[IMP] mis_builder: display blank cells instead of 0.0 when there is no data

[ThomasBinsfeld]

In KPI values computed on accounts where there are no moves during the period, show a blank instead of 0.0. This is nicer for the eye, and sometimes mandatory in some kind of reports.

This is implemented with an AccountingNone singleton which is a null value that "dissolves" in basic arithmetic operations with non-None values.

[oca-clabot]

Hey @ThomasBinsfeld, thank you for your Pull Request.

It looks like some users haven't signed our Contributor License Agreement, yet.
You can read and sign our full Contributor License Agreement here: http://odoo-community.org/page/website.cla
Here is a list of the users:

• @ThomasBinsfeld (login unknown in OCA database)

Appreciation of efforts,
OCA CLAbot

[sbidoul]

👍 (code review and functional test)

@ThomasBinsfeld can you please fill and send your individual CLA as explained here https://odoo-community.org/page/cla

[sbidoul]

@jbeficent @alexis-via you may want to look into this one

[ThomasBinsfeld]

@sbidoul ICLA is signed and sent 📧

[alexis-via]

yes, better to put 0.0 instead of bank field for those kind of reports which are mainly financial 👍

[sbidoul]

@alexis-via I'm not sure about your last sentence. To be clear:

• before this PR, KPI with no move lines in the period are displayed 0.0
• with this PR: KPI with no move lines in the period are displayed as an empty cell

This is important because:

• it distinguishes clearly KPI with no activity from KPI which have a 0.0 ending balances (for instance)
• in some cases (eg Luxemburg offical reporting) this distinction is mandatory

[adrienpeiffer]

👍 (Functionnal test & Code review)

[laetitia-gangloff]

Is it normal than comparison doesn't work with AccountNone value ? (compare with)

Before :

After :

[sbidoul]

@laetitia-gangloff good catch

[oca-clabot]

Hey @ThomasBinsfeld,
We acknowledge that the following users have signed our Contributor License Agreement:

• @ThomasBinsfeld

Appreciation of efforts,
OCA CLAbot

[sbidoul]

@laetitia-gangloff it should be ok, now.

[laetitia-gangloff]

It seems not ok in all cases (percentage) :

I don't know if it is ok to get +0 if there is no move (difference)?

[sbidoul]

@laetitia-gangloff p1 vs p3 is blank, that is normal because (-4960 - 0)/abs(0) does not make sense. p3 vs p1 is (0-(-4960))/abs(4960) = 100%. So that part is ok.

For the second part, we can hide the +0 indeed. I'll look into it.

[coveralls]

Coverage decreased (-0.6%) to 68.484% when pulling 266d055 on ThomasBinsfeld:8.0_add_AccountingNone into 631e6d9 on OCA:8.0.

[sbidoul]

@laetitia-gangloff 0 comparisons now show blank too.

[laetitia-gangloff]

👍

[sbidoul]

Ok, ready to merge.

[adrienpeiffer]

👍 Again ! Thanks !

[pedrobaeza]

What is the added value of using AccountingNone object instead of embedding the logic? I see that at the end, you have to add the clause if x is AccoutingNone in some if clauses.

[sbidoul]

What is the added value of using AccountingNone object instead of embedding the logic?

@pedrobaeza because the logic would be complex and would need to be repeated in user-entered kpi expressions.

The tests for AccountingNone are at the "boundaries" only (when rendering or returning the value to clients).

[pedrobaeza]

OK then

👍