Adding changes from PR 763

NationalSecurityAgency · May 9, 2024 · 1b7c662 · 1b7c662
1 parent ed76e7c
commit 1b7c662
Show file tree

Hide file tree

Showing 2 changed files with 6 additions and 4 deletions.
diff --git a/src/main/java/emissary/util/FlexibleDateTimeParser.java b/src/main/java/emissary/util/FlexibleDateTimeParser.java
@@ -57,7 +57,7 @@ public final class FlexibleDateTimeParser {
      * Remove other junk -- anything in an html tag, all parenthesis and quotes, and any non-word characters at the
      * beginning or end
      */
-    private static final Pattern REMOVE = Pattern.compile("<.+?>|=0D$|\\(|\\)|\"|\\[|]|\\W+$|^\\W+", Pattern.DOTALL);
+    private static final Pattern REMOVE = Pattern.compile("<.+?>$|=0D$|\\(|\\)|\"|\\[|]|\\W+$|^\\W+", Pattern.DOTALL);
 
     private static final Pattern PHT_REPLACE = Pattern.compile("PHT");
     private static final String PT_TIMEZONE = "PT";
@@ -264,7 +264,8 @@ private static String cleanDateString(final String date) {
             return date;
         }
 
-        String cleanedDateString = date;
+        // date strings over 100 characters are more than likely invalid
+        String cleanedDateString = StringUtils.substring(date, 0, 100);
         cleanedDateString = REPLACE.matcher(cleanedDateString).replaceAll(SPACE);
         cleanedDateString = REMOVE.matcher(cleanedDateString).replaceAll(EMPTY);
 

diff --git a/src/test/java/emissary/util/FlexibleDateTimeParserTest.java b/src/test/java/emissary/util/FlexibleDateTimeParserTest.java
@@ -2,6 +2,7 @@
 
 import emissary.test.core.junit5.UnitTest;
 
+import org.apache.commons.lang3.RandomStringUtils;
 import org.junit.jupiter.api.Assertions;
 import org.junit.jupiter.api.BeforeAll;
 import org.junit.jupiter.api.Test;
@@ -622,15 +623,14 @@ void parse_yyyy_DDD() {
     @Test
     void testCleanDateString() {
         test("2016-01-04 18:20<br>", EXPECTED_NO_SECS, "HTML");
+        test("2016-01-04 18:20<br>br>", EXPECTED_NO_SECS, "HTML");
         test("2016-01-04\t\t18:20", EXPECTED_NO_SECS, "TABS");
         test("2016-01-04        18:20", EXPECTED_NO_SECS, "SPACES");
         test("2016-01-04 18:20=0D", EXPECTED_NO_SECS, "qp'ified ending");
         test("$$2016-01-04 18:20:00$$", EXPECTED_NO_SECS, "Extra characters at the beginning and end");
         test("2016-01-04 (18:20:00)", EXPECTED_NO_SECS, "Extra parenthesis");
         test("2016-01-04 18:20:00 [GMT]", EXPECTED_NO_SECS, "Extra brackets");
         test("\"Mon\", 4 Jan 2016 18:20 +0000 \"EST\"", EXPECTED_NO_SECS, "Extra quotes");
-        test("2016-01-04 18:20:00 +0000.1234", EXPECTED_NO_SECS, "");
-        test("2016-01-04 18:20:00", EXPECTED_NO_SECS, "");
     }
 
     @Test
@@ -641,5 +641,6 @@ void testBad() {
         assertNull(FlexibleDateTimeParser.parse("1234", Collections.singletonList(null)));
         test("17.Mar.2016", 0L, "UNKNOWN");
         test("Mon, 2 Feb 2017 06:20:30 PM +0000", 0L, "UNKNOWN");
+        test("2016:01:04 18:20:30 GMT+0000<" + RandomStringUtils.randomAlphanumeric(75) + ">", 0L, "UNKNOWN");
     }
 }