previously we were doing all kinds of checks to see if we should rewrite the url or not and we
were missing many scenarios. not any more.
with this change, we rewrite the url unless:
- the href is not set
- link has target attribute
- the absolute url of the link doesn't match the absolute prefix for all urls in our app
This also means that ng-ext-link attribute which we previously used to distinguish external
links from app links is not necessary any more. apps can just set target=_self to prevent
rewriting.
BREAKING CHANGE: ng-ext-link directive was removed because it's unnecessary
apps that relied on ng-ext-link should simply replace it with target=_self
previously it would create a new instance which wasn't configured as the one in the app,
which resulted in incorrect values being returned in html5 mode with base url set
Often it is impossible to set the http defaults during the config phase,
because the config info is not available at this time.
A good example is authentication - often the app needs to bootstrap,
allow user to enter credentials and only then it gains access to
session token which then should be sent to the server with every request.
Without having the ability to set the defaults at runtime, the developer
either has to resort to hacks, or has to set the session token header
with every request made by the app.
Create build for other modules as well (ngResource, ngCookies):
- wrap into a function
- add license
- add version
Breaks `$sanitize` service, `ngBindHtml` directive and `linky` filter were moved to the `ngSanitize` module. Apps that depend on any of these will need to load `angular-sanitize.js` and include `ngSanitize` in their dependency list: `var myApp = angular.module('myApp', ['ngSanitize']);`
this was never meant to be a public api used by apps. I refactored
the code to hide the functionality.
BREAKING CHANGE: $browser.addJs method was removed
apps that depended on this functionality should either use many of the
existing script loaders or create a simple helper method specific to the
app.
this api was never supposed to be public. nobody should be relying
on it.
I'm removing it since angular doesn't need it.
BREAKING CHANGE: $browser.addCss was removed
apps the depend on this functionality should write a simple utility
function specific to the app (see this diff for hints).
closure compiler is stubborn and puts the flag to the top of the file, so
we have to post-process the minified file to move the flag into the angular
closure.
Previously one had to write:
$routeProvider.when('/foo', {...});
$routeProvider.when('/bar', {...});
$routeProvider.otherwise({...});
After this change it's just:
$routeProvider.
when('/foo', {...}).
when('/bar', {...}).
otherwise({...});
Breaks #when which used to return the route definition object but now
returns self. Returning the route definition object is not very useful
so its likely that nobody ever used it.
The purpose of allowing the scope to be specified was to enable the $route service to work
together with ngInclude. However the functionality of creating scopes was in the recent past
moved from the $route service to the ngView directive, so currently there is no valid use case
for specifying the scope for ngInclude. In fact, allowing the scope to be defined can under
certain circumstances lead to memory leaks.
Breaks ngInclude does not have scope attribute anymore.
It turns out that listening only on "blur" event is not sufficient in many scenarios,
especially when you use form validation you always had to use ngModelnstant
e.g. if you want to disable a button based on valid/invalid form.
The feedback we got from our apps as well as external apps is that the
ngModelInstant should be the default.
In the future we might provide alternative ways of suppressing updates
on each key stroke, but it's not going to be the default behavior.
Apps already using the ngModelInstant can safely remove it from their
templates. Input fields without ngModelInstant directive will start propagating
the input changes into the model on each key stroke.
Previously if there was a white-space in fn: fn( ) {} we failed to infer no args.
This was originally reported by recht, but I decided to use a different fix.
Closes#829
We have many instances of this object and we clone them as well (e.g. ng-repeat).
This should save some memory and performance as well.
Double prefixed private properties of attr object:
attr.$element -> attr.$$element
attr.$observers -> attr.$$observers
Update shallowCopy to not copy $$ properties and allow passing optional destination object.
The `attr` object was only shallow copied which caused all observers to be shared.
Fixing similar issue in ng-* boolean attributes as well as ng-src and ng-href.
Instead of using our custom serializer we now use the native one and
use the replacer function to customize the serialization to preserve
most of the previous behavior (ignore $ and $$ properties as well
as window, document and scope instances).
$httpProvider.defaults.transformRequest and $httpProvider.defaults.transformResponse
are now arrays containing single function. This makes it easy to add an
extra transform fn.
adding an extra fn before had to be done in this cluncky way:
$httpProvider.defaults.transformResponse =
[$httpProvider.defaults.transformResponse, myTransformFn];
after this change, it's simply:
$httpProvider.defaults.transformResponse.push(myTransformFn);
Breaks angular.fromJson which doesn't deserialize date strings into date objects.
This was done to make fromJson compatible with JSON.parse.
If you do require the old behavior - if at all neeeded then because of
json deserialization of XHR responses - then please create a custom
$http transform:
$httpProvider.defaults.transformResponse.push(function(data) {
// recursively parse dates from data object here
// see code removed in this diff for hints
});
Closes#202
So that we can have non string values, e.g. ng-value="true" for radio inputs
Breaks boolean attrs are evaluated rather than interpolated
To migrate your code, change: <input ng-disabled="{{someBooleanVariable}}">
to: <input ng-disabled="someBooleanVariabla">
Affected directives:
* ng-multiple
* ng-selected
* ng-checked
* ng-disabled
* ng-readonly
* ng-required
This service has been accidentaly documented in the past, it should not be considered
to be public api.
I'm also removing fallback to Modernizr since we don't need it.
Breaks any app that depends on this service and its fallback to Modernizr, please
migrate to custom "Modernizr" service:
module.value('Modernizr', function() { return Modernizr; });
It's now possible to register controllers as:
.register('MyCtrl', function($scope) { ... });
// or
.register('MyCtrl', ['$scope', function($scope) { ... });
Additionally a module loader shortcut api was added as well:
myModule.controller('MyCtr', function($scope) { ... });
For typical app that has ng-app directive on the html element, we now can do:
angular.element(document).injector() or .injector()
angular.element(document).scope() or .scope()
instead of:
angular.element(document.getElementsByTagName('html')[0]).injector()
...
