When a char field contains a value which represents a number (e.g. "00036"),
the value is inserted as a number in the formula instead of a string.
Because of this, the function value is not found.
actual: =ODOO.PIVOT.HEADER(1,"x_studio_barcode",00003456799)
expected: =ODOO.PIVOT.HEADER(1,"x_studio_barcode","00003456799")
closesodoo/odoo#152018
Opw: 3623662
Task: 3631998
X-original-commit: 9fedd9a5c3daacb71862864ede466d25420749bb
Signed-off-by: Lucas Lefèvre (lul) <lul@odoo.com>
Signed-off-by: Pierre Rousseau (pro) <pro@odoo.com>
In a spreadsheet with multiple data sources (2 pivots), each data source
initially loads and triggers a new evaluation upon loading.
This results in two evaluations, even if both data sources resolve in less
than 10ms apart. In such cases, the first re-evaluation becomes redundant,
as a new one is immediately triggered.
The issue is worse when more than 6 RPCs are required, as most browsers limit
network calls to 6 in parallel. Consequently, the 7th RPC will unnecessarily
wait after the evaluation triggered by the first RPC to resolve.
For spreadsheets with many many data sources, the accumulation of these
pointless evaluations significantly impacts performance.
In a real-life scenario with 18 data sources from our production database,
the spreadsheet took approximately ~33s to fully load and become reactive.
With this commit, the loading time is reduced to ~7s (only one evaluation
instead of 18) (tested in 17.0).
Note that this testing was conducted locally, with minimal latency, and with
a limited amount of data.
One consequence of this commit is that cells won't load incrementally as
each data source loads. Instead, all cells will display "Loading..." until
all data sources are loaded. Given the substantial speed improvement, we
consider this trade-off worthwhile.
This fix only impacts loadable datasources (pivot, lists, graphs),
it could also include data sources using individual RPCs (currency,
accounting). Maybe for master.
closesodoo/odoo#150015
X-original-commit: 70877d29cc2368298f8716f64c248a86ed0416ce
Signed-off-by: Vincent Schippefilt (vsc) <vsc@odoo.com>
Signed-off-by: Lucas Lefèvre (lul) <lul@odoo.com>
Currently, if you have a pivot grouped by a date field, aggregated by week
month or quarter, sorting all the pivot cells does not work as expected
when sorting based on the date header column.
"April 2023" would end up being before "March 2020" just because "A" is
before "M". Similarly, "W1 2023" is before "W2 2020" and "Q1 2023" is before
"Q2 2020"
With this commit, for months aggregates, the result of
`=ODOO.PIVOT.HEADER(1,"create_date:month","04/2023")` is currently the
string "April 2023".
The result now becomes a real date just like any other date value
in a spreadsheet. It's the number corresponding to the first day of the
month.
For week and quarter aggregates, we could move the year first ("W1 2023"
becomes "2023 W1"). However, we decided not to do it to keep consistency:
- with other places in odoo (pivot views)
- with the way we talk/think (quarter/week comes first)
closesodoo/odoo#139295
Task: 3570281
Related: odoo/enterprise#49300
Signed-off-by: Vincent Schippefilt (vsc) <vsc@odoo.com>
This commit fixes 2 very similar issue.
See enterprise commit.
- Group a pivot by any date field, with the day aggregate
- insert it in spreadsheet
- autofill a pivot function
=> the autofill tooltip displays the day number instead
of the day formatted as a date.
- on the same pivot
- click on menu "Data > Insert pivot > insert pivot cell > [you pivot]"
=> in the dialog, numbers appears for the headers.
Task: 3570281
Part-of: odoo/odoo#139295
This commit factorizes how the format is computed for ODOO.PIVOT and
ODOO.PIVOT.HEADER functions. It was essentially duplicated.
Also move the date(time) format responsibilty to each time adapters,
instead of handling the different aggregate cases separatly.
This commit also prepares the next commit which fixes a formatting bug.
