Improve warning value extraction performance in Response (#50208)

This commit improves the performance of warning value extraction in the
low-level REST client, and is similar to the approach taken in
be connected to Elasticsearch through a proxy that injects its own
warnings.

Author: darrenfoong

client/rest/src/main/java/org/elasticsearch/client/Response.java

@@ -114,16 +114,95 @@ public HttpEntity getEntity() {
             "GMT" + // GMT
             "\")?"); // closing quote (optional, since an older version can still send a warn-date)
 
+    /**
+     * Optimized regular expression to test if a string matches the RFC 1123 date
+     * format (with quotes and leading space). Start/end of line characters and
+     * atomic groups are used to prevent backtracking.
+     */
+    private static final Pattern WARNING_HEADER_DATE_PATTERN = Pattern.compile(
+            "^ " + // start of line, leading space
+            // quoted RFC 1123 date format
+            "\"" + // opening quote
+            "(?>Mon|Tue|Wed|Thu|Fri|Sat|Sun), " + // day of week, atomic group to prevent backtracking
+            "\\d{2} " + // 2-digit day
+            "(?>Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov|Dec) " + // month, atomic group to prevent backtracking
+            "\\d{4} " + // 4-digit year
+            "\\d{2}:\\d{2}:\\d{2} " + // (two-digit hour):(two-digit minute):(two-digit second)
+            "GMT" + // GMT
+            "\"$"); // closing quote (optional, since an older version can still send a warn-date), end of line
+
+    /**
+     * Length of RFC 1123 format (with quotes and leading space), used in
+     * matchWarningHeaderPatternByPrefix(String).
+     */
+    private static final int WARNING_HEADER_DATE_LENGTH = 0
+            + 1
+            + 1
+            + 3 + 1 + 1
+            + 2 + 1
+            + 3 + 1
+            + 4 + 1
+            + 2 + 1 + 2 + 1 + 2 + 1
+            + 3
+            + 1;
+
+    /**
+     * Tests if a string matches the RFC 7234 specification for warning headers.
+     * This assumes that the warn code is always 299 and the warn agent is always
+     * Elasticsearch.
+     *
+     * @param s the value of a warning header formatted according to RFC 7234
+     * @return {@code true} if the input string matches the specification
+     */
+    private static boolean matchWarningHeaderPatternByPrefix(final String s) {
+        return s.startsWith("299 Elasticsearch-");
+    }
+
+    /**
+     * Refer to org.elasticsearch.common.logging.DeprecationLogger
+     */
+    private static String extractWarningValueFromWarningHeader(final String s) {
+        String warningHeader = s;
+
+        /*
+         * The following block tests for the existence of a RFC 1123 date in the warning header. If the date exists, it is removed for
+         * extractWarningValueFromWarningHeader(String) to work properly (as it does not handle dates).
+         */
+        if (s.length() > WARNING_HEADER_DATE_LENGTH) {
+            final String possibleDateString = s.substring(s.length() - WARNING_HEADER_DATE_LENGTH);
+            final Matcher matcher = WARNING_HEADER_DATE_PATTERN.matcher(possibleDateString);
+
+            if (matcher.matches()) {
+                warningHeader = warningHeader.substring(0, s.length() - WARNING_HEADER_DATE_LENGTH);
+            }
+        }
+
+        final int firstQuote = warningHeader.indexOf('\"');
+        final int lastQuote = warningHeader.length() - 1;
+        final String warningValue = warningHeader.substring(firstQuote + 1, lastQuote);
+        assert assertWarningValue(s, warningValue);
+        return warningValue;
+    }
+
+    /**
+     * Refer to org.elasticsearch.common.logging.DeprecationLogger
+     */
+    private static boolean assertWarningValue(final String s, final String warningValue) {
+        final Matcher matcher = WARNING_HEADER_PATTERN.matcher(s);
+        final boolean matches = matcher.matches();
+        assert matches;
+        return matcher.group(1).equals(warningValue);
+    }
+
     /**
      * Returns a list of all warning headers returned in the response.
      */
     public List<String> getWarnings() {
         List<String> warnings = new ArrayList<>();
         for (Header header : response.getHeaders("Warning")) {
             String warning = header.getValue();
-            final Matcher matcher = WARNING_HEADER_PATTERN.matcher(warning);
-            if (matcher.matches()) {
-                warnings.add(matcher.group(1));
+            if (matchWarningHeaderPatternByPrefix(warning)) {
+                warnings.add(extractWarningValueFromWarningHeader(warning));
             } else {
                 warnings.add(warning);
             }

client/rest/src/test/java/org/elasticsearch/client/RestClientSingleHostTests.java

@@ -392,6 +392,7 @@ public void testHeaders() throws IOException {
     public void testDeprecationWarnings() throws IOException {
         String chars = randomAsciiAlphanumOfLength(5);
         assertDeprecationWarnings(singletonList("poorly formatted " + chars), singletonList("poorly formatted " + chars));
+        assertDeprecationWarnings(singletonList(formatWarningWithoutDate(chars)), singletonList(chars));
         assertDeprecationWarnings(singletonList(formatWarning(chars)), singletonList(chars));
         assertDeprecationWarnings(
                 Arrays.asList(formatWarning(chars), "another one", "and another"),
@@ -401,6 +402,9 @@ public void testDeprecationWarnings() throws IOException {
                 Arrays.asList("ignorable one", "and another"));
         assertDeprecationWarnings(singletonList("exact"), singletonList("exact"));
         assertDeprecationWarnings(Collections.<String>emptyList(), Collections.<String>emptyList());
+
+        String proxyWarning = "112 - \"network down\" \"Sat, 25 Aug 2012 23:34:45 GMT\"";
+        assertDeprecationWarnings(singletonList(proxyWarning), singletonList(proxyWarning));
     }
 
     private enum DeprecationWarningOption {
@@ -486,8 +490,12 @@ private void assertDeprecationWarnings(List<String> warningHeaderTexts, List<Str
      * Emulates Elasticsearch's DeprecationLogger.formatWarning in simple
      * cases. We don't have that available because we're testing against 1.7.
      */
+    private static String formatWarningWithoutDate(String warningBody) {
+            return "299 Elasticsearch-1.2.2-SNAPSHOT-eeeeeee \"" + warningBody + "\"";
+    }
+
     private static String formatWarning(String warningBody) {
-        return "299 Elasticsearch-1.2.2-SNAPSHOT-eeeeeee \"" + warningBody + "\" \"Mon, 01 Jan 2001 00:00:00 GMT\"";
+        return formatWarningWithoutDate(warningBody) + " \"Mon, 01 Jan 2001 00:00:00 GMT\"";
     }
 
     private HttpUriRequest performRandomRequest(String method) throws Exception {