This makes for a much more flexible route matching:
- route /foo matches /foo and redirects /foo/ to /foo
- route /bar/ matches /bar/ and redirects /bar to /bar/
Closes#784
Compiler should not reassign values to element attributes if its not neccessary due
to interpolation or special attribute magic (ng-src -> src)
This resolves several issues on IE caused by reassigning script.src attribute which
caused all of the scripts to be reloaded.
In IE window.console.log and friends are functions that don't have apply or call fns.
For this reason we have to treat them specially and do our best to log at least
something when running in this browser.
Closes#805
In ie7 all of the input fields are set to readonly and disabled, because ie7 enumerates over all attributes even if the are not declared on the element.
Added support of timezone in dates not just zulu timezone.
This fixes issues for date filter which uses json deserialization under the hood. (for now)
Closes #/800
Fixed an issue where a directive that uses transclusion (such as ngRepeat) failed to link if it was declared on the root element of the compilation tree. (For example ngView or ngInclude including template where ngRepeat was the top most element).
corrected omitted assignment of controller to the element data object. Without this fix the controller created by ngView is not accessible from the browser debugger.
this is to enable nicer tests:
describe('fooSvc', function() {
var fooSvc;
beforeEach(inject(function(_fooSvc_) {
fooSvc = _fooSvc_;
}));
it('should do this thing', function() {
//test fooSvc
});
});
we can't provide this functionality because the directives are lazy
loaded when the module loads, which is too late for the shiv to do
anything useful.
- change custom onload directive to special arguments recongnized by both
ng-view and ng-include
- rename $contentLoaded event to $viewContentLoaded and $includeContentLoaded
- add event docs
It's more likely you are using angular.fromJson() inside Angular world, which means you get proper
exception handling by $exceptionHandler.
There is no point to explicitly push it to console and it causes memory leaks on most browsers
(tried Chrome stable/canary, Safari, FF).
Problems:
- controller was instantiated immediately on $afterRouteChange (even if no content), that's
different compare to ng:controller, which instantiates controllers after compiling
- route listened on current scope ($afterRouteChange), so if you were listening on $rootScope
($afterRouteChange), you get called first and current.scope === undefined, which is flaky
- route handles scope destroying, but scope is created by ng:view
- route fires after/before route change even if there is no route (when no otherwise specified)
Solution:
- route has no idea about scope, whole scope business moved to ng:view (creating/destroying)
- scope is created (and controller instantiated) AFTER compiling the content
- that means on $afterRouteChange - there is no scope yet (current.scope === undefined)
- added $contentLoaded event fired by ng:view, after linking the content
- remove $formFactory completely
- remove parallel scope hierarchy (forms, widgets)
- use new compiler features (widgets, forms are controllers)
- any directive can add formatter/parser (validators, convertors)
Breaks no custom input types
Breaks removed integer input type
Breaks remove list input type (ng-list directive instead)
Breaks inputs bind only blur event by default (added ng:bind-change directive)
Reason to fix this was the fact that with undefined url, it ended up with weird exception
(Cannot call method 'replace' of undefined), which was more confusing than helpful.
jQuery.ajax() does request to current url, if url is not specified, so I decided for this solution.
These methods cause IE8 holds the whole jqLite in the memory, even when page is reloaded.
jqLite's cache keeps element's data (event handlers, attached scopes, injector, etc…), so almost all used memory is never released in IE8.
jQuery creates its own Event object (wrapper around native Event) instead.
currently we run into infinite digest if a function is being
watched as an expression. This is because we generate bound
function wrapper when the watch is processed via parser.
I'm not too keen on the solution because it relies on the unbound
fn that is being exposed for other reasons, but I can't think
of a better way to deal with this right now
As scopes are injected into controllers now, you have the reference anyway, so having scope as first argument makes no sense…
Breaks $watcher gets arguments in different order (newValue, oldValue, scope)
Controller is standalone object, created using "new" operator, not messed up with scope anymore.
Instead, related scope is injected as $scope.
See design proposal: https://docs.google.com/document/pub?id=1SsgVj17ec6tnZEX3ugsvg0rVVR11wTso5Md-RdEmC0kCloses#321Closes#425
Breaks controller methods are not exported to scope automatically
Breaks Scope#$new() does not take controller as argument anymore
Chrome (probably other browsers as well) fires 'hashchange' event synchronously, so if you change raw location from within $apply/$digest, we don't want to $apply twice. (It would throw an exception)
- there are too many unknowns about PATCH, so I'm dropping its support until we know that this is actually useful
- expectGET, expectHEAD and expectJSON (and the same for whenXXX) should not require response data to be specified
Now, that we have autoscroll attribute on ng:include, there is no reason to disable the service completely, so $anchorScrollProvider.disableAutoScrolling() means it won't be scrolling when $location.hash() changes.
And then, it's not $autoScroll at all, it actually scrolls to anchor when it's called, so I renamed
it to $anchorScroll.
$http:
- use promises internally
- get rid of XhrFuture that was previously used internally
- get rid of $browser.defer calls for async stuff (serving from cache),
promises will take care of asynchronicity
- fix transformation bugs (when caching requested + multiple request
pending + error is returned)
- get rid of native header parsing and instead just lazily parse the
header string
$httpBackend:
- don't return raw/mock XMLHttpRequest object (we don't use it for
anything anymore)
- call the callback with response headers string
mock $httpBackend:
- unify response api for expect and when
- call the callback with response headers string
- changed the expect/when failure error message so that EXPECTED and GOT
values are aligned
Conflicts:
src/service/http.js
test/service/compilerSpec.js
test/service/httpSpec.js
The input event is fired on all non-ie browsers whenever the contents of an input
field changes. This means that we now support cut&paste via mouse which
was previously unsupported.
