Before this commit iteration in a t-foreach containing null or
undefined values crashed because they have no attributes.
closesodoo/odoo#31828
Signed-off-by: Fabien Meghazi <amigrave@users.noreply.github.com>
- Set the OS in a timezone such as the current day is different from the
day in UTC (e.g. America/Nome before 10:00 AM or Australia/Melbourne
after 3:00 PM)
- Open any datepicker
- The 'little triangle' indicating the current day is wrongly set (one
day before or after)
Knowing that Odoo always creates momentjs date and datetime with the
`UTC` flag set to `true`, the `bootstrap-datetimepicker` does something
which seems inconsistent.
First, it retrieves the `viewDate`, and sets it to the beginning of the
month and week in:
https://github.com/odoo/odoo/blob/1c6c504215f3ef09e6336c92c9d350e87599eaa1/addons/web/static/lib/bootstrap-datetimepicker/src/js/bootstrap-datetimepicker.js#L725
In this part, it is important to note that each `startOf` functions
called sets the hours/minutes/seconds to zero. It means that the
reference time is changed.
Then, it iterates on this newly created date, and determines `today` by
comparing it to `getMoment()` in:
https://github.com/odoo/odoo/blob/1c6c504215f3ef09e6336c92c9d350e87599eaa1/addons/web/static/lib/bootstrap-datetimepicker/src/js/bootstrap-datetimepicker.js#L748
However, `getMoment()` returns the current date and time, but with the
`UTC` flag set to `false`.
Therefore, we compare a UTC datetime on which the reference time has
been changed to a non-UTC datetime, which fails to give the appropriate
current day.
There are two approaches to solve this. The first possibility is to
change the way Odoo defines momentjs dates and datetimes, maybe by
removing the `UTC` flag at creation. This sounds like a bad idea, since
other widgets or views (such as the calendar or the pivot view) make use
of them. This is likely to introduce a bunch of new issues with TZ in
these views. The second approach is patching the library to fit our use.
Although we usually don't do such a thing, this allows to specifically
solve this use case, and in particular placing the 'small triangle' at
the appropriate date without impacting any other part of the system or
the library. It can be easily performed by comparing the dates and the
months to make it work.
opw-1915251
closesodoo/odoo#30538
In arabic, with the version localization of bootstrap we use the week
starts on saturday. But the version of bootstrap-datetimepicker we use
in 10.0 and saas-14 is 3.0.3 which only fully supports week starting
on sunday and monday.
Thus in arabic, the column title in the date picker would be:
Sunday <- Monday (ﻥ, ﺙ, ﺭ, ﺥ, ﺝ, ﺱ, ﺡ)
Instead of what the numbered days in the calendar are shown as:
Friday <- Saturday (ﺱ, ﺡ, ﻥ, ﺙ, ﺭ, ﺥ, ﺝ)
This issue doesn't happen in saas-15, 11.0 and over since the library
has been updated with ad6fd5f8 and the as of version 4 supports any
starting day of week, with this change:
https://github.com/Eonasdan/bootstrap-datetimepicker/commit/557b657
opw-1825640
closes#24200
Before this commit, there was an issue with input method using
composition in which typed characters were duplicated.
This was occuring until the composition ended.
This is a bug specific to Google Chrome v53+.
It has been fixed in the most recent version of ace.js [1],
and we just port the fix into our custom code.
Note that we do not simply import ace.js as a whole, because
we must update lots of custom code at the same time.
We plan to do so in the future.
Fixes#23629
[1] https://github.com/ajaxorg/ace/pull/3116
Note that we do not do a custom build anymore, since we lazyload the
files anyway. A custom build will save a few kb, but will not do any
meaningful change to the user experience. However, it dies impact
negatively our ease of maintenance, so standard build it is.
Since we are able to lazy load Qweb templates, there was
an issue to generate Awesome Timesheet iOS native app with
Cordova properly. It seems that under Cordova, iOS always
returns 0 for the status on a xml http request object.
The problem is that native applications are requesting
a local file from something that isn't a webserver.
The local file is delivered with no http headers that means no
"200 OK" header and no "404 Not Found" errors.
In this case, the request is still be completing as intended even
if the status code used by browsers is 0.
For more info, see:
https://developer.mozilla.org/fr/docs/Web/API/XMLHttpRequesthttps://bugzilla.mozilla.org/show_bug.cgi?id=331610
In particular, qunit 2.3.0 solves this issue: https://github.com/qunitjs/qunit/issues/1119
A consequence of the above issue is that some tests could have completed successfully,
although an error was raised.
Before this commit the bootstrap version we used was inconsistent.
Indeed, the javascript files were the 3.3.4 ones and the less files
were the 3.3.5 ones. This was due to a failed merged resolution in 9.0.
This should however not cause any issue as the difference between
3.3.x versions are very minor.
Basically, this commit is more of a fix than an improvement and may
even be backported if we find reasons to do so. This commit is mainly
made to prepare the transition to version 4.0.0, which requires to
switch our LESS files to SCSS first (and it will be easier to go by
bootstrap 3.3.7 with SCSS files first).
