MiquelRForgeFlow]

@guewen I have looked into it, and yes, it seems OCA/queue#923 fixes the issue @HviorForgeFlow faced.