IE8 and older don't support this events and IE9 has a problematic
support for it, so we can't rely solely on this event and drop keydown
and change events.
Previously we used to put callbacks on the window object, but that
causes problems on IE8 where it is not possible to delete properties
from the window object
closure compiler is smarter than we expected and drops the unused fn
argument - this breaks the meta-programing logic of jqLite.
The fix special cases JQLiteHasClass since its the only fn that needs
this treatment in a way that is minification-proof.
This fix is similar to what I've done in ng:view, if a new template has been requested before the
callback for the previous template returned, ignore it. Otherwise weird race conditions happen
and users might end up getting the content for the previous include rendered instead of the most
recent one.
Parser now builds expressions that can detect promises and transparently
evaluate them to undefined or the promise value.
If promiseA is resolved with value 'A', then {{promiseA}} evals to 'A';
If promiseA is unresolved, then {{promiseA}} evals to undefined;
Following invocations are supported:
- {{promise}}
- {{promise.futureProp}}
- {{[promise][0]}}
- {{object.promise}}
- {{object[promise]}}
- {{array[promise]}}
- {{fn(promise)}}
- combinations of the above
If jsonp is not successfull, we return internal status -2.
This internal status should by normalized by $xhr into 0,
but $xhr needs to distinguish between jsonp-error/abort/timeout (all status 0).
- since NaN !== NaN in javascript digest can get into an infinite loop
when model value is set to NaN
- angular.equals(NaN, NaN) should return true since that's what we
expect when comparing primitives or objects containing NaN values
Previously NaN because of its special === properties was used as the
initial value for watches, but that results in issues when NaN is used
as model value.
In order to allow for model to be anything incuding undefined and NaN we
need to mark the initial value differently in a way that would avoid
these issues, allow us to run digest without major perf penalties and
allow for clients to determine if the listener is being called because
the watcher is being initialized or because the model changed. This
implementation covers all of these scenarios.
BREAKING CHANGE: previously to detect if the listener was called because
the watcher was being initialized, it was suggested that clients check
if old value is NaN. With this change, the check should be if the newVal
equals the oldVal.
Closes#657
- turn scope into a $rootScope service.
- injector is now a starting point for creating angular application.
- added inject() method which wraps jasmine its/beforeEach/afterEach,
and which allows configuration and injection of services.
- refactor tests to use inject() where possible
BREAK:
- removed angular.scope() method
- better compatibility with 3rd party code - we clober 3rd party
style only if it direcrtly collides with 3rd party styles
- better perf since it doesn't execute stuff on every digest
- lots of tests
Commit 5a2dcb9a doesn't properly modify angular-boostrap.js.
This fix resolves issues and makes both the regular and scenario
version of angular-boostrap.js functional.
The last script element in the dom is always us if the script that
contains angular is loaded synchronously.
For async loading manual bootstrap needs to be performed.
Close#621
Along the way I also changed the repeater impl to use for loop instead
of for in loop.
Iteration over objects is handled by creating an array of keys, which is
sorted and this array then determines the order of iteration over an
element. This makes repeating over objects deterministic and
cross-browser compatible.
A lot of badness happens when we don't ignore stale xhrs. These
raceconditions are only apparent when user clicks through the app very
quckly without waiting for routes to fully load.
Closes#619
This functionality was previously available only as obscure $browser.defer.cancel.
I also added docs and tests and fixed an issue in .defer.cancel mock.
For example:
<a href="some/link">inner <span>text</span></a>
If you click on "text", then the span element is event.target, so we need to traverse the DOM.
Change introduced by me in 8611ebe6 results in considerable inefficiencies when the compiler
and linker is used from within a widget, in which case, we call $digest unnecessary since it
will be called by the $apply which called the directive/widget in the first place.
There are only two places when the extra $digest call can be useful - when manually bootstrapping
the app or in tests. However even in tests this behavior can result in unwanted results (especially
when ng:controller is involved). So it is better to leave it for the developer to call $digest
when it is really needed.
Because only controllers don't have currying, we can infer its arguments, all other APIs needing currying, automatic inference complicates the matters unecessary.
this fix is needed for Firefox or other browsers that strictly follow
dom/css spec which states that element.style should make properties
available in camelCased form.
Closes#569
defaults definned per action should take precedence over defaults
defined for the whole resource.
This is potentialy a BREAKING CHANGE in case someone relied on the buggy
behavior.
- Configure our docs app to use new $location with html5 history api!
- Update simple node web server to serve index.html for all links
(rewritting).
- Update .htaccess file to serve index.html for all links (rewritting).
- At runtime determine the base href path and attach it to the DOM. We
needed the absolute URL to get all browsers to work well.
- Because of the above, we also need to dynamically determine all needed
js/css resources and add them to the DOM. This was needed because FF6
would eagerly fetch resources with wrong URL since the base element is
added to the dom at runtime.
- All content html files were moved to the partials directory, because
with the new html5 urls it was impossible to tell if request for
http://domain/api/angular.filter.html was an html5 url for the html
filter doc page, or an xhr/appcache request for the content html file
for the html filter.
f
Because of changes in jQuery, we need to use element().prop() instead of element().attr() to retrieve className and other element properties.
Additionally all attribute selectors (e.g. input[name=value]) must have value quoted if it contains dots (".").
since jQuery 1.6.4 attr() focuses only on work with element attributes and doesn't deal well with element properties, so adding prop() support is required for getting many e2e tests to pass after upgrading the runner to jQuery 1.6.4.
Contains 3 fixes:
- the internal model was by mistake using "checked" property instead of
"selected"
- use jqLite.prop() to set 'selected' property
- added inChangeEvent check - we should not interfere with the browser
selecting elements when not necessary
since jQuery 1.6.4 prop() became very important because attr() does't have access to certain properties any more (e.g. className), so I'm adding it to jqLite as well so that jqLite preserves the feature-set it had before the jQuery upgrade.
