- according to previous change(commit) we need to adjust ID in cypress test files
Signed-off-by: Darshan-upadhyay1110 <darshan.upadhyay@collabora.com>
Change-Id: Ic5510392b101ff7c7daa28719a60963c7bb2c743
- Because we have appended '-input' to add ui-edit-control in jsDialog builder filer
- now we do not need any id/string which have '-input' at very end of that element
- for some cases we need correct GetElement by ID in js files
Signed-off-by: Darshan-upadhyay1110 <darshan.upadhyay@collabora.com>
Change-Id: Id9b4f53e8deaa2f26c6abc93dc80504895107bd0
- To unify editControl, added additional "parent" for edit's input
- this will be consistent with other input type controls
- and will help to establish link between lable and input[type='edit']
- changed parameter from input to it's input container which will help to pass correct ID to Core
Signed-off-by: Darshan-upadhyay1110 <darshan.upadhyay@collabora.com>
Change-Id: If5ab78bdf3d3ad61438898f5e4cacc19bd2ecb63
in 24.04 more elements are grid now, it seems to be a
regression in button box:
1. enable macro execution in coolwsd.xml
2. open spreadsheet with macro
Result: you see the dialog with buttons in 2 rows
Expected: all buttons are in the same row
it seems to be missing grid placement for left and right sections
Signed-off-by: Szymon Kłos <szymon.klos@collabora.com>
Change-Id: Ic1a3479114e4f832e852b9bdf3d39d2669ae69bb
problem:
when user selected disable macro in the opening dialog
document used to close and return to integrator
but now cancel button in csv import dialog is hidden,
so no need to distinguish there and we can continue to load without macros
In addition we still close the file if user clicks "X" button on title bar
Signed-off-by: Pranam Lashkari <lpranam@collabora.com>
Change-Id: Ie0c7d77954e099cb3c8a72c400c688b3ae1343c7
Previously when there was a frozen pane we calculated an incorrect
size on the unfrozen pane, leading to slight blank edges when you were
mid-zoom-out.
Signed-off-by: Skyler Grey <skyler.grey@collabora.com>
Change-Id: Id26f6ba815429addd0cdf3af37f5c54363a08b2b
Previously when panning with zoom, the freeze shadow would stay in its
initial state (either hidden or shown) whether or not it should actually
be shown at the current position.
Signed-off-by: Skyler Grey <skyler.grey@collabora.com>
Change-Id: Ic0d05c1204b1e15abbb6d8ad34235206fdc7740c
Previously the freeze shadow would stay where it was, creating a hard
edge and confusion about where was frozen. With this change, the freeze
shadow moves with the zoom.
This change needs a followup to show/hide the shadow when we reach/move
away from the document edge.
Signed-off-by: Skyler Grey <skyler.grey@collabora.com>
Change-Id: I8bec41be6ce5eff3c1bfd3ac0234b6d53833614b
- Fix alignment between rows by:
- Setting some of the columns' width
- Removing margin from all buttons & set width. Add column gap
- Fix vertical alignment
- Fix header alignment (Name, Type, Value) so it aligns to the content
- Make "Add property" occupy full width (ideally this button would be
part of the inner grid "last row" but for now this is already an
improvement)
- Remove lengthly css targets and the use of nth-child
Signed-off-by: Pedro Pinto Silva <pedro.silva@collabora.com>
Change-Id: I85d9bdbcc365c735757ee086f6e4912d694a011c
coolforkit: common/Log.cpp:677: void Log::shutdown(): Assertion
`ThreadLocalBufferCount <= 1 && "Unstopped threads may have unflushed
buffered log entries"' failed.
Was caused by mis-accounting. Also join document threads on 'exit'
which happens during unit tests to ensure we don't hit the above.
Signed-off-by: Michael Meeks <michael.meeks@collabora.com>
Change-Id: I523e723e54e4252ae0d65de36af086e97dd79f04
Otherwise forkit tends to loop aggressively re-starting that for the
duration of the test wastefully.
