This change brings Angular's JSONP behaviour closer in line with jQuery's. It will no longer treat
a callback called with no data as an error, and will no longer support IE8 via the onreadystatechanged
event.
BREAKING CHANGE:
Previously, the JSONP backend code would support IE8 by relying on the readystatechanged events. This
is no longer the case, as these events do not provide adequate useful information for deeming whether
or not a response is an error.
Previously, a JSONP response which did not pass data into the callback would be given a status of -2,
and treated as an error. Now, this situation will instead be given a status of 200, despite the lack
of data. This is useful for interaction with certain APIs.
Previously, the onload and onerror callbacks were added to the JSONP script tag. These have been
replaced with jQuery events, in order to gain access to the event object. This means that it is now
difficult to test if the callbacks are registered or not. This is possible with jQuery, using the
$.data("events") method, however it is currently impossible with jqLite. This is not expected to
break applications.
Closes#4987Closes#6735
In order to make the behavior compatible with $rootScope.$watch and $rootScope.$on methods, and
make it possible to deregister an attribute observer, Attributes.$observe method now returns a
deregistration function instead of the observer itself.
BREAKING CHANGE: calling attr.$observe no longer returns the observer function, but a
deregistration function instead.
To migrate the code follow the example below:
Before:
```
directive('directiveName', function() {
return {
link: function(scope, elm, attr) {
var observer = attr.$observe('someAttr', function(value) {
console.log(value);
});
}
};
});
```
After:
```
directive('directiveName', function() {
return {
link: function(scope, elm, attr) {
var observer = function(value) {
console.log(value);
};
attr.$observe('someAttr', observer);
}
};
});
```
Closes#5609
Originally we destroyed the oldValue by incrementaly copying over portions of the newValue
into the oldValue during dirty-checking, this resulted in oldValue to be equal to newValue
by the time we called the watchCollection listener.
The fix creates a copy of the newValue each time a change is detected and then uses that
copy *the next time* a change is detected.
To make `$watchCollection` behave the same way as `$watch`, during the first iteration
the listener is called with newValue and oldValue being identical.
Since many of the corner-cases are already covered by existing tests, I refactored the
test logging to include oldValue and made the tests more readable.
Closes#2621Closes#5661Closes#5688Closes#6736
PR #5547 introduced conversion of all 0 status codes to 404 for cases
where no response was recieved (previously this was done for the
file:// protocol only). But this mechanism is too eager and
masks legitimate cases where status 0 should be returned. This commits
reverts to the previous mechanism of handling 0 status code for the
file:// protocol (converting 0 to 404) while retaining the returned
status code 0 for all the protocols other than file://
Fixes#6074Fixes#6155
On older browser that don't support the new HTML5 inputs
and display a text input instead, the user is required to enter
the data in the corresponding ISO format. The value in `ng-model`
will always be a date.
E2e tests contain a workaround to a bug in webdriver,
see https://github.com/angular/protractor/issues/562.
Also adds weeks as format to the `dateFilter`.
Related to #757.
Closes#5864.
Add css animations when form or field status change to/from dirty,
pristine, valid or invalid. This works like animation system present
with ngClass, ngShow, etc.
Closes#5378
If enter -> leave -> enter -> leave occurs then the first leave animation will
animate alongside the second. This causes the very first DOM node (the view in ngView
for example) to animate at the same time as the most recent DOM node which ends
up being an undesired effect. This fix takes care of this issue.
Closes#5886
This reverts commit cad717b117.
This change causes regressions in existing code and after closer inspection
I realized that it is trying to fix an issue that is should not be considered
a valid issue.
The location service was designed to work against either "hash" part of the
window.location when in the hashbang mode or full url when in the html5 mode.
This change tries to merge the two modes partially, which is not right. One
reason for this is that the search part of window.location can't be modified
while in the hashbang mode (a browser limitation), so with this change part
of the search object should be immutable and read-only which will only cause
more confusion.
Relates to #5964
Per ECMAScript 5.1 specification trailing commas are allowed in object and
array literals. All modern browsers as well as IE>8 support this syntax.
This commit adds support for such syntax to Angular expressions.
Before this fix, search queries in hashbang mode were ignored if the hash was not present in the
url. This patch corrects this by ensuring that the search query is available to be parsed by
urlResolve when the hashbang is not present.
Closes#5964
There are always going to be false positives here, unfortunately. But
testing different properties will hopefully reduce the number of false
positives in a meaningful way, without harming performance too much.