The behavior of attr() getter and setter changed in jQuery 1.6 and now they treat element properties and attributes as two different things, but in order to not break everyone there is a partial backwards compatibility for checking and updating element properties as well. see http://api.jquery.com/prop/ for more info.
- move all script load order into angularFiles.js
- rakefile and angular-bootstrap.js use angularFiles.js to get script orders
- gen_jstd_configs.js uses angularFiles.js to generate various jstd config files
- run gen_jstd_configs.js whenever we run server.sh
Closes#470
This is jQuery incompatible hack.
But we were doing monkey patching there anyway...
`$(...).trigger('click')` returns an array of return values, so that scenario
runner knows, whether the event default action was cancelled.
Without this fix, scenario runner was doing navigation even if JS code called
`event.preventDefault()`.
Note, this does not work in FF6
https://developer.mozilla.org/en/DOM/element.dispatchEvent
dispatchEvent method returns false if at least one of the event handlers called
preventDefault(), true otherwise.
It's helpful when browserTrigger method returns this value, as we can assert,
whether the default operation was cancelled or not.
* update $route to reflect new $location
* add some more unit tests to $route
* fix some other failing unit tests
* redirect overrides the url now
Breaks $route custom redirect fn has only 3 params now
This is just basic implementation of $browser.url, $browser.onUrlChange methods:
$browser.url() - returns current location.href
$browser.url('/new') - set url to /new
If supported, history.pushState is used, location.href property otherwise.
$browser.url('/new', true) - replace current url with /new
If supported, history.replaceState is used, location.replace otherwise.
$browser.onUrlChange is only fired when url is changed from the browser:
- user types into address bar
- user clicks on back/forward button
- user clicks on link
It's not fired when url is changed using $browser.url()
Breaks Removed $browser.setUrl(), $browser.getUrl(), use $browser.url()
Breaks Removed $browser.onHashChange(), use $browser.onUrlChange()
- buzz api keeps on throttling our requests which makes our build fail
so I'm disabling the buzz demo e2e test
- the $xhr service jsonp test was modified to use jsonp on angularjs.org
instead of buzz api for the same reason as mentioned above
ng:class as well as ng:class-odd and ng:class-even always reset the
class list to whatever it was before compilation, this makes it
impossible to create another directive which adds its own classes on the
element on which ng:class was applied.
the fix simply removes all classes that were added previously by
ng:class and add classes that the ng:class expression evaluates to.
we can now guarantee that we won't clobber stuff added before or after
compilation as long as all class names are unique.
in order to implement this I had to beef up jqLite#addClass and
jqLite#removeClass to be able to add/remove multiple classes without
creating duplicates.
- both $watch and $on now return a function which when called
deregisters the listener
- $removeListener was removed and replaced with the above
- added more tests for $watch and $on
Closes#542
- special attrs such as ng:href, ng:check did not work as intended when
their values do not contain bindings. And this commit is to fix that
Closes#534
- register listeners with $on
- remove listeners with $removeListener
- fire event that bubbles to root with $emit
- fire event that propagates to all child scopes with $broadcast
In order to avoid unnecesary route reloads when just hashSearch part
of the url changes, it is now possible to disable this behavior by
setting reloadOnSearch param of the route declaration to false.
Closes#354
Change doc_widget.js to:
- render "edit in jsfiddle" button next to all examples
- make opt out certain examples by adding jsfiddle="false" attribute to
doc:source element
it turns out that even with our tricks, jqLite#show is not usable in
practice and definitely not on par with jQuery. so rather than
introducing half-baked apis which introduce issues, I'm removing them.
I also removed show/hide uses from docs, since they are not needed.
Breaks jqLite.hide/jqLite.show which are no longer available.
we commonly assign stuff in if statments like this:
if (variable = someFn()) {
//do something with variable
}
This results in lint and IDE warnings (did you mean ==?).
It is better to be explicit about our intention and wrap the assignement
into parens:
if ((variable = someFn())) {
//do something with variable
}
Doing so suppresses warnings + is easier to understand the intention.
I verified that the closure compiler strips the extra parens, so there
is no byte overhead for this safety practice.
We should use this style going forward...
The support for the 'z' formatting flag was removed becase the timezone
info can't be retrieved from the browser apis (except for en-US locale
on some but not all browsers). For this reason we don't want to support
this flag at all.
Related to this, since the 'long' and 'longtime' datetime formats require
the 'z' flag in the formatting string, we are removing support for this
format as well.
- filter.number, filter.currency and filter.date are injected with
$locale service so that we can just swap the service to localize these
- date filter was beefed up in order to support literal strings found in
localization rules
These widgets are useless and only trigger extra $updateViews.
The only reason we had them was to support ng:change on these widgets,
but since there are no bindings present in these cases it doesn't make
sense to support ng:change here. It's likely just a leftover from
getangular.com
Breaking change: ng:change for input[button], input[submit], input[reset], input[image]
and button widgets is not supported any more
- Speed improvements (about 4x on flush phase)
- Memory improvements (uses no function closures)
- Break $eval into $apply, $dispatch, $flush
- Introduced $watch and $observe
Breaks angular.equals() use === instead of ==
Breaks angular.scope() does not take parent as first argument
Breaks scope.$watch() takes scope as first argument
Breaks scope.$set(), scope.$get are removed
Breaks scope.$config is removed
Breaks $route.onChange callback has not "this" bounded
The previous implementation didn't handle situation when in css
something was hidden with a cascaded display:none rule and then we
wanted to show it.
Unfortunatelly our test doesn't test this scenario because it's too
complicated. :-/
both numbers and currency need to be formatted using a generic pattern
which can be replaced for a different pattern when angular is working in
a non en-US locale
for now only en-US locale is supported, but that will change in the
future
- fixed copyright overnship
- updated copyright years
- added @license tag so that closure compiler preserves the header
- added version number into headers (finally!)
