Also, drop isIndexable instance property, unmasking class property, which
already has the value we need.
Also, factorise catalog clearing by reusing ERP5Site_reindexAll's.

Allows having a more consistent API between this tool and all other
Nexedi-maintained tools.

Reduces code duplication.

portal_activities contains documents which need to be (and are) indexed.
For consistency, the tool itself should be indexable.

So it is not just an unused constant (OFS_HANDLER) and no implementation.

Also, clarify what exactly was the broken BTree marker value, fixing
isBTree accessor. Update associated tests.

erp5.util.testbrowser: Refactor of openNoVisit() and {Image,Submit}Control for recent zope.testbrowser (followup of 5a0b3763).

And specifically, Domains (more portal types may be added to the relation
string field widget as needed).
Predicates get expanded to the list of documents they match.

Can be used to identify whether a document should be used to match other
documents, rather than as a final document itself.

Reorganize code by parsing the appcache with jio.

Interface checking is done by a standalone gadget whch directly display the status.

Drop not needed code (do not reimplement getInterfaceList for example).

Force interface strict HTML format

/reviewed-on nexedi/erp5!706

/reviewed-on nexedi/erp5!705

Send 'on' value when checked.
Do not add checkbox in the request body if unchecked.

for example: SSLError: ('The read operation timed out',)

/reviewed-on nexedi/erp5!703

erp5.util.testbrowser: Require zope.testbrowser >= 5.0.0 since 5a0b3763 so update dependency accordingly.

Otherwise, the query will include a pointless "node.uid > 0" condition
(because internal API has no choice but to return a condition) which
confuses the query optimiser and may lead to very suboptimal execution
plans: a range scan of node.uid (catalog as node, node.uid=stock.node_uid)
despite provided account uid list constraining stock.node_uid.

Remove trailing whitespaces.
Make it easier to extend the list of indexed columns without causing
unnecessary diffs (sadly, SQL being SQL, those lines will still cause
extra diffs).

/reviewed-on !699

This Merge Request adds ```reachability_workflow``` and adds it to all Coordinate types  (Adress, Email etc.)
so that the can be marked as ```reachable``` or ```unreachable```.
It also adds ```coordinate_interaction_workflow```, so that when a coordinate text changes actions can be triggered.
The default behaviour is that, when coordinate text changes for an 'unreachable' Coordinate, it is marked as reachable.

We already discussed this with @jerome , but I cc @nexedi , since this can break compatibility in projects, because
in the past Coordinates acquired ```validation_state``` by their parent document (Person, Organisation, etc).
So now, if some code checks ```getValidationState``` (though it should be checking on the parent) for Coordinate, or searches Catalog with ```validation_state``` (though it should be ```parent_validation_state```), behaviour will change.

Test Result <a href="https://nexedijs.erp5.net/#/test_result_module/20180608-2D68AF16">here</a> seems good (3 failures in UI tests that seem unrelated)


/reviewed-on !691

Because the helper script for invoice, (order, packing lists, etc) printout is using `DateTime.strftime` it does not support dates before 1900.
This is a [documented behavior](https://docs.python.org/2/library/datetime.html#strftime-and-strptime-behavior) in python that dates before 1900 cannot be used.

In these cases it's not supposed to happen, but if by mistake user enter a date before 1900, printing the invoice would fail, which is not good.

These patterns were very simple and did not need strftime.

/reviewed-on nexedi/erp5!694