Some parts of the file list contents (file name and actions) had a
higher z-index than the file list multiselect header. That header is
fixed in place, so when the file list contents were scrolled and those
parts with a higher z-index overlapped the multiselect header they were
fully visible. Now the z-index for the multiselect header has a higher
value (the same used in the controls bar) to ensure that the contents
are shown behind the header.
Fixes#7540
Signed-off-by: Daniel Calviño Sánchez <danxuliu@gmail.com>
Fixes#7184
The SyncJob can be very resource intensive. Since it requests all users
on the system to create the system addressbook. In order to do this it
creates a vcard for every user and updates the addressbook.
There is no need for this job since the proper signals are emitted and
handled in the carddav backend to update the addressbook live.
Worst comes to worst there is always the occ command to bring the
address book in sync again.
Signed-off-by: Roeland Jago Douma <roeland@famdouma.nl>
When the checkbox was moved to where the favourite icon was shown before
the layout of the file list was modified. The checkbox is no longer a
descendant of the ".filename" element, so it is no longer removed by the
"External storages" file list.
However, even before the checkbox was moved, explicitly removing it was
not the best approach, as file list rows could still be selected using
"Ctrl/Shift+click". This did not provide much value, as the selection
header has no actions; it simply states the number of selected elements.
The proper way to disable the selection is by setting "_allowSelection"
to false in the file list instead.
Signed-off-by: Daniel Calviño Sánchez <danxuliu@gmail.com>
When the checkbox was moved to where the favourite icon was shown before
the layout of the file list was modified. The first column is no longer
the file name, so neither the thumbnail nor the name link were found.
Due to this the thumbnail was not set to the appropriate icon, and the
dummy event handler was not removed from the name link, so clicks on the
name were basically ignored. Now the selectors are based on the
".filename" CSS class instead of relying on the column position.
Signed-off-by: Daniel Calviño Sánchez <danxuliu@gmail.com>
When an empty area of a file row was clicked and the "Details" action
was executed "fileActions.currentFile" was not guaranteed to be set to
the appropriate object (it depended on the previous actions of the
user), so when it was used by "getCurrentMimeType()" and other
FileActions functions they may not work as expected. Now it is
explicitly set to the appropriate value before its use.
Signed-off-by: Daniel Calviño Sánchez <danxuliu@gmail.com>
The share menu toggle and some share menu items included an 'href="#"'
attribute, so they were handled as internal links by the browser, which
changed the current anchor when they were clicked. However, there was no
real need to change the anchor in those cases, and it could interfere
with other apps (for example, the PDF viewer sets the current anchor to
"#pdfviewer" when it is shown and it hides itself when that anchor is
modified). According to the HTML 5 spec the "href" attribute is not
mandatory for "a" elements, so they were removed.
Other options would have been to change the elements from "a" to "div"
or something like that, but that would have required changes to the CSS
rules too, or to prevent the default event handling for those elements
through JavaScript, which would have been a workaround instead of the
proper solution.
Signed-off-by: Daniel Calviño Sánchez <danxuliu@gmail.com>
When a "mouseup" event was triggered on any element except on the share
menu or its descendants the share menu was closed. The share menu toggle
is not a descendant of the share menu, so clicking on it when the share
menu was shown closed it by removing its "open" CSS class. However, once
that happened the click event was then handled by the share menu toggle,
which toggled the "open" CSS class in the share menu and thus added it
again. So, from the user point of view, nothing happened when clicking
on the share menu toggle if the share menu was open.
Now a "mouseup" event on the share menu toggle no longer closes the
share menu, and thus toggling the "open" CSS class when handling the
"click" event works as expected.
Signed-off-by: Daniel Calviño Sánchez <danxuliu@gmail.com>
Clicking on an empty space in a file row causes the details view to be
shown. As it is a user initiated action on the file list now it is done
by triggering the Details action instead of directly calling
"_updateDetailsView"; the result is the same in both cases, but using
the action is more consistent (clicking on the file name triggers the
default action, and clicking on the inline actions triggers those
actions) and also makes possible to use the "beforeTriggerAction" and
"afterTriggerAction" listeners.
Signed-off-by: Daniel Calviño Sánchez <danxuliu@gmail.com>
In the same way that other elements can know when a FileAction is
registered or a default action is set this commit makes possible to be
notified before and after a FileAction is executed.