- add support for full,long, medium, short datetime formats in en
Breaks MMMMM. now we don't support MMMMM anymore as old implementation differs
from Unicode Locale Data format we are following.
- removed support for fullDateTime and fullTime as it means too much
trouble with full timeZone names
- added docs for the new features
- If the third param of TzDate constructor is defined, toStirng will
just return this third parameter. Otherwise, toString will still
be treated as unimplemented method
This reverts commit 2428907259.
We decided to revert this because it is not bullet proof. The issue is
that we can't reliably have both angular and non-angular code in charge
of the DOM. We could work around some issues here and there, but we
can't do it reliably, so it's better not to support DOM manipulation
that happens outside of angular. There is a good chance that once we
integrate with MDVs our possition will change, but until then our
position is that only angular or angular widgets/directives can change
change DOM that was compiled.
- browser should remember the last value retrieved via browser.getUrl
- browser should update window.location only if the new value is
different from the current window.location value
the flag must be in all src and test files so that we get the benefit of
running in the strict mode even in jstd
the following script was used to modify all files:
for file in `find src test -name "*.js"`; do
echo -e "'use strict';\n" > temp.txt
cat $file >> temp.txt
mv temp.txt $file
done
- add 'use strict'; statement to the prefix file
- configure closure compiler to use the ES5 strict mode
- strip all file-specific strict mode flags after concatination
Closes#223
Breaks $browser.poll() method is moved inline to $browser.startpoll()
Breaks $browser.startpoll() method is made private
Refactor tests to reflect updated browser API
Closes#387
- e2e tests will run index.html (without jquery) and with
index-jq.html(with jquery).
- many small changes to make e2e tests work withough JQuery as we
discover problems that were previously hidden by using real JQuery.
Rewrite $route example a bit, as it required $location and $route services
to be eager published in the root scope.
Fix small typos in formatter and ng:options docs.
The var eventHandler was defined outside forEach loop, so registering more
events caused calling listeners registered by the last one.
Regression:
elm.bind('click keyup', callback1);
elm.bind('click', callback2);
elm.bind('keyup', callback3);
Firing click event would have executed callback1, callback3 !
$xhr header defaults are now exposed as $xhr.defaults.headers.common and
$xhr.default.headers.<httpmethod>. This allows applications to configure
their defaults as needed.
This commit doesn't allow headers to be set per request, only per
application. Per request change would require api change, which I tried
to avoid *for now*.
Creating <script> tags would require a lot of extra work if we want all browsers
to load and execute these scripts. We decided to not implement that in jqLite.
See #369 for more information.
Closes#369
Support new date format, specifically day of week/Month of year in string
e.g. {{ someDate | data:"EEE, MMM d, yyyy" }} -> "Wed, Jul 10, 2011"
Closes#396
8cad231 broke $eager services
Problem is that the injector.eager function is not invoked when a new scope
is created. Added a test to make sure service is eagerly instantiated.
Closes#403
This behavior was changed by e83465c362
So this commit should have been part of e83465c362
Also removed hiding and navigating to about:blank as it makes no sense now...
Other browsers prepend "Error: <Exception name>" to stack, but FF4 and Opera do not.
So when formatting error we prepend it by hand, when not present...
ng:href was producing unclickable links, as the event propagation was stopped by 'a' widget
All links in regression/issue-352.html were tested in:
* Chrome 11
* Opera 11
* Firefox 4
* IE7, IE8
Closes#352
Couple of changes into angular.scenario runner:
- add autotest config (runs tests when document ready)
- update ObjectModel (forwards events)
- use only one ObjectModel instance for all outputters
- expose error msg and line number in ObjectModel.Spec and ObjectModel.Step
- fix generating spec.ids
- fix 'html' output so that it does not mutate ObjectModel
Couple of changes into docs / generator:
- rename copy -> copyTpl
- move docs/static into docs/examples (to avoid conflict with jstd proxy)
Running all docs e2e tests:
========================================================
1/ compile angular-scenario, jstd-scenario-adapter
>> rake compile
2/ build docs
>> rake docs
3/ start jstd server
>> ./server-scenario.sh
4/ capture some browser
5/ run node server to serve static content
>> node ../lib/nodeserver/server.js
6/ run tests
>> ./test-scenario.sh
Removed angular.scenario.Application.checkUrlStatus_ method and these tests:
* should call error handler if status check fails
* should perform a HEAD request to verify file existence
* should call error handler if status code is less than 200
* should call error handler if status code is greater than 299
* should call error handler if status code is greater than 299
This is a combination of 4 commits:
* Fix some small typos, missing semi-colons, etc.
* Fix comment for angular.scenario.SpecRunner.run method
* Fixed some missing semi-colons in cookbook
* Fixed missing semi-colon in nodeserver/server.js
when we stopped exposing $location service on the root scope
the scenario runner was not modified to access the $location
service via $service
The following apis were affected:
- browser().location().hashSearch()
- browser().location().hashPath()
- browser().location().search()
sync caching in ng:view must be reverted becase ng:view uses
$route.onChange to listen for changes.
$route fires all onChange events before it calls $become(Controller)
which means that if the template being included via ng:view contains
ng:controller, ng:include or other widget that create new scopes,
these scopes will be created and initialized before the parent scope
is fully initialized (happens after $become is called).
For this reason ng:view must be async.
The new scope implemenetation will resolve this issue by providing
us with an api to register one-off tasks to be executed during the
flush phase. We'll be able to compile and link the included template
safely at this time.
