logsplit.py: Add script to split apache logs per day.

Change-Id: I2f3489f7d6e880c5397b3519c02ed89f394ddbe4

1 files changed, 34 insertions, 0 deletions

diff --git a/logsplit.py b/logsplit.py
new file mode 100755
index 0000000..76a18ea
--- /dev/null
+++ b/logsplit.py
@@ -0,0 +1,34 @@
+#!/usr/bin/env python
+from __future__ import print_function
+import sys
+import re
+import datetime
+import gzip
+import fileinput
+
+
+def main(file_names):
+    fi = fileinput.FileInput(file_names, openhook=fileinput.hook_compressed)
+    f = None
+    last_date = None
+    for line in fi:
+        match = re.match(r".+?\[(.+?):.+?\]", line)
+        try:
+            date = match.group(1)
+            if date != last_date:
+                last_date = date
+                dt = datetime.datetime.strptime(date, "%d/%b/%Y")
+                fname = dt.strftime("access.log-%Y%m%d.gz")
+                if f is not None:
+                    f.close()
+                #f = open(fname, "w")
+                f = gzip.GzipFile(fname, "w")
+            f.write(line)
+        except:
+            print("Error context: %s:%d\n%s" % (fi.filename(), fi.filelineno(), line))
+            raise
+    f.close()
+
+
+if __name__ == "__main__":
+    main(sys.argv[1:])