Extract executeScript-like functionality into a single ExtensionScriptExecutor class.

chrome/browser/extensions/script_executor.h

Index: chrome/browser/extensions/script_executor.h
diff --git a/chrome/browser/extensions/script_executor.h b/chrome/browser/extensions/script_executor.h
new file mode 100644
index 0000000000000000000000000000000000000000..50d17e85adc344220d5de272b54e7064bd6edcf2
--- /dev/null
+++ b/chrome/browser/extensions/script_executor.h
@@ -0,0 +1,47 @@
+// Copyright (c) 2012 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.
+
+#ifndef CHROME_BROWSER_EXTENSIONS_SCRIPT_EXECUTOR_H_
+#define CHROME_BROWSER_EXTENSIONS_SCRIPT_EXECUTOR_H_
+#pragma once
+
+#include <string>
+
+#include "base/callback_forward.h"
+#include "chrome/common/extensions/user_script.h"
+
+namespace content {
+class WebContents;
+}
+
+namespace extensions {
+
+// Interface for executing extension content scripts (e.g. executeScript) as

[koz (OOO until 15th September), 2012/05/11 01:09:16]

Woah, an interface! What is this, Java? :-P

[not at google - send to devlin, 2012/05/14 02:27:16]

I wish!

+// described by the ExtensionMsg_ExecuteCode_Params IPC, and notifying the
+// caller when responded with ExtensionHostMsg_ExecuteCodeFinished.
+class ScriptExecutor {
+ public:
+  virtual ~ScriptExecutor() {}
+
+  // Callback from ExecuteScript. The arguments are (success, error).
+  typedef base::Callback<void(bool, const std::string&)> ExecuteScriptCallback;
+
+  // Executes a script. The arguments match ExtensionMsg_ExecuteCode_Params in
+  // extension_messages.h (request_id is populated automatically).
+  //
+  // |callback| will always be called even if the IPC'd renderer is destroyed
+  // before a response is received (in this case the callback will be with a
+  // failure and appropriate error message).
+  virtual void ExecuteScript(const std::string& extension_id,

[koz (OOO until 15th September), 2012/05/11 01:09:16]

I notice that the call sites here become less readable because these parameters
are now unnamed. Up to you but I think it wouldn't be the end of the world if
this took an ExtensionMsg_ExecuteCode_Params. It would also mean you don't have
to worry about the arguments here matching up with those in the IPC struct.

[not at google - send to devlin, 2012/05/14 02:27:16]

I agree these parameters are confusing; particularly the bools. I don't want to
pass in the IPC directly because there are some fields I'll want to pull out and
treat separately (e.g. extension_id) and some that only make sense to set from
within here (e.g. request_id).

I'll make all the bools into enums instead. That should make things much nicer.

+                             bool is_javascript,
+                             const std::string& code,
+                             bool all_frames,
+                             UserScript::RunLocation run_at,
+                             bool in_main_world,
+                             const ExecuteScriptCallback& callback) = 0;
+};
+
+}  // namespace extensions
+
+#endif  // CHROME_BROWSER_EXTENSIONS_SCRIPT_EXECUTOR_H_