- unless we are repeating OPTION elements, buffer new nodes in document
fragment and append them to the DOM in one go at the end
- for OPTION elements we have to keep on using the old way
because of how option widget communicates with select widget
this should be change, but that change is out of scope of this CL
- modify jqLite to support wrapping of document fragments
- fix jqLite documentation typo
This change unintentionally avoids the following webkit bug that
that affects repeater growth:
https://bugs.webkit.org/show_bug.cgi?id=57059
However the following bug affecting shrining of repeaters is still
unresolved https://bugs.webkit.org/show_bug.cgi?id=57061
- extend size() to take size(obj, ownPropsOnly)
- add specs for size()
- update docs to mention string support
- use size() in ng:repeat
including the hasOwnProp check for all object doesn't create
significant perf penalty:
http://jsperf.com/dedicated-code-branch-for-hasownprop
- add `sync` flag xhr.cache
- change ng:include to use the sync flag
- change ng:view to use the sync flag
The end result is that there are fewer repaints in the browser,
which means less "blinking" that user sees.
they have no significant effect on minified and gziped size. in fact
they make things worse.
file | before | after removal
----------------------------------------
concat | 325415 | 325297
min | 62070 | 62161
min + gzip | 25187 | 25176
The bottom line is that we are getting 0.05% decrease in size after
gzip without all of the hassle of using underscores everywhere.
so it is possible to easily compile just a part of a document.
e.g.:
<html>
<head>
<title>partially compiled doc</title>
<script src="angular.js" ng:autobind="compileThis"></script>
</head>
<body>
this part won't be compiled: {{1+2}}
<div id="compileThis" ng:init="i=0" ng:click="i = i+1">
Click count: {{i}}
</div>
</body>
</html>
")]}\',\n" is a commonly used security prefix added to json http
responses iat google and elsewhere in order to prevent certain
cross-site attacks
$xhr service now autodetects the prefix and strips it before
deserializing the json.
the implementation should be more flexible to allow for wider range
of prefixes, but we need this one right now and can address other
usecases later.
angular.compile()() returns {scope:scope, view:view},
this isn't useful at all and only makes tests more verbose.
Instead, this change makes the linking function return scope directly
and if anyone needs the linked dom there are two ways to do it
documented in angular.compile.
other changes:
- moved angular.compile docs to the compiler so that they are closer to
the compiler
- fixed some typos and updated angular.compile docs with the new return
value
- split up services into files under src/service
- split up specs into files under test/service
- rewrite all specs so that they don't depend on one global forEach
- get rid of obsolete code and tests in ng:switch
- rename mock $log spec from "$log" to "$log mock"
I extracted mock $log factory into stand alone function, so we can access it and test, because this service is rewritten by real service during testing, so we can't access it through angular.$service('$log')...
- split mocks between angular-mocks.js and mocks.js
- src/angular-mocks.js now contains only mocks that we want to ship
- test/mocks.js contains mocks that we use internally for testing
angular
- created angular.mock namespace
- created public $exceptionHandler mock rethrows errors
- created public $log mock stores all logs messages in an array that can
be accessed to make assertions
- internally we now have factory to create $exceptionHandler
that we can assert on
- internally we also keep track of all messages logged and
fail tests if messages were not expected and cleaned up (checked
via global beforeEach and afterEach)
- updated RakeFile and docs reader.js to point to the new
angular-mocks.js location
- made real $exceptionHandler and $log factories accessible from tests
and simplified their specs
- fixed typos in several spec descriptions
- added log assertions throughout the test suite
- this change is needed because of previously reverted $route changes
that used to propagate evals automatically.
- also added docs to highlight how the eval propagation works
This reverts commit 17ee0f031a.
this and a5eb3ed1 introduced scope leakage that can't be resolved
until we fix up scope relationships and eval propagation.
I'm reverting both this and a5eb3ed1 until we can properly address
the issue.
- retrieveScope run into infinite loop if called on DOM tree
that doesn't contain scope reference (happens only with jQuery)
- added missing specs for retrieveScope function
select (one/multiple) could not chose from a list of objects, since DOM requires string ids.
Solved by adding index formatter, which exposed incorrect handling of formatters in select
widgets.
- don't update browser before and after eval instead
- sync location properties before eval
- sync location properties and update browser after eval
- added tests
- symplified the code
- removed $location.toString() because it was not idempotent and useless
This resolves the issue with issuing two $route.onHashChange calls
when the $location was updated with a hashPath that needs to be encoded
camelcase is used for other angular functions and forEach is also
used by EcmaScript standard.
- rename the internal as well as the external function name
- tweak the implementation of the function so that it doesn't
clober it self when we extend the angular object with an
object that has a forEach property equal to this forEach function
Closes#85
- use jqLite api to bootstrap angular
- when jQuery is present DOMContentLoaded or hacks for IE are used
- when jqLite is present DOMContentLoaded is used for modern browsers
and IE9 and window.onload is used for other browsers.
- test html for comparing DOMContentLoaded with window.onload
Closes#224
in IE window object has length property which makes it look like a
collection to jqLite. This commit makes jqLite properly identify
window as an element even in IE.
IE6 doesn't have Window type, so we need to check against window
object and only then do a more general check against Window. This
is not perfect, but I say screw IE6.
IE9 added support for addEventListener and it looks like some
new events (e.g. DOMContentLoaded) can't be registered via attachEvent,
only addEventListener correctly registeres these. So to make sure
we are using the latest and greatest we need to look for addEventListener
instead of attachEvent in the jqLite feature detection code.
- There was a perceived lag when typing do to the fact that we were
listening on the keyup event instead of keydown. The issue with
keydown is that we can not read the value of the input field. To
solve this we schedule a defer call and perform the model update
then.
- To prevent calling $eval on root scope too many times as well as to
prevent drowning the browser with too many updates we now call the
$eval only after 25ms and any additional requests get ignored. The
new update service is called $updateView
This reverts commit 00ca67e4be.
Now that we don't have published services, we don't need to worryi
about this any more and in fact this behavior is confusing because
to override a service that has dependencies with a services with
no dependencies one must specify $inject:[] or else the old
dependencies will be injected.
