Graham Gough wrote: > I use jpluckc to generate pdb files and would like to copy them to my > Palm only if they have changed since the last time they were plucked, > Unfortunately cmp doesn't do the trick, since, even if the essential > contents of the files are the same, the files are not identical > (presumably some sort of timestamp). Is there a simple way of checking > this, other than looking at the sizes of the files, which obviously is > not fool-proof?
Comparing the files won't help because the order of the records varies from conversion to conversion, even if the site is identical. This is because the spider is multithreaded and data records will rarely be added in exactly the same order because of varying network bandwidth. The way around this is to maintain a HashMap from within JPluck that maintains checksums for each individual record, sorting them alphabetically by URL. Then calculate a checksum based on that. This is very easy to implement, will try to add this tonight to JPluck 2. Regards -Laurens _______________________________________________ plucker-list mailing list [EMAIL PROTECTED] http://lists.rubberchicken.org/mailman/listinfo/plucker-list
