-This script isn't perfectly accurate, because the time stamps in
-Exim's log files are only accurate to a seond whereas internally Exim
-computes sender rates to the accuracy of your computer's clock
-(typically 10ms).
-
-The log files to be processed can be specified on the command line
-after the other arguments; if no filenames are specified the script
-will read from stdin.
-
-The first command line argument is the smoothing period, as defined by
-the documentation for the ratelimit ACL condition. The second argumetn
-is a regular expression.
-
-Each line is matched against the regular expression. Lines that do not
-match are ignored. The regex may contain 0, 1, or 2 () capturing
-sub-expressions.
-
-If there are no () sub-expressions, then every line that matches is
-used to compute a single rate. Its maximum value is reported when the
-script finishes.
-
-If there is one () sub-expression, then the text matched by the
-sub-expression is used to identify a rate lookup key, similar to the
-lookup key used by the ratelimit ACL condition. For example, you might
-write a regex to match the client IP address, or the authenticated
-username. Separate rates are computed for each different client and
-the maximum rate for each client is reported when the script finishes.
-
-If there are two () sub-expressions, then the text matched by the
-first sub-expression is used to identify a rate lookup key as above,
-and the second is used to match the message size recorded in the log
-line, e.g. " S=(\\d+) ". In this case the byte rate is computed instead
-of the message rate, similar to the per_byte option of the ratelimit
-ACL condition.
+options:
+
+-d Show debugging information to stderr
+-p Show progress of parse the log to stderr
+
+<period> The smoothing period in seconds, as defined by the
+ documentation for the ratelimit ACL condition.
+
+ This script isn't perfectly accurate, because the time
+ stamps in Exim's log files are only accurate to a second
+ whereas internally Exim computes sender rates to the
+ accuracy of your computer's clock (typically 10ms).
+
+<regex> The second argument is a regular expression.
+
+ Each line is matched against the regular expression.
+ Lines that do not match are ignored. The regex may
+ contain 0, 1, or 2 () capturing sub-expressions.
+
+ If there are no () sub-expressions, then every line that
+ matches is used to compute a single rate. Its maximum
+ value is reported when the script finishes.
+
+ If there is one () sub-expression, then the text matched
+ by the sub-expression is used to identify a rate lookup
+ key, similar to the lookup key used by the ratelimit
+ ACL condition. For example, you might write a regex
+ to match the client IP address, or the authenticated
+ username. Separate rates are computed for each different
+ client and the maximum rate for each client is reported
+ when the script finishes.
+
+ If there are two () sub-expressions, then the text matched
+ by the first sub-expression is used to identify a rate
+ lookup key as above, and the second is used to match the
+ message size recorded in the log line, e.g. "S=(\\d+)".
+ In this case the byte rate is computed instead of the
+ message rate, similar to the per_byte option of the
+ ratelimit ACL condition.
+
+<logfile> The log files to be processed can be specified on the
+ command line after the other arguments; if no filenames
+ are specified the script will read from stdin.
+
+examples:
+
+./ratelimit.pl 1 ' <= .*? \[(.*?)\]' <logfile>
+
+ Compute burst sending rate like ACL condition
+ ratelimit = 0 / 1s / strict / \$sender_host_address
+
+./ratelimit.pl 3600 '<= (.*?) ' <logfile>
+
+ Compute sending rate like ACL condition
+ ratelimit = 0 / 1h / strict / \$sender_address
+