Anonymize UUIDs

components/feedback/anonymizer_tool.cc

 // Copyright 2015 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/feedback/anonymizer_tool.h"
 
 #include <utility>
 
 #include "base/memory/ptr_util.h"
 #include "base/strings/string_number_conversions.h"
@@ skipping 147 matching lines @@
 #define IRI SCHEME ":" IHIER_PART OPT_NCG("\\?" IQUERY) OPT_NCG("#" IFRAGMENT)
 
 #define IRI_REFERENCE NCG(IRI "|" IRELATIVE_REF)
 
 // TODO(battre): Use http://tools.ietf.org/html/rfc5322 to represent email
 // addresses. Capture names as well ("First Lastname" <foo@bar.com>).
 
 // The |kCustomPatternWithoutContext| array defines further patterns to match
 // and anonymize. Each pattern consists of a single capturing group.
 CustomPatternWithoutContext kCustomPatternsWithoutContext[] = {
-  {"URL", "(?i)(" IRI ")"},
-  // Email Addresses need to come after URLs because they can be part
-  // of a query parameter.
-  {"email", "(?i)([0-9a-z._%+-]+@[a-z0-9.-]+\\.[a-z]{2,6})"},
-  // IP filter rules need to come after URLs so that they don't disturb the
-  // URL pattern in case the IP address is part of a URL.
-  {"IPv4", "(?i)(" IPV4ADDRESS ")"},
-  {"IPv6", "(?i)(" IPV6ADDRESS ")"},
+    {"URL", "(?i)(" IRI ")"},
+    // Email Addresses need to come after URLs because they can be part
+    // of a query parameter.
+    {"email", "(?i)([0-9a-z._%+-]+@[a-z0-9.-]+\\.[a-z]{2,6})"},
+    // IP filter rules need to come after URLs so that they don't disturb the
+    // URL pattern in case the IP address is part of a URL.
+    {"IPv4", "(?i)(" IPV4ADDRESS ")"},
+    {"IPv6", "(?i)(" IPV6ADDRESS ")"},
+    // Universal Unique Identifiers (UUIDs).
+    {"UUID",
+     "(?i)([0-9a-zA-Z]{8}-[0-9a-zA-Z]{4}-[0-9a-zA-Z]{4}-[0-9a-zA-Z]{4}-"
+     "[0-9a-zA-Z]{12})"},
 };
 
 // Like RE2's FindAndConsume, searches for the first occurrence of |pattern| in
 // |input| and consumes the bytes until the end of the pattern matching. Unlike
 // FindAndConsume, the bytes skipped before the match of |pattern| are stored
 // in |skipped_input|. |args| needs to contain at least one element.
 // Returns whether a match was found.
 //
 // Example: input = "aaabbbc", pattern = "(b+)" leads to skipped_input = "aaa",
 // args[0] = "bbb", and the beginning input is moved to the right so that it
@@ skipping 179 matching lines @@
     }
 
     skipped.AppendToString(&result);
     result += replacement_id;
   }
   text.AppendToString(&result);
   return result;
 }
 
 }  // namespace feedback