Auto merge of #17083 - danielj41:javascript-url-global-3, r=jdm

"javascript:" urls: execute in correct global scope

<!-- Please describe your changes on the following line: -->

#### Summary

This pull request makes `javascript:` urls execute in the correct global scope.

#### Example

```html
<script> var x = 4; </script>

<!-- this branch: logs "4" -->
<!-- master: undefined variable error -->
<a href="javascript:console.log(x)">link</a>
```

#### Questions

I'm new to servo and rust, so I'm unsure about these changes. In particular:
  * What's the appropriate place to evaluate the js?
    * I moved it to `handle_navigate`, but I'm not sure if this will catch all occurrences of `javascript:` urls. I also don't know if this will execute in the correct thread and the correct window.
  * What should I do with the result of the js evaluation?
    * I just ignored it. The previous behavior displayed it as the content of a new page load.

---
<!-- Thank you for contributing to Servo! Please replace each `[ ]` by `[X]` when the step is complete, and replace `__` with appropriate data: -->
- [x] `./mach build -d` does not report any errors
- [x] `./mach test-tidy` does not report any errors
- [x] These changes fix #15147, #16718

<!-- Either: -->
- [x] There are tests for these changes

<!-- Also, please make sure that "Allow edits from maintainers" checkbox is checked, so that we can help you if you get stuck somewhere along the way.-->

<!-- Pull requests that do not address these steps are welcome, but they will require additional verification as part of the review process. -->

<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/17083)
<!-- Reviewable:end -->
This commit is contained in:
bors-servo 2017-09-09 04:39:43 -05:00 committed by GitHub
commit 40c8a6389a
17 changed files with 154 additions and 93 deletions

View file

@ -44,7 +44,7 @@ use msg::constellation_msg::{FrameType, BrowsingContextId, PipelineId, TopLevelB
use net_traits::response::HttpsState;
use script_layout_interface::message::ReflowQueryType;
use script_thread::{ScriptThread, Runnable};
use script_traits::{IFrameLoadInfo, IFrameLoadInfoWithData, LoadData, UpdatePipelineIdReason};
use script_traits::{IFrameLoadInfo, IFrameLoadInfoWithData, JsEvalResult, LoadData, UpdatePipelineIdReason};
use script_traits::{MozBrowserEvent, NewLayoutInfo, ScriptMsg};
use script_traits::IFrameSandboxState::{IFrameSandboxed, IFrameUnsandboxed};
use servo_atoms::Atom;
@ -114,7 +114,7 @@ impl HTMLIFrameElement {
}
pub fn navigate_or_reload_child_browsing_context(&self,
load_data: Option<LoadData>,
mut load_data: Option<LoadData>,
nav_type: NavigationType,
replace: bool) {
let sandboxed = if self.is_sandboxed() {
@ -140,11 +140,26 @@ impl HTMLIFrameElement {
// document; the new navigation will continue blocking it.
LoadBlocker::terminate(&mut load_blocker);
if let Some(ref mut load_data) = load_data {
let is_javascript = load_data.url.scheme() == "javascript";
if is_javascript {
let window_proxy = self.GetContentWindow();
if let Some(window_proxy) = window_proxy {
ScriptThread::eval_js_url(&window_proxy.global(), load_data);
}
}
}
//TODO(#9592): Deal with the case where an iframe is being reloaded so url is None.
// The iframe should always have access to the nested context's active
// document URL through the browsing context.
if let Some(ref load_data) = load_data {
*load_blocker = Some(LoadBlocker::new(&*document, LoadType::Subframe(load_data.url.clone())));
match load_data.js_eval_result {
Some(JsEvalResult::NoContent) => (),
_ => {
*load_blocker = Some(LoadBlocker::new(&*document, LoadType::Subframe(load_data.url.clone())));
}
};
}
let window = window_from_node(self);