Since we are doing a assert call checking a url without no page.waitForSelector
or similar calls, we add this to avoid flakes where the assert is called
when we are in process of navigation to the page. Also, we use Promise.all
here, and call page.waitForNavigation first, to avoid race conditions where
form is submitted quickly and navigation occurs and then we are stuck waiting
for a navigation. There were no flakes in the CI related to this issue, but
I expect someone with a slow hardware would probably stumble upto this in future.
We were missing the step for clicking the checkbox to show the subdomain
input field. There was a flake because of this issue in CI where the
"testsubdomain" input was getting typed into the name field. This fixes
that flake.
There were two reason there were flakes in login tests. One problem was
that the assert call for checking that the login page url was happening
too early in rare cases. To fix this issue we wait for email input field
element in the login page. The second issue was the sometimes the url was
/login/ instead of /accounts/login/. The time between the redirect were
super close here too (around ~10ms) so I am suprised this didn't occur flake
locally more often.
Server logs that show the redirect:
2020-05-26 18:55:27.065 INFO [zr] 127.0.0.1 GET 200 36ms (db: 5ms/2q) (+start: 14ms) /accounts/login/ (unauth@zulip via ?)
2020-05-26 18:55:27.074 INFO [zr] 127.0.0.1 GET 200 28ms (db: 3ms/2q) (+start: 23ms) /login/ (unauth@zulip via ?)
To take a screenshot on failure where we have our common error
handling code in common.run_test method we need to have access to
the Page instance. Currently, we create a new Page in the test method
we pass into run_test, so we pass in a new Page instance to that method
so the test and the run_test method have access to the same Page. And,
then we take a screenshot on failure. It will be saved as `failure-${num}`
in var/puppeteer directory.
This is mostly just moving code from the `activity.js`
tests, but I also now explicitly cover the "100%"
use case (i.e. all four folks in the huddle are present).
Computed indexes into these raw objects should be guarded with
Object.prototype.hasOwnProperty; make our accessors do this
automatically and use them consistently.
Signed-off-by: Anders Kaseorg <anders@zulip.com>
The previous styling was brittle and ended up breaking in very small
phone-size views with the text overflowing the boundaries of the page.
The right fix is to move those heading outside the portico-header
class, since the CSS for that isn't generally appropriate here.
This is a prep commit which combines the previous `#searchbox`
block with the newly updated `#searchbox_legacy` block which
contains the modifications related to the new navbar display.
This only consists of changes to `#searchbox` and is still broken.
But it integrates the searchbox with the new tab_bar changes so that
only one searchbox is shown (instead of two, previously).
This is helpful because if the user pastes multiple queries in the
searchbox and there are invalid search operators, then it is visible
through the typeahead.
The main reasoning for this change is as follows:
* When the search bar contains multiple search queries
but no search results, the last search operand does
not get displayed.
This happens due to the fact that filter object
contained 2 terms having the operator key value as
"search" instead of a single term where operator is
"search" and operand is a single string containing
the space seperated search queries. This condition
occurs for search_pills_enabled case only because
we used to Filter.parse the query twice
(once for the `base_operators` and once for the
`suggestion_operator instead of doing both at once).
Thus the `search_query` value inside the
`narrow.show_search_query` function which only
selected the operands of the first term displayed
an incomplete result.
* Another benefit of this commit is to display the narrow
operators in the URL fragment the same way as when
search_pills_enabled = False.
For example, On entering the queries in the mentioned
order -> 'is: starred', 'abc', 'def', 'is: private',
'ghi'. This is the URL:
Previously:
/#narrow/is/starred/is/private/search/abc.20def/search/ghi
Now (same as pills disabled case):
/#narrow/is/starred/is/private/search/abc.20def.20ghi
* We are also able to de-duplicate the non-typeahead search
query code path.
As mentioned in the comment for `KEY.BACKSPACE` event
in `input_pills.js`, we do normal character deletion
if there is input present. However this wasn't the case
if spaces were present. Also the input wasn't cleared
after the last pill was removed.
