If an element has a directive whose content is loaded using `templateUrl`,
and the element is cloned using a linking function before the template arrives,
the clone needs to be updated as well.
This also updates `ngIf` and `ngRepeat` to keep the connection to the clone
of a tranclude function, so that they know about the changes a directive with
`templateUrl` does to the element in the future.
Fixes to #4930.
Stop dirty-checking during $digest after the last dirty watcher has been re-checked.
This prevents unneeded re-checking of the remaining watchers (They were already
checked in the previous iteration), bringing a substantial performance improvement
to the average case run time of $digest.
Closes#5272Closes#5287
The priority of ngInit is adjusted to occur before ngInclude, and after
ngController. This enables ngInit to initiallize values in a controller's
scope, and also to initiallize values before ngInclude executes.
Closes#5167Closes#5208
Due to an earlier change, ngModelWatch() no longer returns a value to the
caller. This means the digest loop has no way to tell if the watch actually
modified anything and so can not schedule another pass.
This means any watches that watch form or model controller changes
(e.g. watches on form.$valid) that are scheduled prior to an ngModelWatch()
will not be able to see any changes made therin.
This commit fixes this behavior by returning the latest evaluated ng-model
value.
Closes#5258Closes#5282
When using two-way binding with isolate scope, under some circumstances
the lastValue variable captured in the parentValueWatch function can get
out of sync.
Specifically, if both the value in the origin scope as well as the value
in the isolate scope get independently updated to the same value within
one digest cycle, the lastValue is never updated. This potentially causes
the watch to make the wrong decision as to which side to update on subsequent
passes.
This fixes things by ensuring lastValue is always set to the last seen
value even if the watch's logic was short circuited because there was no
difference between the values in the original and isolate scopes.
Closes#5182
The urlResolve method was fixed to automatically remove the
volume label from path names to fix issues with the file
protocol on windows where $location.path() was returning
paths where the first segment would be the volume name,
such as "/C:/mypath". See #4942 and #4928
However, the solution was specific to the $location non-
HTML5 mode, and was implemented at a lower level of
abstraction than it should have been. This refactor moves
the fix to inside of the LocationHashBangUrl $$parse method.
Closes#5041
`$sanitize` now uses the same mechanism as `$compile` to validate uris.
By this, the validation in `$sanitize` is more general and can be
configured in the same way as the one in `$compile`.
Changes
- Creates the new private service `$$sanitizeUri`.
- Moves related specs from `compileSpec.js` into `sanitizeUriSpec.js`.
- Refactors the `linky` filter to be less dependent on `$sanitize`
internal functions.
Fixes#3748.
Previously if an app was running from file:// origin we would always return either
http 200 or 404 depending on whether the response was present.
This changes the behavior so that we do this only if the protocol of the request
(not the origin) is file:// and only if the status code is 0.
Closes#4436Closes#4587Closes#4514
Although modern browser support the "input" event, they still only fire
the "change" event when they auto complete form elements
other than the currently selected one.
Related to #1460
IE8, IE9 and IE10 can use `script.onreadystate` so up till now we have been using this
if the sniffer says we are on IE.
But IE11 now does not support `script.onreadystate` and only supports the more standard
`script.onload` and `script.onerror`.
IE9 and IE10 do support `script.onload` and `script.onerror`. So now we only test whether
we are on IE8 or earlier before using `script.onreadystate`.
See http://pieisgood.org/test/script-link-events/
jQuery just uses all these handlers at once and hopes for the best, but since IE9 and IE10
support both sets of handlers, this could cause the handlers to be run more than once.
jQuery also notes that there is a potential memory leak in IE unless we remove the handlers
from the script object once they are run. So we are doing this too, now.
Closes#4523Closes#4527Closes#4922
Previously, when unwrapping promises was set to `true`,
an error would occur if a parsed expression had a
new line in it.
This was because when generating the `evaledFnGetter` code,
a new line in an parsed expression would create a new line
in a JS string in that code, which is illegal. That is:
```js
pw("A+
B")
```
Closes#4718
When a request is aborted, it makes no sense to read the response headers or text.
