closes #4768

- delete a forgotten print
- allow pg_dump custom dumps to be larger than the available disk size, the
  previous commit allowed dumps to be larger than memory, this one remove this
  limitation. zip dumps are still limited to by the disk size.

- let the user choose between the pg_dump custom format or the zip format including the filestore
- use file objects to allow dumps larger than memory
- postgres subprocess invocation is now clean and thread-safe, we dont touch the local process environ anymore
- add a manifest to the zip dump format with version information about odoo, postgres (pg_dump doesnt output it) and modules

If a survey has "allow users to go back" option enabled, users can go back to previous pages and change their answers before submitting the whole survey.

This commit fixes the very special case, when the user reaches the last page of the survey, then click on "Previous" button, then reopen the survey from the invitation URL (/survey/fill/<survey_id>/<token> without the /prev flag in the URL). This won't crash anymore.

This commit fixes #2658 and #2680.

When user answers to multiple suggestion questions with a comment box,
the comment box was prefilled with the ID's of the suggestion instead of
user-entered data.

This commit fixes #3149 and closes #3171.

Avoid creation of doublon survey_user_input entries when a user loads the landing page of a survey (/survey/start/... route).

Due to some strange spec, jQuery.ajax() function called with "undefined" URL will do an extra call to the URL of the webpage where the script lies (http://api.jquery.com/jQuery.ajax/). Now, we check that URL is not "undefined" to avoid those calls.

By the way, this problem probably happened in every page that had survey.js in its assets... (correct loading of survey.js is fixed in saas-6 at 4dd5dbb2, this fix is complementary.)

This commit fixes #3032 and closes #3337 #3338 #3092.

[FIX] Purchase: replace SQL by ORM because ORM replace date=False by date is null automatically. Fix #4691

Payment info such as "You will be redirected to..." should be translatable.
opw 619786

When url_for was looking for a route which match, it was only looking for GET route.
So routes which were restricted to be used only with a POST method, were never found.

The result was that urls in website for route post (form in most cases) was never prefixed with the lang.
So the request.lang was always the default lang from website...

If you was creating a sale order (in ecommerce), the lang used in sale order was wrong and the description not in the current lang.

Adding a new context as, when creating a new journal from the pos.config,
that journal should automatically be defined as a POS payment journal.

[FIX] Correct resupply pull rules to have the correct warehouses to propagate / serve so it works with MTO rules also

Before 8.0, the field journal_entry_id did not exist.

For database coming from older release, like 7.0, this field is not filled in during the migration, because this is not possible.
Set the needaction to depend only on the journal_entry_id will have as effect to have every bank statement line entered when the database was under 7.0 to match the domain, while the needaction is made to display the number of records that need an action.

Besides, even in 8.0, this is possible that a line has not the journal_entry_id set, while not needing any actions (see 2bb38ca8)

For bank statement line having an account_id, but no journal_entry_id, it is not possible to reconcile the line in the bank statement reconciliation tool, as a filter is applied to only reconcile lines having journal_entry_id AND account_id not set.

As written in the help message of the account_id field:
This technical field can be used at the statement line creation/import time in order to avoid the reconciliation process on it later on. The statement line will simply create a counterpart on this account

Not allowing the reconciling should not prevent to close the statement in such a case. The button "close" was displayed only when all lines had journal_entry_id set.

Damn you commit -a

Make the field description on a forum translatable and add it in the forum view. opw 619786

Upstream traceability on produced goods (serial number on finished product) was
broken due to wrong values in cache for production.move_lines2 after production.
Refresh the value of production after each action_consume to make such the state
of the cache is correct. opw 609450

Similar fix for manufactruing order not going in done state in some specific
configrations (e.g. some components being phantom BOM).
Again due to wrong cache state after consumption. opw 610515
Fixes #1296

The tracking reference and other delivery references are not relevant to
duplicated pickings. Overwrite copy to remove carrier_tracking_ref, volume and
number_of_packages.

Add fallback on stock.picking.in and out to use copy method of stock.picking.

For partial delivery, the duplicated picking is the delivered order and
the existing picking is the backorder of the delivery (why so much hate?).
This means we have to switch the delivery info between the backorder and
the delivered picking.
Combo opw 615593 and 618802

The method test_if_product, used in the workflow to test that the mrp production is for a product (!= service), used to call the method _action_compute_lines in order to compute the production lines and determine from them the production type.

The thing is, the method _action_compute_lines, despite the fact it returns the lines of the production, actually creates the lines. So, just to test if the production was of product type, the productin lines were created, in database.

This rev. introduces a _prepare_lines method, which returns the computed production lines, without actually creating them in database, so the test_if_product method can test if the production is of product type without creating the production lines.

Therefore, production lines are now computed and created during the action_compute method, instead of computing them when the production was tested to get the production type.

Computing the lines before the action_compute has as side effect to not set the scheduled date of the work orders in module mrp_operations, at MO confirmation (as, on confirmation, the action_compute method is called only for productions for which the lines are not yet computed, and mrp_operations overide action_compute to set the scheduled date)

opw-620189

name_get of pos.category should use a browse instead of a read.
For a company having thousands of products and a few categories, using a browse
will greatly improve the load time as it is cached.

The constraint prevent sheets to overlap was broken because it relies on a check on the
user_id.  The problem is that the latter uses an old-api function field
that is not recomputed yet at the time of the validation.  The fix consists in
using a non-computed field instead.

This rev. is related to:
 - 4fb9c8f0
 - 6c785419

Closes #4217
opw-620175

Issue was the propagation of contextual values across actions, more
precisely conserving the selected fiscal year when selecting an account
from the chart of accounts tree view: the chart of accounts tree view is
generally opened for a specific fiscal year, and it seemed sensible that
opening an account would show only the journal items for the previously
selected fiscal years rather than all items ever.

PR #649 altered action.read by tentatively evaluating the action's
context, however this has the side-effect of providing evaluated
contexts when creating or editing actions via the UI, usually breaking
them in the process (as the context at this point is basically
nonsensical for the action's purpose).

This backs out the previous fix, and creates a fix restricted to the
tree view's JS (thereby removing the feature for window actions not
invoked from a tree view).

closes #4677, closes #4690

As date_start and date_stop are date field and not datetime, should use
DEFAULT_SERVER_DATE_FORMAT for search. opw 619592

POS users should not be able to create nor modify payment methods (account.journal)
POS users should not be able to create nor modify point of sales (pos.config)

At first opened session, if no payment methods was set, this is possible that the pos user should temporary have accesses granted to mark a payment method as pos payment method. This is done by the openerp.SUPERUSER_ID added by this rev.

opw-625489

Once the bank statement reconcilation done, the back button should not come back to Home when it does not found the bank statement list in the breadcrumb history, but simply perform a history_back action, which will come back to the previous action, the statement form for instance.

opw-625397

Without this fix, the 'Total' line of the pivot view does not display any inventory value, because there is no __domain as we are not asking the inventory value for any specific product.

This rev. is associated with rev. 38aa984f

closes #4660

Only admin was able to create product.putaway records. Gives all access
to warehouse manager.
If a putaway strategy was present on a location, a warehouse user was not able
to transfer goods as he had no access to the rule lines (no read to
stock.fixed.putaway.strat). Give read access. opw 619774

Colors were no more applied. colors attribute from tree expects a semicolon separated string, not a dict.

Regression introduced during b414e05a

Searching by location in stock.history failed as field name does not exists
Fixes #4650

When comment is created, emails are sent with subject: "Re: False" and footer: "About Forum False".
Now, when the post is a comment, we fallback to the name of the parent (the main forum post).