Signed-off-by: Michael Meeks <michael.meeks@collabora.com>
Change-Id: Ia1c684a5d995f54f29290c9631b1ee14266445d7
Simple example to re-instate previous unit test.
Signed-off-by: Michael Meeks <michael.meeks@collabora.com>
Change-Id: I26da1178bc897797a656eb5ae9f838d17cbaf75f
This added 200+ lines of code, made things harder to
understand, and was used only in a single test that can be
covered in another way.
Signed-off-by: Michael Meeks <michael.meeks@collabora.com>
Change-Id: I4ed8d1d52d533f8b24be7dd5e12dbb7702ef1a80
It sets a reply's status to "resolved" only if its parent is resolved.
I can't remember why we implemented this like this before.
But now, we don't need to.
Signed-off-by: Gökay Şatır <gokaysatir@collabora.com>
Change-Id: Ia31dab0237d3b89e5d391cf87394514d18b595f7
Copy an image to the clipboard, click Paste -> Paste special on the
notebookbar in Firefox: the image is pasted, but the paste special
dialog doesn't close.
What happens here is that Firefox lacks navigator.clipboard.read(), so
the relevant code is Clipboard.js dataTransferToDocumentFallback(), not
the new _navigatorClipboardGetTypeCallback(). Here e.g. copy from
desktop Writer (html+text) would lead to a _doInternalPaste(), which
would call _checkAndDisablePasteSpecial() for us, but the image case
returns early, so the closing doesn't happen.
Fix the problem by explicitly calling _checkAndDisablePasteSpecial()
before returning early. In contrast with the other uses of
_checkAndDisablePasteSpecial(), we don't need to call the paste or paste
special uno command (since _asyncReadPasteImage() takes care of pasting
with a websocket comment), so it's fine to ignore the return value.
No testcase, our cypress tests run on Chrome, and this happens only with
Firefox.
Signed-off-by: Miklos Vajna <vmiklos@collabora.com>
Change-Id: I751d234edfdf9e8fb37f2bf2d6b3f3d8e65a27f4
Some tests are designed to exercise the reconnection
logic. Once the test validates that the correct
reconnection logic is used, the test is considered
finished. In those cases, we need to break the
test framework's retry logic.
Change-Id: Ie0bb5dcc430a954910c2af4817a500f5b576bd4c
Signed-off-by: Ashod Nakashian <ashod.nakashian@collabora.co.uk>
It's only relevant for regular URLs that contain version hash.
Signed-off-by: Aron Budea <aron.budea@collabora.com>
Change-Id: I73afaa0559c3dddc09a8ed5f786fb5126e1c8f0c
Failed because mobile wizard opened too soon.
Fix: Wait for selection before opening mobile wizard.
Signed-off-by: Neil Guertin <neil.guertin@collabora.com>
Change-Id: Ie29e3f5be17bafa2dd09b3f8f5b6415104456cb7
Most users of selectTextOfShape also need to select all text.
Sometimes the cursor appears but does not allow CTRL+a to select
all text. Work around this by including the selection in the retry
using waitUntil.
Fixes mobile/impress/apply_paragraph_props_text_spec.js, which started
failing sporadically after commit e0a0646ce1.
I also adjusted the dblclick selector because it was finding multiple
elements but I think that is unrelated
I also created a separate function editTextInShape, for tests
that do not need to select all text
Signed-off-by: Neil Guertin <neil.guertin@collabora.com>
Change-Id: Icdfa731ba9a5f8d08a858192a4f1f5bc79b09b76
insert shape now shows all the items in one column.
fix it so we have rows
Signed-off-by: Szymon Kłos <szymon.klos@collabora.com>
Change-Id: I371f7fa233e46a62dbb6637d934bc5ceb5cdecb0
problem:
when you try to reply to long comment, it would be resized
and subsequntly reply box will take space and other comments
will not make enough space for it
Signed-off-by: Pranam Lashkari <lpranam@collabora.com>
Change-Id: I4d81904721e5bb196947598d0dca211ee9d480a6
problem:
when comment was on the edge of the document/screen(vertically)
resize was called which ultimately collapsed the comments in small screen
which could not be expanded unless scrolled further from edge.
