bluetooth: Add Device service for chrome://bluetooth-internals.

chrome/browser/ui/webui/bluetooth_internals/bluetooth_internals_ui.cc

 // Copyright 2016 The Chromium Authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 #include "chrome/browser/ui/webui/bluetooth_internals/bluetooth_internals_ui.h"
 
 #include "chrome/browser/profiles/profile.h"
 #include "chrome/common/url_constants.h"
 #include "chrome/grit/browser_resources.h"
 #include "content/public/browser/web_ui_data_source.h"
 
 BluetoothInternalsUI::BluetoothInternalsUI(content::WebUI* web_ui)
     : content::WebUIController(web_ui) {
   // Set up the chrome://bluetooth-internals source.
   content::WebUIDataSource* html_source =
       content::WebUIDataSource::Create(chrome::kChromeUIBluetoothInternalsHost);
 
   // Add required resources.
   html_source->AddResourcePath("bluetooth_internals.css",
                                IDR_BLUETOOTH_INTERNALS_CSS);
   html_source->AddResourcePath("bluetooth_internals.js",
                                IDR_BLUETOOTH_INTERNALS_JS);
   html_source->AddResourcePath(
       "device/bluetooth/public/interfaces/adapter.mojom",
-      IDR_BLUETOOTH_MOJO_JS);
+      IDR_BLUETOOTH_ADAPTER_MOJO_JS);
+  html_source->AddResourcePath(
+      "device/bluetooth/public/interfaces/device.mojom",

[dpapad, 2016/10/12 19:42:45]

+dbeam@chromium.org

Still catching up with the design doc at go/fizzbluetoothinternals. I am
noticing a discrepancy of the C++ code backing up chrome://bluetooth-internals
with the rest of WebUI. Specifically

C++ code handling messages from WebUI for all other WebUI pages resides at
chrome/browser/ui/webui/. For the case of bluetooth-internals, the only code
residing there (chrome/browser/ui/webui/bluetooth_internals) is a dummy
content::WebUIController subclass that simply serves the CSS and JS.

The JS files at chrome/browser/resources/bluetooth_internals/ are reaching out
to device/bluetooth directly, which seems a bit odd. Notice how all other
Mojo-ified WebUI pages (plugins/site-engagement,omnibox) have their *.mojom
files (and their implementations of those) inside chrome/browser/ui/webui.

Is it OK to
a) Have WebUI handling code outside of the so far established location
chrome/browser/ui/webui/?
b) Mix device/bluetooth code with WebUI handling code?

I am looping in more WebUI OWNERS. @dbeam: WDYT? Is this breaking existing
conventions in WebUI code structure?

[dpapad, 2016/10/12 21:02:32]

Talked about this with @dbeam in person. It seems that reaching outside of
chrome/browser/ui/webui directly is fine, there are already some examples of
doing so (some WebUI pages exist in content/ instead). Regarding mixing
device/bluetooth code with WebUI handling code, it still seems odd from my
perspective (Chrome's UI related code should live somewhere under chrome/,
top-level directories like device/, printing/ etc are considered lower-level
services not UI), but if device/bluetooth OWNERS are fine with that (already
LGed), then it is OK.

[mbrunson, 2016/10/12 23:20:03]

Great. scheib@ and ortuno@ are the device/bluetooth OWNERS, and I always add and
get LGTM from them first on this project.

+      IDR_BLUETOOTH_DEVICE_MOJO_JS);
   html_source->SetDefaultResource(IDR_BLUETOOTH_INTERNALS_HTML);
   html_source->DisableI18nAndUseGzipForAllPaths();
 
   Profile* profile = Profile::FromWebUI(web_ui);
   content::WebUIDataSource::Add(profile, html_source);
 }
 
 BluetoothInternalsUI::~BluetoothInternalsUI() {}