Task: 3570281
Part-of: odoo/odoo#139295
For the ODOO.PIVOT.HEADER functions, "special" values such as measure and
total were managed in different places (total management was also duplicated
see enterprise commit).
Now the measure and total are managed in one place, in the high level method.
There's also now a dedicated method to get a measure display name. The method
`getGroupByDisplayLabel` was perverted at that purpose (see enterprise commit)
Task: 3570281
Part-of: odoo/odoo#139295
Method names in the pivot data source/model are not particularly clear
and self-explanatory.
To commit renames some methods (and their argument names) with hopefully
more meaningful names.
I'm also moving `getDisplayedPivotHeaderValue` (now
`computeOdooPivotHeaderValue`) from the model to the data source. It's
a high level function, the implementation can be in the data source.
Task: 3570281
Part-of: odoo/odoo#139295
- define a relational global filter without any default value.
- reference that filter with `ODOO.FILTER.VALUE`
=> when loading the spreadsheet, a `read` RPC is triggered
with an empty list of ids.
This is:
- useless network call
- useless evaluation when the RPC resolves
closesodoo/odoo#149741
Signed-off-by: Rémi Rahir (rar) <rar@odoo.com>
A date filter with an empty value would still create a domain filtered
on the current year instead of being bypassed.
closesodoo/odoo#148984
Task: 3668916
X-original-commit: 763f88e2d45cb7ba107db135eecd0365ba2b8c8b
Related: odoo/enterprise#54095
Signed-off-by: Pierre Rousseau (pro) <pro@odoo.com>
Signed-off-by: Rémi Rahir (rar) <rar@odoo.com>
- Inserting a cumulated graph view inside spreadsheet
(e.g. cumulated subscription MRR breakdown),
- apply a global filter to filter, let's say on the current
month
=> the first data point does not include data from before
the current month (the accumulation starts at 0, even though
there is data before)
You can also check the MRR evolution subscription dashboard,
"MRR over time" chart.
Note: with this fix, we assume all "cumulated" charts are also
"cumulated_start". Which is true in practice (only one cumulated
graph view in the entire codebase)
closesodoo/odoo#149084
Task: 3680601
Related: odoo/enterprise#54146
Signed-off-by: Lucas Lefèvre (lul) <lul@odoo.com>
Change your language and open any dashboard.
The table titles (e.g. "Top countries" of the Leads dashboard) are not
translated.
It was lost in commit odoo/o-spreadsheet@9616681
With this commit not all link labels are translated. Only odoo links.
I don't think other (regular) links should be translated.
closesodoo/odoo#146574
X-original-commit: a2e406db138d51e710caf61a2809cfa991ea6cb3
Signed-off-by: Rémi Rahir (rar) <rar@odoo.com>
Signed-off-by: Lucas Lefèvre (lul) <lul@odoo.com>
Since https://github.com/odoo/odoo/pull/103510, the read_group parameter "groupby" can no longer contain
implicit duplicates. For example groupby=['date', 'date:month] doesn't work
anymore. Here we remove all duplicates from groupby before making a read_group.
closesodoo/odoo#143792
Signed-off-by: Julien Mougenot (jum) <jum@odoo.com>
The arrow in the inputs of the side panel was stuck to the side of the input,
without any padding.
closesodoo/odoo#143191
Task: 3376873
X-original-commit: 97c2f7832669ce01f70ee0b5474328382b3f8c2f
Signed-off-by: Lucas Lefèvre (lul) <lul@odoo.com>
Signed-off-by: Adrien Minne (adrm) <adrm@odoo.com>
This commit modifies the functioning of `ODOO.PIVOT.TABLE` formula when the
boolean `includeTotal` is set to 'false'.
Earlier, total values of both row and column were not displayed in this case.
But now, if a groupby is not applied on any of the two dimensions, it displays
total values of the dimension on which groupby has been applied.
Otherwise the formula serves no purpose since it doesn't display any
data.
