added Email normalization preprocessors

ahmedkotb · Feb 13, 2012 · e7efd64 · e7efd64
1 parent dc9ef51
commit e7efd64
Showing 1 changed file with 17 additions and 0 deletions.
diff --git a/smart_email/src/preprocessors/EmailAddressesNormalization.java b/smart_email/src/preprocessors/EmailAddressesNormalization.java
@@ -0,0 +1,17 @@
+package preprocessors;
+
+import general.Email;
+
+public class EmailAddressesNormalization implements Preprocessor {
+
+	private static final String ADDRESS_TAG = " EMAILADDR ";
+
+	private static final String EMAIL_REGEX = ".+\\@.+\\..+";
+
+	@Override
+	public void apply(Email email) {
+		email.setSubject(email.getSubject().replaceAll(EMAIL_REGEX, ADDRESS_TAG));
+		email.setContent(email.getContent().replaceAll(EMAIL_REGEX, ADDRESS_TAG));
+	}
+
+}