Update effective-TLD data files using the latest data from Mozilla, ignoring all private domains.

Private domains are now ignored by the tld_cleanup.cc preprocessor due to bug 96086 (also see https://bugzilla.mozilla.org/show_bug.cgi?id=687165), in order to allow users to navigate to the "TLDs" directly.

This data file is Mozilla's changeset 290afd57d2a8, from 2012-07-04 16:08 +0100.

It includes changes (some ignored, as they involved private domains) from a number of Mozilla bugs, listed on http://hg.mozilla.org/mozilla-central/filelog/ba8463beab13/netwerk/dns/effective_tld_names.dat between 9411dffc948b (2011-09-02 14:08 -0400) and 290afd57d2a8 (2012-07-04 16:08 +0100).

BUG=37436, 96086
TEST=none
Review URL: https://chromiumcodereview.appspot.com/10789035

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@147236 0039d316-1c4b-4281-b951-d872f2087c98

1 files changed, 33 insertions, 16 deletions

diff --git a/net/tools/tld_cleanup/tld_cleanup.cc b/net/tools/tld_cleanup/tld_cleanup.cc
index d6e42e4..2094a37 100644
--- a/net/tools/tld_cleanup/tld_cleanup.cc
+++ b/net/tools/tld_cleanup/tld_cleanup.cc
@@ -1,4 +1,4 @@
-// Copyright (c) 2006-2008 The Chromium Authors. All rights reserved.
+// 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.
 
@@ -8,7 +8,7 @@
 // generate a perfect hash map.  The benefit of this approach is that no time is
 // spent on program initialization to generate the map of this data.
 //
-// Running this program finds "effective_tld_names.cc" in the expected location
+// Running this program finds "effective_tld_names.dat" in the expected location
 // in the source checkout and generates "effective_tld_names.gperf" next to it.
 //
 // Any errors or warnings from this program are recorded in tld_cleanup.log.
@@ -21,6 +21,9 @@
 //  * Logs a warning if GURL reports a rule as invalid, but keeps the rule.
 //  * Canonicalizes each rule's domain by converting it to a GURL and back.
 //  * Adds explicit rules for true TLDs found in any rule.
+// TODO(pamg): Remove this comment when http://crbug.com/96086 is fixed.
+//  * Ignores any entries in the file between "// ===BEGIN PRIVATE DOMAINS==="
+//    and "// ===END PRIVATE DOMAINS===".
 
 #include <map>
 #include <set>
@@ -47,6 +50,9 @@ struct Rule {
 
 typedef std::map<std::string, Rule> RuleMap;
 typedef std::set<std::string> RuleSet;
+
+const char kBeginPrivateDomainsComment[] = "// ===BEGIN PRIVATE DOMAINS===";
+const char kEndPrivateDomainsComment[] = "// ===END PRIVATE DOMAINS===";
 }
 
 // Writes the list of domain rules contained in the 'rules' set to the
@@ -55,18 +61,18 @@ typedef std::set<std::string> RuleSet;
 bool WriteRules(const RuleMap& rules, const FilePath& outfile) {
   std::string data;
   data.append(
-      "%{\n"
-      "// Copyright (c) 2009 The Chromium Authors. All rights reserved.\n"
-      "// Use of this source code is governed by a BSD-style license that\n"
-      "// can be found in the LICENSE file.\n\n"
-      "// This file is generated by net/tools/tld_cleanup/.\n"
-      "// DO NOT MANUALLY EDIT!\n"
-      "%}\n"
-      "struct DomainRule {\n"
-      "  const char *name;\n"
-      "  int type;  // 1: exception, 2: wildcard\n"
-      "};\n"
-      "%%\n"
+"%{\n"
+"// Copyright (c) 2012 The Chromium Authors. All rights reserved.\n"
+"// Use of this source code is governed by a BSD-style license that can be\n"
+"// found in the LICENSE file.\n\n"
+"// This file is generated by net/tools/tld_cleanup/.\n"
+"// DO NOT MANUALLY EDIT!\n"
+"%}\n"
+"struct DomainRule {\n"
+"  const char *name;\n"
+"  int type;  // 1: exception, 2: wildcard\n"
+"};\n"
+"%%\n"
   );
 
   for (RuleMap::const_iterator i = rules.begin(); i != rules.end(); ++i) {
@@ -177,11 +183,22 @@ NormalizeResult NormalizeFile(const FilePath& in_filename,
   size_t line_end = 0;
   RuleMap rules;
   RuleSet extra_rules;
+  int begin_private_length = arraysize(kBeginPrivateDomainsComment) - 1;
   while (line_start < data.size()) {
-    // Skip comments.
-    if (line_start + 1 < data.size() &&
+    // Skip the entire section of private domains.
+    // TODO(pamg): remove this when http://crbug.com/96086 is fixed.
+    if (line_start + begin_private_length < data.size() &&
+        !data.compare(line_start, begin_private_length,
+                      kBeginPrivateDomainsComment)) {
+      line_end = data.find(kEndPrivateDomainsComment, line_start);
+      if (line_end == std::string::npos) {
+        LOG(WARNING) << "Private-domain section had no end marker.";
+        line_end = data.size();
+      }
+    } else if (line_start + 1 < data.size() &&
         data[line_start] == '/' &&
         data[line_start + 1] == '/') {
+      // Skip comments.
       line_end = data.find_first_of("\r\n", line_start);
       if (line_end == std::string::npos)
         line_end = data.size();