Development is still in its early stages, so feedback is welcome.
There are three supported operations: upload, status, and get.
use the 'upload' operation to submit an input file of target addresses to the service:
export MIDAR_API_KEY=<...> $ ./midar-api upload my-targets.addrs result ID: 5
The input file should have 1 address per line.
use 'status' to check on a run:
$ ./midar-api status 5 status: queued task name: (unnamed) submission date: Thu Jun 14 14:43:03 2018 $ ./midar-api status 5 status: finished task name: (unnamed) submission date: Thu Jun 14 17:05:32 2018 completion date: Thu Jun 14 17:11:29 2018 found aliases: False
The 'status' command really only tells you two things; 'queued' means successfully added to the job queue (input file was in the right format, etc.) or that the job is now running, and 'error' means the run failed for some reason. The 'found aliases' tells you whether there were any discovered alias sets.
use the 'get' operation to retrieve MIDAR alias sets on a successful run:
$ ./midar-api get 5
This writes out results to midar-5.sets or returns an HTTP 404 error.
In the resulting output, lines starting with '#' are comments and can be ignored with the exception of lines of the following form, which denotes the beginning of an alias set:
# set 0: 151 11325 0 0 0 0 0 0 0 11325The addresses in the alias set are then given one per line. All addresses in an alias set were tested and confirmed (multiple times) to share a counter with each other (in a full mesh).
Currently, a run of a few hundreds targets takes 10-15 mins. A run of 10k takes about 1-2 hours. The system currently won't allow you to do a MIDAR run with more than 100k targets.
Getting access to the MIDAR API
We currently limit access to academic researchers. Please send access requests, questions, or feedback to firstname.lastname@example.org using an email address from your institution (we don't accept access requests sent from gen eral email providers like gmail, yahoo, etc.