From 152af0d74ac6154c1b76675f7c094a92aa1f0e67 Mon Sep 17 00:00:00 2001
From: "Christian P. MOMON" <christian.momon@devinsy.fr>
Date: Sat, 22 May 2021 01:11:04 +0200
Subject: [PATCH] Improved page metric computation in probing.

---
 .../metrics/http/HttpAccessLogAnalyzer.java   | 41 ++++++++++++++++++-
 1 file changed, 40 insertions(+), 1 deletion(-)

diff --git a/src/fr/devinsy/statoolinfos/metrics/http/HttpAccessLogAnalyzer.java b/src/fr/devinsy/statoolinfos/metrics/http/HttpAccessLogAnalyzer.java
index 1db92a8..fc85bb5 100644
--- a/src/fr/devinsy/statoolinfos/metrics/http/HttpAccessLogAnalyzer.java
+++ b/src/fr/devinsy/statoolinfos/metrics/http/HttpAccessLogAnalyzer.java
@@ -169,7 +169,7 @@ public class HttpAccessLogAnalyzer
             }
 
             // metrics.http.pages.* =
-            if (StringUtils.endsWithAny(log.getRequest(), ".html", ".htm", ".xhtml", ".cgi", "/"))
+            if ((isPage(log.getRequest())) && (log.getStatus().getCategory() == HttpStatusCategory.SUCCESS))
             {
                 this.counters.inc("metrics.http.pages", year, yearMonth, yearWeek, date);
             }
@@ -214,6 +214,45 @@ public class HttpAccessLogAnalyzer
         }
     }
 
+    /**
+     * Checks if is page.
+     *
+     * @param request
+     *            the request
+     * @return true, if is page
+     */
+    public static boolean isPage(final String request)
+    {
+        boolean result;
+
+        if (StringUtils.isBlank(request))
+        {
+            result = false;
+        }
+        else
+        {
+            Pattern PAGE_PATTERN = Pattern.compile("^.* (?<path>/[^ \\?]*)(\\?.*| .*)?$");
+
+            Matcher matcher = PAGE_PATTERN.matcher(request);
+            if (matcher.find())
+            {
+                String path = matcher.group("path").toLowerCase();
+
+                result = path.matches("^.*(/|/[^/]*\\.html|/[^/]*\\.htm|/[^/]*\\.xhtml|/[^/]*\\.cgi|/[^/]*\\.php|/[^/\\.]*[a-z0-9])$");
+
+                // System.out.println(result + " [" + path + "] [" + request +
+                // "]");
+            }
+            else
+            {
+                result = false;
+            }
+        }
+
+        //
+        return result;
+    }
+
     /**
      * Parses the log.
      *