Conflicts:
src/Angular.js
test/AngularSpec.js
- Browser now starts the poller on first call to addPollFn()
- Many services ($location, $cookies, $router) are no longer eager-publish. The result is that
unless someone needs the $cookies, they will not cause the Browser to start polling for them.
- Closes#152 ($resource().query() sometimes calls callback before
returning, and it shouldn't)
- add $browser.defer method
- add $defer service
- integrate $browser.defer with outstandingRequests counter in $browser
- fix all old tests that relied on buggy behavior
- move @ng:repeat to widgets.js and its specs to widgetsSpecs.js
- move @ng:non-bindable to widgets.js and its specs to widgetsSpecs.js
- make widget.template suitable for attribute widgets
- fix up the js docs for attribute widgets
- if initRun param is set to false, listener doesn't execute
- the oldValue should equal newValue during the initial execution
- added docs
- added specs
- improve json filter example
- improve filter overview doc
- improving validator overview jsdocs
- simplify number filter examples and make them live + add specs
- various doc fixes
Support ISO 8601 extended format datetime strings (YYYY-MM-DDTHH:mm:ss.SSSZ) as defined
in EcmaScript 5 throughout angular. This means that the following apis switched from
YYYY-MM-DDTHH:mm:ssZ to YYYY-MM-DDTHH:mm:ss.SSSZ (note the added millis) when representing dates:
- angular.Date.toString
- angular.String.toDate
- JSON serialization and deserialization (used by json filter, $xhr and $resource)
If user override existing extension, angular properties ($) will be preserved.
This piece of logic could be refactored into separate method:
Something like we have extend(), addMissingProperties() - I can't find a name
for this method...
Closes#51
The reason to void these to objects is that they cause all sorts
of problems like exceptions being thrown and infinite loops occuring
when we iterate over object properties.
e.g.
Before:
code: element('.actions ul li a').click();
output: element .actions ul li a click
After
code: element('.actions ul li a', "'Configuration' link").click();
output: element 'Configuration' link ( .actions ul li a ) click
- By default the runner now creates multiple output formats as it runs. Nodes are created in the DOM with ids: json, xml, and html.
ex. $('#json').html() => json output of the runner
ex. $('#xml').html() => json output of the runner
$result is also an object tree result.
The permitted formats are html,json,xml,object.
If you don't want certain formats you can select specific ones with the new ng:scenario-output attribute on the script tag.
<script src="angular-scenario.js" ng:scenario-output="xml,json">
- Added element(...).count() that returns the number of matching elements for the selector.
- repeater(...).count() now returns 0 if no elements matched which can be used to check if a repeater is empty.
- Added toBe() matcher that does strict equality with ===
- Implement iit and ddescribe. If iit() is used instead of it() then only that test will run. If ddescribe() is used instead of describe() them only it() statements inside of it will run. Several iit/ddescribe() blocks can be used to run isolated tests.
- Implement new event based model for SpecRunner. You can now listen for events in the runner. This is useful for writing your own UI or connecting a remote process (ex. WebDriver). Event callbacks execute on the Runner instance.
Events, if fired, will always be in the below order. All events always happen
except for Failure and Error events which only happen in error conditions.
Events:
RunnerBegin
SpecBegin(spec)
StepBegin(spec, step)
StepError(spec, step, error)
StepFailure(spec, step, error)
StepEnd(spec, step)
SpecError(spec, step, error)
SpecEnd(spec)
RunnerEnd
- Only allow the browser to repaint every 10 steps. Cuts 700ms off Firefox in benchmark, 200ms off Chrome.
- Bug Fix: Manually navigate anchors on click since trigger wont work in Firefox.
Sanitization works in two phases:
1) We parse the HTML into sax-like events (start, end, chars).
HTML parsing is very complex, and so it may very well be that what
most browser consider valid HTML may not pares properly here,
but we do best effort. We treat this parser as untrusted.
2) We have safe sanitizeWriter which treats its input (start, end, chars)
as untrusted content and escapes everything. It only allows elements
in the whitelist and only allows attributes which are whitelisted.
Any attribute value must not start with 'javascript:'. This check
is performed after escaping for entity (&xAB; etc..) and ignoring
any whitespace.
- Correct linky filter to use safeHtmlWriter
- Correct html filter to use safeHtmlWriter
Close#33; Close#34
It sort of worked since the browser would throw stack too deep
exception and the angular would then print the error to console.
So as long as you did not have console open you would not notice
this as an error.
The location should be based on the base path of the angular script
and the version identifier of the angular script.
ex: angular.js -> angular-ie-compat.js
js/angular-0.9.0.min.js -> js/angular-ie-compat-0.9.0.js
Polish the Scenario Runner UI to include:
- a scroll pane that steps appear in since the list can be very long
- Collapse successful tests
- Show the line where the DSL statements were when there's an error (Chrome, Firefox)
Also:
- Remove lots angular.bind calls to reduce the amount of stack space used.
- Use setTimeout(...,0) to schedule the next future to let the browser breathe and have it repaint the steps. Also prevents overflowing the stack when an it() creates many futures.
- Run afterEach() handlers even if the it() block fails.
- Make navigateTo() take a function as the second argument so you can compute a URL in the future.
- Add wait() DSL statement to allow interactive debugging of tests.
- Allow custom jQuery selectors with element(...).query(fn) DSL statement.
Known Issues:
- All afterEach() handlers run even if a beforeEach() handler fails. Only after handlers for the same level as the failure and above should run.
This commit was produced by a combination of 4 commits:
- Added URL_MATCH test for basic url
- Moved two tests from $location to URL_MATCH, as they should be here
- Added test for host without "/" ending and fix the regexp to pass the test
- Added another test for matching empty abs path ("/") and fix the regexp
- "this" always means the current chain scope inside a DSL
- addFutureAction callbacks now take ($window, $document, done)
- $document has a special method elements() that uses the currently selected nodes in the document as defined by using() statements.