For example, one row groupby and zero column groupbys are applied on pivot:
The formula `=ODOO.PIVOT.TABLE(1,,false)` would display all cells of pivot
(including the row totals), except the last row (i.e. the column total).
Task ID : 3580623
closesodoo/odoo#140970
Signed-off-by: Lucas Lefèvre (lul) <lul@odoo.com>
Steps to reproduce:
- Insert a list/pivot in a blank spreadsheet from a module (say Sales).
- Apply a global filter on it.
- Insert an odoo chart in the same sheet but from a different module (eg. CRM)
This works just fine when chart is inserted from the same model because it matches
the existing filter (from the pivot/list), and directly returns that field matching
without check the field matching from charts.
However in case of different model, the issue is that when there are no charts in the
sheet, the existing code of `getOdooChartIds` retrieves the incorrect chart ids
(`getChartIds` getter returns all chart ids, including the id of chart being inserted).
This leads to a traceback as the code tries to fetch fieldMatchings for a non-existent
chart within the sheet.
This commit resolves the issue by modifying the `getOdooChartIds` method to now utilize
`this.charts` instead of `getChartIds` getter, which correctly provides the ids of
charts already present in the sheet.
Task ID: 3573402
closesodoo/odoo#140700
Signed-off-by: Lucas Lefèvre (lul) <lul@odoo.com>
The current spreadsheet layout is not adapted to print the document as
the spreadsheet component is designed to only be rendered in the visible
part of the client viewport, the invisible parts of the spreadsheet are
not rendered.
This revision adds the possibility to render the whole spreadsheet when
the user wants to print the document (either via a top menu button or
with the generic shortcut Ctrl+P).
closesodoo/odoo#140281
Task: 3575484
Related: odoo/enterprise#49829
Signed-off-by: Lucas Lefèvre (lul) <lul@odoo.com>
This revision extends the caching introduced in [1] to cache the
Pivot cells of a `PivotTable` for every combination of `includeTotal`
and `includeColumnHeaders`. the conditions for the `clickableCell` of
`ODOO.PIVOT.TABLE` cells fetch the pivot cells a LOT, which becomes
problematic for big pivots.
For a spreadsheet with a formulat `ODOO.PIVOT.TABLE` that spreads over
2562 (122*21) cells:
| | Before | After |
|----------------------------------|---------|--------|
| `getPivotDomainArgsFromPosition` | 1227 ms | 124 ms |
| load spreadsheet as dashboard | 7.97 s | 2.4 s |
| | | |
[1] https://github.com/odoo/odoo/pull/138594closesodoo/odoo#141116
Signed-off-by: Lucas Lefèvre (lul) <lul@odoo.com>
In order to identify an ODOO.PIVOT.TABLE formula cell, we set the
display name of the pivot as the evaluated value of the cell containing
the formula.
closesodoo/odoo#140674
Task: 3580153
Related: odoo/enterprise#50003
Signed-off-by: Rémi Rahir (rar) <rar@odoo.com>
Purpose
=======
This task answers two use cases making text filters more useful and easier to
use:
Always using the same few values
--------------------------------
The text filter allows to type any arbitrary text. However, some users will
always use the same few values. They will always again and again type the
same values, every time they want to set or change the filter value. That's
cumbersome. There's also a risk of making a typo while typing.
Parametric dashboards
---------------------
Some dashboards are built to be parametric (our business analysts do that a
lot). e.g. They put the measure field in a cell, then reference that cell in
PIVOT functions. If you want to have the same dashboard, but with another
measure: just update the cell with the new measure and that's it.
In read-only mode however, you can't update that cell :(
But you can mimic a variable parameter with a text filter. Create the text
filter, then get the filter value in a cell with the function
=ODOO.FILTER.VALUE("..."). Now you can update the cell value by setting
different values in the filter input.
