When a document gets rows inserted in category table while there was none
before (typically: first document indexation), it may trigger
  IntegrityError: (1062, "Duplicate entry '...' for key 'PRIMARY'")
because in the DELETE..INSERT pattern, DELETE finds no matching rows and
does not gap-lock (because we enable innodb_locks_unsafe_for_binlog), then
the second INSERT happening, chronologically speaking, waits for the
transaction of the first to e committed, and on commit it causes such
duplicate key error.

A transient visible effect of this change is that if both competing
indexations see a different document state (because document got modified
in some 3rd transaction), the union of the resulting set of rows will be
visible until the reindexation which must have been triggered by the 3rd
transaction gets executed, at which point only the latest set will be
visible.

A similar issue exists before this change, with stricter conditions: it
needs the intersection of both sets to be empty, because a non-empty
intersection causes the duplicate key error solved here.

This change has been measured to improve scalability of an invoice building
test case (naturally triggering indexations) starting from ~12 activity
nodes:
   8 nodes:  +1.4% invoices/hour
  12 nodes:  +9.5% invoices/hour
  16 nodes: +12.3% invoices/hour
(values are the difference between DELETE..INSERT and DELETE..REPLACE)

asSQLExpression() failed when executed several times on the same instance,
because too much code was skipped when 'column_map' is cached.

Avoid getting the list twice, follow coding style, explicitly access portal
methods on portal.

Remove condition to test if the event path destination predicate applies for this destination. On this this case, we can trust on the previous selection using the domain

call getObject() once explicitly to reduce number of calls.

this code was working just by chance because active_knowledge_pad is a brain, not an ERP5 object.

Revert "fix undefined variables"

This reverts commit 77a53809.

Revert "Listbox: calculate a mapping uid -> object once"

This reverts commit 84029c01.

Revert "fix list method returning duplicate uids"

This reverts commit 386b0e6b.

Revert "test for dialog listbox with editable fields"

This reverts commit e046d8f5.

Instead of looping through the list each time.
Also check that the list method does not return duplicate uids.

This cache is used by AccountingTransactionLine_getNodeItemList

It is useless to clear the cache in activities, because this would only work when using only one zope instance. Instead, clear the cache in the test for now.

Also do not hardcode 'business_process_module'.

>>> self.closure_cache[business_link] = new_business_process
was done after 'business_link' was overridden by the for loop.

This also the following line in getBusinessLinkClosure():
  if simulation_path.startswith(path):

For example, '/erp5/portal_simulation/1/10'
should not match '/erp5/portal_simulation/1/1'

amendments to dd69a61c, unsubcribe / subscribe instead of changing alarm node

mistakenly removed in a8b2267c

With this change, code like the following will use much less memory.

  for i in portal_catalog(...):
    i.activate().method()
Signed-off-by: Vincent Pelletier <vincent@nexedi.com>

so that we can register activity by path from restricted environment.
also set default parameters just same as ActiveObject.activate().

so that we can call ActivityTool.activateObject() with not only by
object but also by it's path.
Signed-off-by: Vincent Pelletier <vincent@nexedi.com>