- $document.elements() allows placeholder insertion into selectors to make them more readable.
ex. $document.elements('input[name="$1"]', myVar) will substitute the value of myVar for $1 in the selector. Subsequent arguments are $2 and so on.
- $document.elements() results have a special method trigger(event) which should be used to events. This method implements some hacks to make sure browser UI controls update and the correct angular events fire.
- futures now allow custom formatting. By default any chain that results in a future can use toJson() or fromJson() to convert the future value to and from json. A custom parser can be provided with parsedWith(fn) where fn is a callback(value) that must return the parsed result.
Note: The entire widgets.html UI is now able to be controlled and asserted through DSL statements!!! Victory! :)
* embedded images as data URIs
* rake task to generate multipart js file with embeded images for IE
* move images into a separate directory outside of src or css and
keep them there for reference
* clean up Rakefile and ruby code
* .gitignore update
* don't penalize IE 8+ with an extra request to the ie-compat.js file
update(objOrString)
updateHash(objOrString [, objOrString])
toString()
cancel()
Examples:
$location.update('http://www.angularjs.org/path#path?a=b');
$location.update({port: 443, protocol: 'https'});
$location.updateHash('hashPath');
$location.updateHash({a: 'b'});
$location.updateHash('hashPath', {a: 'b'});
This commit was produced by squash of more commits, here are the old messages:
- Change tests to use update() instead of parse().
- First implementation of update() method
- Test for update() with object parameter
- Add new tests for location, refactor location code
- Add tests for updateHash()
- Implement updateHash()
- Take one or two arguments, could be string - update hashPath, or hash object - update hashSearch...
- Fixed other service tests, to use new $location.update()
Added $location.cancel() method (with test)
Added $location.parse() for back compatability
Remove parse() method
update(objOrString)
updateHash(objOrString [, objOrString])
toString()
cancel()
Examples:
$location.update('http://www.angularjs.org/path#path?a=b');
$location.update({port: 443, protocol: 'https'});
$location.updateHash('hashPath');
$location.updateHash({a: 'b'});
$location.updateHash('hashPath', {a: 'b'});
This commit was produced by squash of more commits, here are the old messages:
- Change tests to use update() instead of parse().
- First implementation of update() method
- Test for update() with object parameter
- Add new tests for location, refactor location code
- Add tests for updateHash()
- Implement updateHash()
- Take one or two arguments, could be string - update hashPath, or hash object - update hashSearch...
- Fixed other service tests, to use new $location.update()
Added $location.cancel() method (with test)
Added $location.parse() for back compatability
Remove parse() method
Date filter now supports:
yyyy: four digit year
yy: two digit year
MM: two digit month
dd: two digit day of month
HH: two digit hour in 0-23
KK: two digit hour in 0-12
mm: two digit minute
ss: two digit second
a: am/pm
Z: four digit timezone offset
example {{ timestamp | date:'yyyy-MM-dd HH:mm:ss' }} becomes 2010-10-13 14:45:23
- added angular.injector(scope, services, instanceCache) which returns inject
- inject method can return, instance, or call function which have $inject
property
- initialize services with $creation=[eager|eager-publish] this means that
only some of the services are now globally accessible
- upgraded $become on scope to use injector hence respect the $inject property
for injection
- $become should not be run multiple times and will most likely be removed
in future version
- added $new on scope to create a child scope
- $inject is respected on constructor function
- simplified scopes so that they no longer have separate __proto__ for
parent, api, behavior and instance this should speed up execution since
scope will now create one __proto__ chain per scope (not three).
BACKWARD COMPATIBILITY WARNING:
- services now need to have $inject instead of inject property for proper
injection this breaks backward compatibility
- not all services are now published into root scope
(only: $location, $cookie, $window)
- if you have widget/directive which uses services on scope
(such as this.$xhr), you will now have to inject that service in
(as it is not published on the root scope anymore)
- feedback relies on *Controller.init to be called when a Controller is
being created. this with previous angular refactoring this is not happening
in angular any more. To make it easier for feedback to transition, this
change makes $become call controller's init method if present.
- call to Controller.init from $route.updateRoute was removed. this was
left there by accident during the previous refactoring.
* added a widget for A (anchor) tag, that modifies the default behavior
and prevent default action (location change and page reload) for tags
with empty href attribute
* stopped event propagation for all ng:click handlers
- previously the poller initialized the cookie cache too late which
was causing previously existing cookies to be deleted by cookie service
- refactored the poller api so that the addPollFn returns the added fn
- fixed older cookie service tests
- removed "this.$onEval(PRIORITY_LAST, update);" because it is not needed
- remove obsolete code in tests
- add warning logs when maximum cookie limits (as specified via RFC 2965) were reached
- non-string values will now get dropped
- after each update $cookies hash will reflect the actual state of browser cookies
this means that if browser drops some cookies due to cookie overflow, $cookies will reflect that
- $sessionStore got renamed to $cookieStore to avoid name conflicts with html5's sessionStore
- change from using prototype to inner functions to help with better compression
- removed watchers (url/cookie) and introduced a poller concept
- moved the checking of URL and cookie into services which register with poolers
Benefits:
- Smaller minified file
- can call $browser.poll() from tests to simulate polling
- single place where setTimeout needs to be tested
- More testable $browser
When a method foo is called on a Resource object, say myResource there are two copies that happen to the resource:
- one inside Resource.foo() in some dummy function
- another inside myResource.$foo() inside the callback passed to foo()
"partials". The pattern is demostrated in the unittest:
Resource.query returns a list of "keys" to resources, which are
partially defined. They have enough data to allow $get to fetch the
whole gamout. Then $get fetches all the details of the resource.