It was more problematic for the threads,
if entire thread could not fit in the screen it was always failing to expand
regression from bac4161
it nullifies the effect from 81b00e7
now if comment is added at the end of the document and it goes beyond doc boundry
doc will not scroll automatically to make entire comment visible
Signed-off-by: Pranam Lashkari <lpranam@collabora.com>
Change-Id: I198b4887ae19ca0c17ba1453960cd36a66beb04d
problem:
when long thread did not fit on screen and actions were performed
on last comments of thread which put root comment out of view,
it will cause scroll back to root comment (i.e: reply)
Signed-off-by: Pranam Lashkari <lpranam@collabora.com>
Change-Id: I23d699425dc2dff1cfeddf6ab9be7131f80c3a31
select function now can take care of reselection and layout update
Signed-off-by: Pranam Lashkari <lpranam@collabora.com>
Change-Id: I2a061643ec190d5b2d005b940923a28131ed6cb0
which may kick in at an arbitrary point. We shouldn't invalidate if
there is no change, but that's a TODO.
Signed-off-by: Caolán McNamara <caolan.mcnamara@collabora.com>
Change-Id: I910c96f2c2a0efab393ccea52c9f25d5804cd967
This is not part of CI at the moment, so I didn't notice it till Szymon
pointed it out.
make -C cypress_test check-idle spec=calc/idle_spec.js
passes with this, which is the only suite in 'make check-idle'.
Signed-off-by: Miklos Vajna <vmiklos@collabora.com>
Change-Id: Ieeb15653abcc016bc7de87115df136dfe3f27d0d
Copy an image in Firefox, paste in Chrome with the Paste button or with
Ctrl-V. Nothing happens in Calc, Writer gives you a popup to copy the
image, which is already the case.
What happens is that Clipboard.js already had some code to guess when
the HTML only contains an image, but while Chrome puts '<img ...' to the
clipboard, Firefox puts '<meta ...><img ...', so we don't recognize the
content as image.
Fix the problem by adding a new isHtmlImage() function to Clipboard.js
that handles both markups and use it from both places that deal with
paste (Paste button, keyboard).
Note that this manual parsing is needed, because the input is user
provided, so DOMParser or innerHTML would not be safe.
Signed-off-by: Miklos Vajna <vmiklos@collabora.com>
Change-Id: Ia1340cd3777c5597e3fd257e905e9cc637802bcc
New helpers setupDocument, loadDocument, reloadDocument
Clean up beforeEach functions
Replace most calls to beforeAll with setupAndLoadDocument
Replace calls to reload with reloadDocument
Signed-off-by: Neil Guertin <neil.guertin@collabora.com>
Change-Id: I12307f93c8a418e6f55bb36b8afe340d7841b726
Replace beforeAll in most tests with new helper function
Tests that need to close or reopen test files cannot use
the new function and will be updated later
Signed-off-by: Neil Guertin <neil.guertin@collabora.com>
Change-Id: I372280e55993123fc26f83cb2e6e6e78665a6846
Cypress recommends NOT cleaning up after tests
https://docs.cypress.io/guides/references/best-practices#Using-after-Or-afterEach-Hooks
Open documents should not affect other tests
Files will be closed automatically by the server anyways
Tests with more strict requirements can explicitly close
documents before, during, or after those tests
This is necessary to support beforeAll cleanup
Half of tests were verifying the wrong filename was closed anyways
Signed-off-by: Neil Guertin <neil.guertin@collabora.com>
Change-Id: Iab458963a3fbfe87f732dc97140704c2d0ef53d4