Also in IE9, trying to read data (either response headers or text) from an aborted request
throws an Error c00c023f.
Fixes#4913Closes#4940
When $compile interpolates a CSS class attribute expression it will
do so by comparing the CSS class value already present on the element.
This may lead to unexpected results when dealing with ngClass values being
added and removed therefore it is best that both compile and ngClass delegate
addClass/removeClass operations to the same block of code.
ngClass works by removing all the former classes and then adding all the
new classes to the element during each watch change operation. This may
cause transition animations to never render. The ngClass directive will
now only add and remove the classes that change during each watch operation.
Closes#4960Closes#4944
In 1.2, the behavior of ngInclude was modified to use DOM APIs rather than jqLite. This means that
even when jQuery was loaded, ngInclude was not calling into it, and thus scripts were not eval'd
as they had been before. Although the use of ngInclude to eval scripts as a lazy-loading strategy
was never an intentional feature, this patch restores the ability to do so.
Closes#3756
Additional API (backwards compatible)
- Injects `$transclude` (see directive controllers) as 5th argument to directive link functions.
- `$transclude` takes an optional scope as first parameter that overrides the
bound scope.
Deprecations:
- `transclude` parameter of directive compile functions (use the new parameter for link functions instead).
Refactorings:
- Don't use comment node to temporarily store controllers
- `ngIf`, `ngRepeat`, ... now all use `$transclude`
Closes#4935.
Hiding `_*` properties was a feature primarily for developers using Closure compiler and Google JS
style. We didn't realize how many people will be affected by this change.
We might introduce this feature in the future, probably under a config option, but it needs more
research and so I'm reverting the change for now.
This reverts commit 3d6a89e888.
Closes#4926Closes#4842Closes#4865Closes#4859Closes#4849
Conflicts:
src/ng/parse.js
Prior to this fix, the urlResolve method would automatically
strip the first segment of a path if the segment ends in a colon.
This was to correct undesired behavior in the $location service
using the file protocol on windows in multiple browsers (see #4680).
However, there could be cases where users intentionally
have first path segments that end in a colon
(although this conflicts with section 3.3 of rfc3986).
The solution to this problem is an extra check to make sure
the first path segment of the input url does not end with a colon,
to make sure we're only removing undesired path segments.
Fixes#4939
Chrome and other browsers on Windows often
append the drive name to the pathname,
as described in #4680. This would cause
the location service to browse to odd
URLs, such as /C:/myfile.html,
when opening apps using file://.
Fixes #4680
See doc update in the diff for more info.
BREAKING CHANGE: jqLite#scope() does not return the isolate scope on the element
that triggered directive with isolate scope. Use jqLite#isolateScope() instead.
When an isolate scope directive is also a "replace" directive and at the root of its template
it has other directives, we need to keep track remember to use isolate scope when linking
these.
This commit fixes the leakage of this state when this directive is used again later inside
or outside of the isolate directive template.
Fixes an issue when we didn't share the isolate scope with the controller
of the directive from the isolate directive's template when this directive
was replaced onto the isolate directive element.
I had to fix one unit test, as it assumed the broken behavior, where application template gets the
isolate scope of other (isolate) directive, rather than the regular scope.
BREAKING CHANGE: Child elements that are defined either in the application template or in some other
directives template do not get the isolate scope. In theory, nobody should rely on this behavior, as
it is very rare - in most cases the isolate directive has a template.
Fixes issue with isolate scope leaking all over the place into other directives on the same element.
Isolate scope is now available only to the isolate directive that requested it and its template.
A non-isolate directive should not get the isolate scope of an isolate directive on the same element,
instead they will receive the original scope (which is the parent scope of the newly created isolate scope).
Paired with Tobias.
BREAKING CHANGE: Directives without isolate scope do not get the isolate scope from an isolate directive on the same element. If your code depends on this behavior (non-isolate directive needs to access state from within the isolate scope), change the isolate directive to use scope locals to pass these explicitly.
