Unlike other database adaptors, memcache adaptors can be shared among
transactions. So pool them in order to reduce the number of connections
established to memcached servers. And as the connections handle
thread-safety internally (by inheriting from threading.local), threads
can share the same connection instance.
Also, do not create a new connector each time we get disconnected, just
reconnect.
Also, memcached internally retries connection, so no need to retry it
ourselves.
Also, do not log when failing to update & delete entries in cache
server: this is a racy cache, each read much be checked anyway, so
failed writes are just business as usual.
Also, document a bit more why we catch "any" exception happening during
finish, and specify base exception class following python coding best
practices.
Also, use named constant for loggin levels for remaining log directives.
Also, display traceback when catching a connection issue during
__getitem__, as the original exception gets converted into a KeyError.

/reviewed-on nexedi/erp5!305

This includes two parts, first is to change the condition for an organisation to be able to use the bank accounts from organisations from parent groups. This is useful for the case where organisations from subgroups does not represent different legal entities but simply some "business units". 

The condition for that have always been: if organisation does not contain bank account, this organisation can use bank accounts from the parent groups.

To be consistent with other places, such a closing accounting periods or setting grouping references, change the condition to: if the organisation does not contain accounting periods, this organisation can use bank accounts from parent organisations.

The behavior should not change much and allow to create bank accounts or cash registers that can be used only in one sub-group organisation.


Second part is about making it visually easier to understand which organisation the bank account belong to by adding some non-selectable options in the select for the organisations. This is only enabled when available bank accounts are not all contained in the same organisation.

![Screenshot_2017-06-19_at_15.24.35](/uploads/dc538645bb79a00f25444547d524635f/Screenshot_2017-06-19_at_15.24.35.png)

/reviewed-on nexedi/erp5!299

This is fix for #20170614-26135D "Bank Reconciliation only show transactions for the selected organisation"

At the same time, we stop using `portal_catalog.getObject` in `erp5_bank_reconciliation`.

/reviewed-on nexedi/erp5!297

By default in ERP5 all fields are not mandatory

Add transparent SVG and PNG icons to the manifest,
rename all functions following the Nexedi conventions,
link to the TodoMVC homepage, move Handlebars template to head,
and rename all files from tutorial_officejs_* to officejs_todomvc_*.

Fixes failing tests because of always showing "No work to be done" in worklist (lazy bug)

Test Results 20170621-6D7F8E4B

/reviewed-on nexedi/erp5!304

/reviewed-on nexedi/erp5!303

/reviewed-on nexedi/erp5!302

Name is unchanged, but it will use
Products.ERP5Type.Utils.checkPythonSourceCode which uses pylint.

/reviewed-on nexedi/erp5!301

/reviewed-on nexedi/erp5!296

The condition for an organisation to be able to use bank account from
other organisations higher in the group hierarchy should not be the
presence of bank account in this organisation, but whether or not it
acts as an independant accounting entity, and this is defined by the
presence of accounting periods.

Section bank account item list script was using `parent_strict_group_uid` related key but in 772033ed we removed
`parent_strict_group_uid` related key and added support to generate
`strict_parent_group_uid` related key.

 In that case, *Section Bank Account* field on payment transaction was empty  
![Screenshot_2017-06-16_at_18.54.33](/uploads/4f439bf14fb8b4d54df80031d281ce7a/Screenshot_2017-06-16_at_18.54.33.png) 

The 
`TypeError: Unknown columns ['parent_strict_group_uid']`
was only in the log files, because we are still ignoring errors in TALES.

I'm adding tests to cover this script, it's logic should have been tested anyway.

/reviewed-on nexedi/erp5!295

See discussions in nexedi/erp5!257 from background.

Here we try to reduce the differences in indexation methods that use `DELETE` and then `INSERT` (the one who does not always index the same number of rows).

In `z_catalog_predicate_list` the `DELETE` was missing, so in the case of predicate which stopped being a predicate, that predicate was never removed from the table.


/reviewed-on nexedi/erp5!294

@Tyagov @luke 

/reviewed-on !288

getObject is not public API. Bonus is we can do one query for all uids

Since it's allowed for multiple sections from same group to post
transactions on the bank account from the parent sections, the bank
reconciliation should use the same logic to be able to reconcile
transactions from any section.

( and AccountingTransaction_getDestinationPaymentItemList )

772033ed removed
`parent_strict_group_uid` related key and added support to generate
`strict_parent_group_uid` related key

If rows have been deleted before, it's more natural to simply INSERT and
no need for REPLACE

since predicate is scriptable, a document that was once a predicate may
no longer be a predicate next time it's indexed, so we should remove it
from predicate table in that case.

predicate_category was already doing this correctly

/reviewed-on nexedi/erp5!291