|Subject:||Provide parsing mechanism for fractional number of seconds|
There's a %[number]N field for nanoseconds or microseconds or whatever, but if you have a number of seconds in decimal format with a variable number of decimal places (e.g. 12:34:56.789 PM), it can't be parsed directly. The workaround would be to scan through the data, find the maximum number of decimal places, pad to that length, then use %[number]N. This wouldn't be easy because the "scan through the data" part could require the user to write some pretty fancy parsing machinery. One solution could be to allow a dot in the %S field, a la %.S . Not sure whether there's a precedent for that or not in other parsing libraries.