Closes#4805Closes#5675
This change makes the ngHref directive useful for SVGAElements by having it bind
to the xlink:href attribute rather than the href attribute.
Closes#5904
The previous code for filtering out non-finite numbers was broken, as it would convert `null` to `0`,
as well as arrays.
This change fixes this by converting null/undefined/NaN/Infinity/any object to the empty string.
Closes#6188Closes#6261
BREAKING CHANGE: ngClass and {{ class }} will now call the `setClass`
animation callback instead of addClass / removeClass when both a
addClass/removeClass operation is being executed on the element during the animation.
Please include the setClass animation callback as well as addClass and removeClass within
your JS animations to work with ngClass and {{ class }} directives.
Closes#6019
If the first element in a template is a <tr>, <th>, <td>, or <tbody> tag,
the HTML compiler will ensure that the template is wrapped in a <table>
element so that the table content is not discarded.
Closes#2848Closes#1459Closes#3647Closes#3241
Because of a4e6d962, model is not updated on input/change between the
compositionstart and compositionend events. Unfortunately, the compositionend
event does not always happen prior to an input/change event.
This changeset calls the listener function to update the model after a
compositionend event is received.
Closes#6058Closes#5433
Due to 339a165, it became impossible to filter nested properties of an object using the filterFilter.
A proposed solution to this was to enable the use of nested predicate objects. This change enables the
use of these nested predicate objects.
Example:
```html
<div ng-repeat="it in items | filter:{ address: { country: 'Canuckistan'}}"></div>
```
Or
```js
$filter('filter')(items, { address: { country: 'Canuckistan' } });
```
Closes#6215
Related to #6009
This corrects a complicated compiler issue, described in detail below:
Previously, if an element transclusion directive contained an asynchronous directive whose template
contained another element transclusion directive, the inner element transclusion directive would be
linked with the element, rather than the expected comment node.
An example manifestation of this bug would look like so:
```html
<div ng-repeat="i in [1,2,3,4,5]">
<div my-directive>
</div>
</div>
```
`my-directive` would be a replace directive, and its template would contain another element
transclusion directive, like so:
```html
<div ng-if="true">{{i}}</div>
```
ngIf would be linked with this template content, rather than the comment node, and the template element
would be attached to the DOM, rather than the comment. As a result, this caused ng-if to duplicate the
template when its expression evaluated to true.
Closes#6006Closes#6101
The flushNext method of testing is difficult and highly coupled with the behavior
of ngAnimate's $animate workflow. It is much better instead to just queue all
$animate animation calls into a queue collection which is available on the $animate
service when mock.animate is included as a module within test code.
Previously, classes added to asynchronous directive elements during the clone
attach function would not persist after the node is merged with the template, prior
to linking. This change corrects this behaviour and brings it in line with synchronous
directives.
Closes#5439Closes#5617
Before this change, an SVGAElement with an xlink:href attribute and no href or name attribute which
was compiled by the angular HTML compiler would never be clickable, due to the htmlAnchorDirective
calling event.preventDefault() due to the missing href attribute.
This change corrects this behaviour by also testing the xlink:href attribute if the element in
question is determined to be an SVG anchor tag (with the href property having type SVGAnimatedString)
Closes#5896Closes#5897
This change uses the regexp from Chromium/Blink to validate emails, and corrects
an error in the validation engine, which previously considered an invalid email
to be valid. Additionally, the regexp was invalidating emails with capital
letters, however this is not the behaviour recomended in the spec, or implemented
in Chromium.
Closes#5899Closes#5924
An infinite $digest loop can be caused by expressions that invoke a promise.
The problem is that $digest does not decrement ttl unless it finds dirty changes;
it should check also if asyncQueue is empty.
Generally the condition for decrementing ttl should be the same as the
condition for terminating the $digest loop.
Fixes#2622
This patch allows the ng-options value of a <select> element to span
multiple lines, which would previously throw an error when used with filters.
Closes#5602
This reverts commit 63cd873fef.
The change breaks existing tests of Google apps. The problem is that
while we tried to avoid adding #/ to window.location.href unnecessarily
we failed doing so. Likely because by setting $path, at some point
(during a digest) we try to check if $location changed and we mistake the
default '/' with an explicit settign of the path via the `path()` method.
This results in us writing the url with '#/' into $browser.url() which updates
the window.location by adding "#/" to the url - something we tried to avoid
in the first place.
I'll reopen PR #5712.
Before this change, on the root of the application, $location.path() would return
the empty string. Following this change, it will always return a root of '/'.
Closes#5650Closes#5712
