From 3818c4eaf7434b92fd2fc1af4c3b422bd96cf3b3 Mon Sep 17 00:00:00 2001
From: Jan Dittberner <jan@dittberner.info>
Date: Fri, 22 Aug 2008 20:08:09 +0000
Subject: [PATCH] initial import of vlogsplitter

 * addresses #1
---
 vlogsplitter/vlogsplitter | 82 +++++++++++++++++++++++++++++++++++++++
 1 file changed, 82 insertions(+)
 create mode 100755 vlogsplitter/vlogsplitter

diff --git a/vlogsplitter/vlogsplitter b/vlogsplitter/vlogsplitter
new file mode 100755
index 0000000..7376861
--- /dev/null
+++ b/vlogsplitter/vlogsplitter
@@ -0,0 +1,82 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+"""
+This script is a replacement for split-logfile and rotatelogs from the
+Apache webserver distribution suited as a pipe log consumer. It has a
+file handle pool like vlogger.
+
+Copyright (c) 2008 Jan Dittberner <jan@dittberner.info>
+Version: $Id$
+"""
+import os
+import sys
+import time
+
+FHPOOL = {}
+FHPOOLSIZE = 100
+
+
+def getfromfhpool(key, dirname, filename):
+    """Get a file handle from the filehandle pool.
+
+    The function checks whether a filehandle for the given key is in
+    the pool and returns it. If there is no such handle the pool size
+    is checked and when it would be exceeded an LRU algorithm is used
+    to remove the oldest handle from the cache. When removing the item
+    from the cache the file is closed. After the pool size check the
+    file identified by dirname and filename is opened for appending
+    log date..
+    """
+    now = time.time()
+    if key in FHPOOL:
+        FHPOOL[key][0] = now
+        return FHPOOL[key][1]
+    if len(FHPOOL) == FHPOOLSIZE:
+        mintime = now
+        minkey = None
+        for (ckey, value) in FHPOOL.iteritems():
+            if value[0] <= mintime:
+                mintime = value[0]
+                minkey = ckey
+        FHPOOL[minkey][1].close()
+        del FHPOOL[minkey]
+    FHPOOL[key] = [now, open(os.path.join(dirname, filename), "a")]
+    return FHPOOL[key][1]
+
+
+def handlelogs(basedir, logfilename, inputdata):
+    """
+    This function controls the input parsing process.
+
+    It reads one log line at a time and adds it to the appropriate log
+    file.
+    """
+    while 1:
+        line = inputdata.readline()
+        if line.find(" ") == -1:
+            continue
+        (vhost, logentry) = line.split(" ", 1)
+        if vhost and logentry:
+            logdir = os.path.join(basedir, vhost)
+            if not os.path.isdir(logdir):
+                os.makedirs(logdir)
+            fhandle = getfromfhpool(vhost, logdir, logfilename)
+            fhandle.write(logentry)
+            fhandle.flush()
+
+
+def usage():
+    """Usage: %s <logbasedir> <logfilename>
+
+    Logs incoming lines in a format starting with a vhost field to
+    <logbasedir>/<vhost>/<logfilename>. Directories below <logbasedir>
+    are created automatically.
+    """
+    print usage.__doc__ % (sys.argv[0])
+
+
+if __name__ == '__main__':
+    if len(sys.argv) < 3:
+        usage()
+        sys.exit(1)
+    handlelogs(sys.argv[1], sys.argv[2], sys.stdin)