Using filters

SPADE supports the use of filters for stream processing of provenance metadata. Multiple filters can be composed in a specified order. A list of currently configured filters is maintained. At the outset, this does not contain any filters.

---

Adding filters

To add a filter, its name must be specified along with the position in the list where it should be inserted. To add a new filter and make it the first to process incoming provenance, it should be inserted into position 1. For example:

-> add filter IORuns position=1
Adding filter IORuns... done

Another filter may then be added before the IORuns filter by also specifying 1 as the position:

-> add filter CycleAvoidance position=1
Adding filter CycleAvoidance... done

At this point, CycleAvoidance is located at index 1 and IORuns is located at index 2. Another filter can be added after CycleAvoidance and before IORuns by specifying 2 as the index:

-> add filter GraphFinesse position=2
Adding filter GraphFinesse... done

Based on the list maintained by SPADE, data provenance will first be transformed by CycleAvoidance, then GraphFinesse, and finally by IORuns, before being delivered to storage.

---

Removing filters

To remove a filter, specify its position:

-> remove filter 2
Removing filter GraphFinesse... done

---

Listing filters

To see the list of filters currently configured in SPADE, use the following:

-> list filters
2 filter(s) added:
    1. CycleAvoidance
    2. IORuns