This is achieved by wrapping the registered action handler in a custom
function that notifies the listeners before and after executing the
handler itself. Due to this approach only FileActions registered through
"registerAction" trigger the events, although that is not a problem as
this is how the actions should be added to the FileActions anyway.
Signed-off-by: Daniel Calviño Sánchez <danxuliu@gmail.com>
When a file from the file list is dragged a drag shadow (a copy of the
file row that follows the cursor position) is created. The drag shadow
element is created as a direct child of the body element, so it needs a
higher "z-index" than the one used for the file list to be visible.
In narrow screens the "#app-content" uses a "z-index" of 1000 in order
to be visible over the "#navigation-bar" when they overlap, so the
"z-index" of the drag shadow must be at least 1000 to be visible over
the file list.
Instead of updating the hardcoded "z-index" it was removed and replaced
by CSS rules for ".dragshadow" elements to ease theming.
Signed-off-by: Daniel Calviño Sánchez <danxuliu@gmail.com>
* Cache it for a day so we will retry eventually
* Cache the status.php response as well so we will try it once a day as
well
Signed-off-by: Roeland Jago Douma <roeland@famdouma.nl>
Add a "searchPrincipals" function to the NC principal backend.
Fix the "findByUri" function to respect the prefixPath.
Signed-off-by: Christoph Seitz <christoph.seitz@posteo.de>
Case: email is set to null, but the avatar is set. In the old case the
email would set $emptyValue but $noImage would still be false. This we
would set the empty string as email.
Signed-off-by: Roeland Jago Douma <roeland@famdouma.nl>
When a constructor is spied using Sinon it is wrapped by a proxy
function, which calls the original constructor when invoked. When "new
Foo()" is executed a "Foo" object is created, "Foo" is invoked with the
object as "this", and the object is returned as the result of the whole
"new" expression.
Before Sinon 4.1.3 the proxy called the original constructor directly
using the "thisValue" of the spied call; "thisValue" was the object
created by the "new" operator that called the proxy. The proxy assigned
"thisValue" to "returnValue", so it was also the value returned by the
proxy and, in turn, the value returned by the whole "new" expression.
Since Sinon 4.1.3 (see pull request 1626) the proxy calls the original
constructor using "new" instead of directly. The "thisValue" created by
the outermost "new" (the one that called the proxy) is no longer used by
the original constructor; the internal "new" creates a new object, which
is the one passed to the original constructor and returned by the
internal "new" expression. This object is also the value returned by the
proxy ("returnValue") and, in turn, the value returned by the whole
outermost "new" expression.
Thus, now "returnValue" should be used instead of "thisValue" to get the
object created by the spied constructor.
Signed-off-by: Daniel Calviño Sánchez <danxuliu@gmail.com>
Fixes#7158. Adds a # on the color if missing.
Increased maxlength, added hash:true for jscolor, and adding a # if not present on the change event.
Since the input element now allows a hex code, changed values to hexcode.
In addition, added a function to get RGB array from hex or rgb values. Calling it in both methods and using it to perform comparison.
Also changed the way we were determining whether the jscolor component had loaded.
Changed the control to use data-jscolor rather than defining opts in the class.
Signed-off-by: Abijeet <abijeetpatro@gmail.com>
When the autocomplete popover is shown the At.js plugin listens on the
message input field for key down events, and when Enter is pressed it
adds the selected item to the message. However, as "_onTypeComment" also
handles key down events for the message input field, when Enter was
pressed the comment was submitted and At.js had no chance to add the
item before that happened. Now when Enter is pressed and the
autocomplete popover is shown the comment is not submitted, and thus
At.js adds the selected item to the message as expected.
Signed-off-by: Daniel Calviño Sánchez <danxuliu@gmail.com>
We have to double check. Since getting the info of the root returns a
generic entry. But actually the stroage is not available. Else we get
very weird sync and web behavior.
Signed-off-by: Roeland Jago Douma <roeland@famdouma.nl>
Roeland Jago Douma
Morris Jobke
Daniel Calviño Sánchez
Arthur Schiwon
Roeland Jago Douma
John Molakvoæ (skjnldsv)
Marin Treselj
Robin Appelman
Vincent Petry
Thomas Müller
Bjoern Schiessle
Tobia De Koninck
Thomas Citharel
Georg Ehrke
Christoph Seitz
Nextcloud bot
Lukas Reschke
Abijeet
Tobias Kaminsky