However, you need to know exactly what value would be correct/valid for the
parameter. Typing any arbitrary text would lead to errors or unexpected
results. Business analysts can know that kind of technical stuff, but
lambda users don't. The solution is currently to duplicate the dashboard and
change the cell value with the exact parameter value you want.
Specification
=============
In the filter config side panel, allow the user to restrict the set of
possible values to values in a range of cells in the spreadsheet. Let's say
A1, A2, A3 contains Paris, Brussels, Berlin. If the user chooses A1:A3 in the
side panel, the text filter input is no longer a free text input, but becomes
a select with the 3 cities.
If a value from the range is selected, then the cell with this value is
changed to another value: keep the previous selected value selected.
display the values with their cell format but use their raw format for the
logic behind it (using the value as a cell value from ODOO.FILTER.VALUE)
closesodoo/odoo#139191
Task: 3554062
Related: odoo/enterprise#49204
Signed-off-by: Rémi Rahir (rar) <rar@odoo.com>
Change the global filter icon from the `fa-search` to something more
filter-like.
Also use the same icon in shared dashboard.
closesodoo/odoo#139164
Task: 3560817
Related: odoo/enterprise#49166
Signed-off-by: Lucas Lefèvre (lul) <lul@odoo.com>
If a chart was linked to an odoo menu, but the odoo menu didn't have
an action, the user would get a traceback when clicking on the chart.
This commit:
- Improves dashboard validation. Now we test that the menu is linked
to an action, in addition to testing that the menu exists
- Send a "danger" notification when the user clicks on a chart with
a menu without an action linked to it, rather than a traceback
closesodoo/odoo#139324
Task: 3563450
Signed-off-by: Rémi Rahir (rar) <rar@odoo.com>
In the goal of simplify assets loading, in this commit we create a new
assets bundle for chartJS and its luxon adapter.
With this, we can now use loadBundle instead of load these two libraries
with loadJS.
task-3562357
closesodoo/odoo#139544
Signed-off-by: Michaël Mattiello (mcm) <mcm@odoo.com>
In the dashboard action, when switching from dashboard to dashboard, the
size of the control panel flickers. That's because the Share button is not
displays while the dashboard is loading and it takes some place, making
the control panel taller when it's displayed.
With this commit, the Share button is always displayed (disabled when the model
isn't loaded). In addition to fix the size flickering issue, it's also less
things appearing/disappearing from the UI (less sapin de Noël)
closesodoo/odoo#138816
Signed-off-by: Rémi Rahir (rar) <rar@odoo.com>
Adds in date type global filters a new category
"From / To" allowing to define a domain between
two dates.
closesodoo/odoo#138507
Task: 3516362
Related: odoo/enterprise#48855
Signed-off-by: Lucas Lefèvre (lul) <lul@odoo.com>
Most tests that need the user id simply use `uid=7`, which will fail
if the default user id ever change in the tests. This commit replaces
it by `sessuin.user_context.uid`.
closesodoo/odoo#138733
Task: 3553122
Related: odoo/enterprise#48965
Signed-off-by: Lucas Lefèvre (lul) <lul@odoo.com>
Re-inserting a pivot that had a contextual domain would fail because
the data source would try to fetch the data with the domain as a string,
but the domain should be a list at this point.
Task: 3553122
Part-of: odoo/odoo#138733
When importing a pivot with a date field in week format, the date was
formatted as "W2023 40" instead of "W40 2023".
closesodoo/odoo#137682
Task: 3539629
Signed-off-by: Lucas Lefèvre (lul) <lul@odoo.com>
There's no need to recompute the table structure
every time. The pivot model is never mutated, we can safely
store the table.
closesodoo/odoo#138594
Related: odoo/enterprise#48910
Signed-off-by: Rémi Rahir (rar) <rar@odoo.com>
Allows to "See records" or "Set matching filter" on ODOO.PIVOT.TABLE
cell results.
