Develop solution to link GraceDB IDs to specific coinc.xml file paths on disk
Context
We would like to read data from pipeline coinc.xml triggers to produce things like PSD plots, SNR series plots, FAR plots, optimised skymap loading (for SeALGW), etc. Specifically, we would like to instantiate a IGWN Alert Consumer to read JSON payloads from the IGWN Alert topics, but this does not provide all the data that we want from the coinc.xml (i.e. we don't have PSD/SNR, and we don't have file paths to coinc.xml on disk). There is a "files"
key that we can use in the payload to download the coinc.xml, but we still can't reliably match this back to a disk on file in a programmatic manner.
Example Payload from IGWN Alert
See attached file for the example content of a payload.json file: payload.json.
Objectives
We would like to come up with a solution for IGWN Alert Consumers to load the coinc.xml file on disk - and so we need appropriate metadata to map GraceDb ID/IGWN Alert payloads back to files. The plan is approximately as follows:
- We will re-download coinc.xml directly from GraceDb for now as the consumer won't know the exact coinc.xml to read from disk with the information available
- The next reviewed version of SPIIR should implement a means for someone to reverse engineer which file on local disk corresponds to GraceDb ID, for convenient post-processing & analysis