// before
<input ng-model="$parent.value" ng-isolate>
.directive('ngIsolate', function() {
return {
scope: {},
template: '{{value}}'
};
});
// after
<input ng-model="value" ng-isolate>
.directive('ngIsolate', function() {
return {
scope: {value: '=ngModel'},
template: '{{value}}
};
});
Closes#1924Closes#2500
We need to wait until animations have added the content to the document before
trying to `autoscroll` to anchors that may have been inserted.
Fixes#4723
The msie variable is a global variable used within the ng core which contains the
version number for the current Internet Explorer browser that is rendering the
application. Other modules outside of the ng core could make use of this variable
instead of having to rollout duplicate detection code. This code makes it easy to
reuse this simple property within the $sniffer service.
Due to animations, DOM might get destroyed much later than scope and so the element $destroy event
might get fired outside of $digest, which causes changes to the validation model go unobserved
until the next digest. By deregistering on scope event, the deregistration always happens
in $digest and the form validation model changes will be observed.
Closes#4226Closes#4779
BREAKING CHANGE:
This commit introduces the notion of "private" properties (properties
whose names begin and/or end with an underscore) on the scope chain.
These properties will not be available to Angular expressions (i.e. {{
}} interpolation in templates and strings passed to `$parse`) They are
freely available to JavaScript code (as before).
Motivation
----------
Angular expressions execute in a limited context. They do not have
direct access to the global scope, Window, Document or the Function
constructor. However, they have direct access to names/properties on
the scope chain. It has been a long standing best practice to keep
sensitive APIs outside of the scope chain (in a closure or your
controller.) That's easier said that done for two reasons: (1)
JavaScript does not have a notion of private properties so if you need
someone on the scope chain for JavaScript use, you also expose it to
Angular expressions, and (2) the new "controller as" syntax that's now
in increased usage exposes the entire controller on the scope chain
greatly increaing the exposed surface. Though Angular expressions are
written and controlled by the developer, they (1) typically deal with
user input and (2) don't get the kind of test coverage that JavaScript
code would. This commit provides a way, via a naming convention, to
allow publishing/restricting properties from controllers/scopes to
Angular expressions enabling one to only expose those properties that
are actually needed by the expressions.
When using ngIf with ngInclude on the same element, ngIf previously did not remove
elements added by ngInclude. Similarly, when using ngIfStart/End, ngIf will miss
elements added between the start/end markers added after ngIf is linked.
This commit changes the behavior of ngIf to add a comment node at the end of its
elements such that elements between the starting comment and this ending comment
are removed when ngIf's predicate does not hold.
Recently we changed the priority of attribute interpolation directive to -100
to ensure that it executes early in the post linking phase. This causes issues
with when terminal directives are placed on elements with attribute bindings
because the terminal directive will usually have 0 or higher priority which
results in attr interpolation directive not being applied to the element.
To fix this issue I'm switching the priority back to 100 and making moving the
binding setup into the pre-linking function.
This means that:
- terminal directives with priority lower than 100 will not affect the attribute
binding
- if a directive wants to add or alter bindings it can do so in the pre-linking
phase, as long as the priority of this directive is more than 100
- all post-linking functions will execute after the attribute binding has been
set up
- all pre-linking functions with directive priority lower than 100 will execute
after the attribute bindings have been setup
BREAKING CHANGE: the attribute interpolation (binding) executes as a directive
with priority 100 and the binding is set up in the pre-linking phase. It used
to be that the priority was -100 in rc.2 (100 before rc.2) and that the binding
was setup in the post-linking phase.
Closes#4525Closes#4528Closes#4649
When we re-enter compilation either due to async directive templates or element transclude directive
we need to keep track of controllers to instantiate during linking.
This piece of info was missing when re-entering compilation and that's what this commit fixes.
I also reordered the properties in the previousCompileContext object.
Closes#4434Closes#4616