(from the context menu in "spreadsheet mode" or by clicking on cells
in dashboard mode)
Task: 3318865
Part-of: odoo/odoo#138594
This commit introduces a new pivot function:
ODOO.PIVOT.TABLE(<pivot_id>, [row_count], [include_total], [include_column_headers])
Purpose
-------
It answers several current limitations:
- The pivot in spreadsheet not being "dynamic" (growing as new groups/records
are created) is a *very* frequent feedback
- sorted pivots are very cumbersome (# syntax) and takes a lot of cells
- cannot easily compute arbitrary aggregates (max, avg). You only get the
total. ... as soon as new groups appears that are not reflected in the
spreadsheet, you can't guess data is missing and that it's not taken into
account.
Specification
-------------
Add a single and simple function =ODOO.PIVOT.TABLE(<pivot_id>) that would
display the entire table (dynamic, as more groups and records are added).
We can achieve almost everything we want:
- by combining with other functions (INDEX, CHOOSECOLS, CHOOSEROWS, XLOOKUP,
...) if we want to target a specific row or column.
- by using infinite ranges (e.g. =MAX(B2:B)) to account for a growing table
- we could remove all the positional PIVOT version of the function
(=ODOO.PIVOT.HEADER(7,"#country_id",1)). The implementation is a mess
(lots of ifs everywhere to account for the #). Just replace it by a single
function SORT(ODOO.PIVOT.TABLE(1), 1) (or with NSORT, or sorted from the
pivot view directly)
For some situations though, there would be the "Totals" in the way:
if you want =MAX(ODOO.PIVOT.TABLE(1)), you don't want to have totals in the
way (with infinite ranges). We add an optional parameter 'include_total'
(default to true)
For dashboards, we want to have custom header names, we don't want to use
the headers from the pivot. To only have data and not the headers:
new parameter 'include_column_headers' (default to true)
All in all: here is what it looks like:
ODOO.PIVOT.TABLE(<pivot_id>, [row_count], [include_total], [include_column_headers])
Task: 3318865
Part-of: odoo/odoo#138594
Co-authored-by: Lucas Lefèvre <lul@odoo.com>
Everytime `getFirstPivotFunction` is used, the returned
args are evaluated.
This commit adds a getter which factorizes this behavior
that was repeated several times.
This getter is also more robust to EMPTY args AST.
Part-of: odoo/odoo#138594
The PR https://github.com/odoo/odoo/pull/104741 have recently modify how the modifiers required, readonly,
invisible, and column_invisible are defined. They are now given by Python
expressions instead of domains. In this PR we introduce two main new
components ExpressionEditor and ExpressionEditorDialog that allow to
edit/view a Python expression in a way similar to what is done in the
DomainSelector/DomainSelectorDialog. For this we have extracted the main
logic found in DomainSelector and moved it to a new component TreeEditor
that is used by both ExpressionEditor and DomainSelector. For this we had
to add a new type of leaf we call "complex condition" that allows us to
represent indecomposable (sub)expressions, i.e. expressions we cannot
(or don't want) to express as conditions of the form (path, operator, value).
The main mappings allowing to transform trees, domains, expressions into
one another (when possible) are to be found in
@web/core/tree_editor/condition_tree.js.
closesodoo/odoo#136258
Related: odoo/enterprise#48512
Signed-off-by: Michaël Mattiello (mcm) <mcm@odoo.com>
Co-authored-by: Julien Carion <juca@odoo.com>
Co-authored-by: Mathieu Duckerts-Antoine <dam@odoo.com>
Co-authored-by: Christophe Matthieu <chm@odoo.com>
Co-authored-by: Pierre Pulinckx <pipu@odoo.com>
This commit removes the default exports from the spreadsheet module.
They don't bring anything except confusion when importing a mix of
default and named exports.
closesodoo/odoo#139071
Task: 3559536
Related: odoo/enterprise#49129
Signed-off-by: Lucas Lefèvre (lul) <lul@odoo.com>
Lucas Lefèvre (lul)