Caching

Process caching

Once a job is cached, PyPPL will skip running this job. But you have to tell a process how to cache its jobs by setting pXXX.cache with a valid caching method:

Caching method (p.cache=?)	How
True	A signature* of input files, script and output files of a job is cached in <workdir>/<job.index>/job.cache, compare the signature before a job starts to run.
False	Disable caching, always run jobs.

Calculating signatures for caching

By default, PyPPL uses the last modified time to generate signatures for files and directories. However, for large directories, it may take notably long time to walk over all the files in those directories. If not necessary, you may simply as PyPPL to get the last modified time for the directories themselves instead of the infiles inside them by setting p.dirsig = False