MD Settings: For rows containing paper-toggle-buttons, make entire row clickable.

chrome/browser/resources/settings/privacy_page/privacy_page.js

Index: chrome/browser/resources/settings/privacy_page/privacy_page.js
diff --git a/chrome/browser/resources/settings/privacy_page/privacy_page.js b/chrome/browser/resources/settings/privacy_page/privacy_page.js
index 8e0c4b029c5ad5253e57388a4c64b732355b14aa..35ccd677302507fd6f197f1bdebf7b0574e40794 100644
--- a/chrome/browser/resources/settings/privacy_page/privacy_page.js
+++ b/chrome/browser/resources/settings/privacy_page/privacy_page.js
@@ -114,8 +114,11 @@ Polymer({
 // <if expr="_google_chrome and not chromeos">
   /** @private */
   onMetricsReportingControlTap_: function() {
+    if(this.metricsReporting_.managed)
+      return;
+
     var browserProxy = settings.PrivacyPageBrowserProxyImpl.getInstance();
-    var enabled = this.$.metricsReportingControl.checked;
+    var enabled = !this.metricsReporting_.enabled;
     browserProxy.setMetricsReportingEnabled(enabled);
   },
 
@@ -142,9 +145,9 @@ Polymer({
 // </if>
 
   /** @private */
-  onSafeBrowsingExtendedReportingControlTap_: function() {
+  onSafeBrowsingExtendedReportingControlTap_: function(e) {
     var browserProxy = settings.PrivacyPageBrowserProxyImpl.getInstance();
-    var enabled = this.$.safeBrowsingExtendedReportingControl.checked;
+    var enabled = !this.safeBrowsingExtendedReportingEnabled_;

[Dan Beam, 2017/02/10 17:20:36]

i still don't really understand how this works.

but on-change="onSafeBrowsingExtendedReportingControlTap_" means "hey, the value
just changed.  the new value is what you want".  then you invert it and send it
to C++.  how is this correct?

[scottchen, 2017/02/10 18:49:57]

You're thinking of the toggle control's "checked" value, which does get updated
before on-change fires, but this.SBER_Enabled_ doesn't, since it's bound one-way
on the checkbox (check="[[SBER_Enabled_]]").

In other words, if there were no handlers, then when you click on the toggle,
toggle.checked would switch back and forth, but this.SBER_Enabled_ doesn't
actually change until you tell it to. And then when it does change, the toggle
will conform to it's value. This is why I'm saying it's easier to just use that
value as the source of truth as opposed to the checkbox value which might or
might not have de-synced from the value due to weird user interaction (such as
dragging not triggering event handler, etc.)

     browserProxy.setSafeBrowsingExtendedReportingEnabled(enabled);
   },
 
@@ -184,4 +187,8 @@ Polymer({
   getStringTernary_: function(value, trueLabel, falseLabel) {
     return value ? trueLabel : falseLabel;
   },
+
+  doNothing_: function(e) {

[scottchen, 2017/02/10 07:53:39]

This is dipping into the "slightly gross" zone, but not too horrible. 

And I think we can reasonably make this a reusable thing - maybe
[is="contained-toggle"], or make the parent a [is="toggle-container"] and have
its child paper-toggle-button always stop propagating tap, or something. I think
we can achieve this without creating shadow-dom I believe, much like
is=action-link.

+    e.stopPropagation();
+  }
 });