Funny things can happen if you have multiple machines and their clock are not
synchronized.

Also, consider the error list when fixit is False only

Add Business Template with Constraints as a example of bt5 with all configuration required to upgrade

Simplify erp5_upgrader to be possible upgrade any project just enabling the alarms that all Constraints will be called and fixed.

The alarms were renamed to cover each step of upgrader and give the possibility to call each step manually one-by-one. Also, all Scripts that the logic can be done by installing Business Template was removed.

Also added:
 - Constraints that can be used in the upgrader
 - Unit test to check different scenarios
 - Extend external method to be possible dump workflow chains to be compared with all configuration defined on Business Templates

Add test to simulate the issue with the method TemplateTool.installBusinessTemplateListFromRepository

Using the method try to install Business Templates that are already
installed, the Business Templates is uploaded on ERP5 but not used. It
creates a lot of garbage because in the end you have many bt5s not
installed.

The logic of all Configuration Items was not changed, but now the
changes are applied only if fixit is True and the constraints are
filtered by constraint type equal configuration.

Cosmetic: Trailing whitespaces were removed in the same commit

Also, Added constraint property sheets to Script Constraint portal type

Otherwise with CKEditor it's not possible to add some text outside of quoted
content

Forgotten be part of  9a710678

As MKCOL_handler() is not implemented on OFS.Folder.Folder class, the one from
ERP5Site was acquired when calling webdav.NullResource.NullResource.MKCOL().

When a portal type acquire role, it should place the role definitions in the same local role group ids

This reverts commit 1e90e4f8.
missing commit message

- we only want to sum transaction in the period when there is a from date.
- when from date is passed, really get period start date from from date

still keeping backward compatibility with older python_memcached.

instead of waiting for (long) timeout.

[Before]:

Where:

  in Supply or Delivery(Sale Order, Sale Packing List,..)

When:

( document.edit(something=some_value)

  OR

  document.validate(), document.start() something when workflow transition  )

AND

  document contains more than 100 sub objects, like 100 lines.

Then:
  /some_module/document/1/recursiveImmediateReindexObject
  /some_module/document/1/reindexObject
  /some_module/document/1/1/recursiveImmediateReindexObject
  .
  .
  /some_module/document/1/100/recursiveImmediateReindexObject

are triggered.

The first big recursiveImmediateReindexObject recursively reindex all the sub
objects. Thus the following reindex are duplicated.

[After]:

Reindex:
  /some_module/document/1/reindexObject
  /some_module/document/1/1/recursiveImmediateReindexObject
  .
  .
  /some_module/document/1/100/recursiveImmediateReindexObject

The root recursiveImmediateReindexObject is removed.

This has been the expected behavior from the beginning.
Before was only a side effect by giving empty dict to "if activate_kw == None"
expression.