Fix up the Mac installer’s Info.plist, build with branding info

chrome/installer/mac/app/BUILD.gn

Index: chrome/installer/mac/app/BUILD.gn
diff --git a/chrome/installer/mac/app/BUILD.gn b/chrome/installer/mac/app/BUILD.gn
index a84fc30592537e72a8121182833d50e9f0644ed5..2bb19f04a5e8ebf92317bccfa5dba318b7d66e1f 100644
--- a/chrome/installer/mac/app/BUILD.gn
+++ b/chrome/installer/mac/app/BUILD.gn
@@ -2,8 +2,10 @@
 # Use of this source code is governed by a BSD-style license that can be
 # found in the LICENSE file.
 
-import("//testing/test.gni")
 import("//build/config/mac/rules.gni")
+import("//build/mac/tweak_info_plist.gni")
+import("//build/util/branding.gni")
+import("//testing/test.gni")
 
 source_set("mac_installer_base") {
   visibility = [ ":*" ]
@@ -28,8 +30,19 @@ source_set("mac_installer_base") {
   public_configs = [ "//build/config/compiler:enable_arc" ]
 }
 
-mac_app_bundle("mac_installer_app") {
+tweak_info_plist("mac_installer_plist") {
   info_plist = "Info.plist"
+  args = [
+    "--breakpad=0",
+    "--keystone=0",
+    "--scm=1",
+    "--bundle_id=$chrome_mac_installer_bundle_id",
+  ]
+}

[Sidney San Martín, 2016/09/21 21:18:39]

Make sure these flags look OK — the only weird bit I noticed in the output is
that the installer ends up with a Chrome version number like "55.0.2868.0", and
I'm not sure if that's appropriate.

[Robert Sesek, 2016/09/21 21:41:21]

Yeah, I'd maybe skip the tweak step for the time being. I'm not sure how we'll
want to version this.

[Sidney San Martín, 2016/09/21 23:19:20]

Done.

+
+mac_app_bundle("mac_installer_app") {
+  output_name = "$chrome_product_installer_full_name"
+  info_plist_target = ":mac_installer_plist"
   extra_substitutions = [ "MACOSX_DEPLOYMENT_TARGET=10.9" ]

[Robert Sesek, 2016/09/21 21:41:21]

You still need to add PRODUCT_INSTALLER_FULLNAME and MAC_INSTALLER_BUNDLE_ID
here.

[Sidney San Martín, 2016/09/21 23:19:20]

Oof, thanks. I skimmed the output Info.plist for empty or un-substituted fields
and didn't see any, but I wasn't expecting the fields without values to be
omitted completely.

Do you think it would make sense to change mac_info_plist to fail on unbound
substitutions?

[Robert Sesek, 2016/09/22 14:46:56]

When I looked at your binary, I considered the same thing. Yes, I think
gen_plist.py should error on undefined variables, rather than silently removing
the key/value pair.

[Sidney San Martín, 2016/09/22 15:43:48]

Will do.

   sources = [
     "AppDelegate.h",