Add support for providing an external file for extracting content.

components/dom_distiller/core/distiller_page.cc

 // Copyright 2013 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 "components/dom_distiller/core/distiller_page.h"
 
 #include "base/bind.h"
 #include "base/json/json_writer.h"
 #include "base/logging.h"
 #include "base/message_loop/message_loop.h"
 #include "base/metrics/histogram.h"
 #include "base/strings/string_util.h"
 #include "base/strings/utf_string_conversions.h"
 #include "base/time/time.h"
 #include "grit/components_resources.h"
 #include "third_party/dom_distiller_js/dom_distiller.pb.h"
 #include "third_party/dom_distiller_js/dom_distiller_json_converter.h"
 #include "ui/base/resource/resource_bundle.h"
 #include "url/gurl.h"
 
 namespace dom_distiller {
 
 namespace {
 
 const char* kOptionsPlaceholder = "$$OPTIONS";
 
 std::string GetDistillerScriptWithOptions(
+    const std::string& script,
     const dom_distiller::proto::DomDistillerOptions& options) {
-  std::string script = ResourceBundle::GetSharedInstance()
-                           .GetRawDataResource(IDR_DISTILLER_JS)
-                           .as_string();
-  if (script.empty()) {
+  // Make a copy of |script| so the options can be applied.
+  std::string script_with_options = script;
+  if (script_with_options.empty()) {

[cjhopman, 2015/02/05 02:09:40]

this could just check script.empty() rather than working on the copy

[nyquist, 2015/02/06 18:50:32]

Done.

     return "";
   }
 
   scoped_ptr<base::Value> options_value(
       dom_distiller::proto::json::DomDistillerOptions::WriteToValue(options));
   std::string options_json;
   if (!base::JSONWriter::Write(options_value.get(), &options_json)) {
     NOTREACHED();
   }
-  size_t options_offset = script.find(kOptionsPlaceholder);
+  size_t options_offset = script_with_options.find(kOptionsPlaceholder);
   DCHECK_NE(std::string::npos, options_offset);
   DCHECK_EQ(std::string::npos,
-            script.find(kOptionsPlaceholder, options_offset + 1));
-  script =
-      script.replace(options_offset, strlen(kOptionsPlaceholder), options_json);
-  return script;
+            script_with_options.find(kOptionsPlaceholder, options_offset + 1));
+  script_with_options = script_with_options.replace(

[cjhopman, 2015/02/05 02:09:40]

I'd move the copy down here where it's needed.

[nyquist, 2015/02/06 18:50:32]

Done.

+      options_offset, strlen(kOptionsPlaceholder), options_json);
+  return script_with_options;
 }
 
 }
 
 DistillerPageFactory::~DistillerPageFactory() {}
 
-DistillerPage::DistillerPage() : ready_(true) {}
+DistillerPage::DistillerPage() : ready_(true) {
+  distiller_js_script_ = ResourceBundle::GetSharedInstance()
+                             .GetRawDataResource(IDR_DISTILLER_JS)
+                             .as_string();
+}
+
+DistillerPage::DistillerPage(std::string distiller_js_script)
+    : ready_(true), distiller_js_script_(distiller_js_script) {
+}
 
 DistillerPage::~DistillerPage() {}
 
 void DistillerPage::DistillPage(
     const GURL& gurl,
     const dom_distiller::proto::DomDistillerOptions options,
     const DistillerPageCallback& callback) {
   DCHECK(ready_);
   // It is only possible to distill one page at a time. |ready_| is reset when
   // the callback to OnDistillationDone happens.
   ready_ = false;
   distiller_page_callback_ = callback;
-  DistillPageImpl(gurl, GetDistillerScriptWithOptions(options));
+  DistillPageImpl(gurl,
+                  GetDistillerScriptWithOptions(distiller_js_script_, options));
 }
 
 void DistillerPage::OnDistillationDone(const GURL& page_url,
                                        const base::Value* value) {
   DCHECK(!ready_);
   ready_ = true;
 
   scoped_ptr<dom_distiller::proto::DomDistillerResult> distiller_result(
       new dom_distiller::proto::DomDistillerResult());
   bool found_content;
@@ skipping 51 matching lines @@
   }
 
   base::MessageLoop::current()->PostTask(
       FROM_HERE,
       base::Bind(distiller_page_callback_,
                  base::Passed(&distiller_result),
                  found_content));
 }
 
 }  // namespace dom_distiller