Thus `trim()` is removed from the input length check and
the new pill is still created from the trimmed value.
We can remove the typeahead by clicking outside the search box
after we have entered the search string to be filtered and then
focus on the searchbox and press enter or just by pressing enter
on an empty string.
Previously, the narrow would just deactivate for the above condition
as the searchbox value which was passed as the raw_operators parameter
to the narrow.activate function was empty.
This happened because we called the activate function on pressing
enter for the keyup event, while the keydown event in the parent
container made a pill from the text and cleared the input. (as
mentioned in the comment for `KEY.ENTER` case in `input_pill.js`)
The /apps page webapp link now takes the user to /accounts/go to find
their organization's login page, rather than failing to do anything.
Fixes#14977.
Commit 9b78a73e36 (#15005) made some of
our poorly written Casper tests fail. Now they’re just as poorly
written but passing again.
Signed-off-by: Anders Kaseorg <anders@zulip.com>
On invitations panel, invites were being removed when
the user clicked on invitation's link. Now we only remove
it when the user completes registration.
Fixes: #12281
The people.js tests were using _add_user function to add
cross realm bots. The problem is that _add_user function
doesn't properly simulates the adding process as it doesn't
add the user in cross_realm_dict as well.
To solve this and eliminate the need of calling
people.initialize(), which means the params obj needs to be
defined, we extracted the whole logic of adding a cross realm
user into a separete function, add_cross_realm_user.
This fixes some issues with unclear terminology and visual styling in
the pages for the new free trial.
There's probably more we can and should usefully do in the future.
This makes it so that search_suggestion.js
does not depend on activity.js.
That dependency hasn't really been "elegant"
for quite some time, but it will become particularly
unnecessary when we go to remove the "Group PMs"
section from the right sidebar.
This commit introduces a temporary wart
where we have these two functions with the
same name in a sort of unnecessarily
complicated code stack:
activity.process_loaded_messages
huddle_data.process_loaded_messages
But we will eliminate the former function
very soon, and our message-related codepaths
will just call the `huddle_data` version
directly.
TESTING NOTES:
Now that `huddle_data` is a tiny leaf
module, it's super easy to just use the
real implementation of what was formerly
called `activity.get_huddles()` (and is
now in `huddle_data`).
When I first wrote this commit, introducing
the real implementation of `get_huddles` exposed
some bugs that I fixed in the immediately
prior commits to this.
When the tests were originally written,
I believe `activity.js` had some annoying
`jQuery` dependencies that made it hard
to unit test against. We've slimmed it over
time to be mostly just a "controller" module.
But even in its current state it would have
been a bit of a bloated dependency.
The other friction for using the actual
version of `get_huddles` was setting up
the message data, but that's pretty minor.
If you have a group PM where some users have
three-digit user_ids and some with four-digit
user_ids (or similar), a huddle could effectively
be ignored when determining the order of
search search suggestions.
Basically, we need a way to canonically sort
user_ids in "huddle" strings, and it's somewhat
arbitrary whether you sort lexically or sort
numerically, but you do need to be consistent
about it.
And JS is not exactly helpful here:
> [99, 101].sort()
[ 101, 99 ]
This is a pretty obscure bug with pretty low
user-facing consequences, and it was never
reported to us as far as I know, but the fix
here is pretty straightforward.
We have had similar bugs of slightly more consequence
in the past. The reason this bug has shown
up multiple times in our codebase is that every
component that deals with huddles has slightly
different forces that determine how it wants
to serialize the huddle. It's just one of those
annoying things. Plus, bugs with group PMs
do tend to escape detection, since most people
spend most of their time either on streams
or in 1:1 PMs.
This is a pure code extraction. The current
code is buggy with respect to user_ids with
different lengths of digits, i.e. it does
a naive lexical sort instead of a numerical
sort. We'll fix that in the next commit.
orientor
Priyank Patel
Steve Howell
Anders Kaseorg
whoodes
Tim Abbott
Ryan Rehman
sahil9001
arunikaydav42
clarammdantas
Vishnu KS