Before this fix, there was an issue when you changed the value of
a datetime field by editing direclty the input (without using the
datepicker popup).
If this value was equal to the offset of your timezone,
the new value wasn't saved. The problem was that no event was
triggered by the 'bootstrap-datetimepicker' lib because it
considered that the old datetime and the new datatime were the
same...
In odoo JS, we always manipulate UTC moment objects and we
manually add the timezone offset at rendering (we mainly do that
for testing purposes). When the offset is added, calling format()
on the moment object doesn't convert the datetime to the browser
timezone because it's flagged as being UTC.
The datepicker lib creates moment objects in the browser timezone.
So comparing the ones it creates with the one given by Odoo is
like comparing apples and oranges, and it gives wrong results
when the datetimes differ by the exact timezone offset. So in this
case, no 'datetime change' is triggered by the lib.
We could fix this by wrapping moment objects into some kind of
OdooDate class and stop manually adding the timezone offset, but
this would involve a refactoring that can only be done in master.
Instead, we prefered to patch the lib to ignore the comparison
between UTC and non-UTC objects, as this is the safest way we
find to fix this issue in stable version.
A bunch of features had been added to the qweb test files and qwebjs
did not pass them as features were either implemented differently or
not implemented at all:
* Fixed wonky and not cross-impl tests
- added ``not`` word replacemenet to qwebjs as I wasn't sure how to
handle that one cross-impl
* fixed t-esc/t-raw to render content *inside* the host node e.g.
span@t-esc should render as <span>{contents of esc}</span>
* added fallback to body content if t-esc/t-raw yields an empty result
* process t-att in order with t-att-* and t-attf-* rather than before
them both
* suppress attribute generation entirely in case of falsy value
This commit aims to improve the UX of grouped Kanban views on
mobile devices. Only one column is now displayed, full width. The
columns' records are lazy-loaded. The user can swipe horizontally
to navigate through columns (thanks to the touchSwipe library).
There were two main problems with the datepicker since the lib was
updated:
- it appeared broken in the search view "Filters" menu
- it did not overflow the parent which scrolls
Indeed, before the update, the datepicker was placed in the <body/>.
Now this is a lib option. Unfortunately, the option is broken (I don't
know why they broke it since it worked before the lib update...). As the
lib does not expose its internal functions, the only found workaround was
to patch the lib code itself.
opw-745311
opw-745625
With this commit, we introduce a benchmarking infrastructure: a new
controller, accessible at the route /web/benchmarks which will render a
new template (web.benchmark_suite).
This template uses benchmark.js and qunit.js to display a list of
benchmarking informations. For example, the number of op/s for
instantiating and destroying a list view.
I hope that this is the start of the beginning of taking the habit to
check our JS code performance sometimes, and making sure we do not have
large regression without a good reason.
Before this change the t-call can use only receive a literal as parameters:
<t t-call="template.name17" />
This call render the template 'template.name17' with the current context.
Now, we can defined an expression with braces (same syntax that t-attf):
<t t-call="template.name{{ 10 + 7 }}" />
This call render the template 'template.name17' with the current context.
This change allows for greater flexibility for applications using qweb.
(The performance due to this change is unchanged during rendering time.)
On IE (from 9.0 up to at least IE EDGE 14) we have this behavior
for the method serializeToString of XMLSerializer:
> (new XMLSerializer()).serializeToString($('<b>"</b>')[0])
'<b xmlns="http://www.w3.org/1999/xhtml">"</b>'
> (new XMLSerializer()).serializeToString($('<b>"</b>')[0].firstChild)
'"'
Whilst browser such as chromium or firefox have:
> (new XMLSerializer()).serializeToString($('<b>"</b>')[0])
'<b xmlns="http://www.w3.org/1999/xhtml">"</b>'
> (new XMLSerializer()).serializeToString($('<b>"</b>')[0].firstChild)
'"'
Hence for IE9 and over, if in a `<t t-extend/>` a `t-jquery`
sub-directive (without `t-operation`) is available, we can have
broken javascript if a " is transformed into an " at a unfortunate
location.
This commit favour node.data over XMLSerializer serializeToString to
avoid the possibility of this issue when a text node is processed.
opw-727283
This commit introduce a full redesign of all JS views. We started
basically from scratch. The goal was to unify all the various views
under a common framework, to make them testable, to make then usable in
different conditions (in studio, or in the frontend), and to make our
lives easier.
Some important points are:
- we introduced new coding guidelines (camelCase, 80 chars width, ...)
- we have a brand new testing framework (still QUnit based)
- kanban view moved to the web addon
- calendar view (formerly web_calender) moved to web as well
- the tree view was removed
- all new code should be documented
We hope that this code is the start of a new era for the Odoo web
client, we want to have a high quality codebase, well documented, well
tested, well designed.
Work done by the framework team: mostly aab, ged, chm, dmo, qsm
IE 11.0 doesn't support `Childnode.remove`:
http://caniuse.com/#feat=childnode-remove
As a workaround, we use `Node.removeChild`,
which is more widely supported.
opw-710617
Christophe Simonis