Python client for LLRP-based RFID readers
Python PowerShell Other
Latest commit 1ac4887 Sep 21, 2016 @ransford Merge branch 'release/0.1.6'
Bug fixes

README.md

sllurp

Build Status

IMPORTANT: This is beta-quality software that changes frequently. Use at your own risk.

This repository began in August 2013 as a snapshot of the dormant LLRPyC project's code on SourceForge. It has been updated to work with Impinj Speedway readers (and probably still other readers), and to provide a simple callback-based API to clients.

These readers are known to work well with sllurp, but it should be adaptable with not much effort to other LLRP-compatible readers:

  • Impinj Speedway (R1000)
  • Impinj Speedway Revolution (R220, R420)
  • Impinj Speedway xPortal
  • Motorola MC9190-Z (handheld)

File an issue on this GitHub project if you would like help getting another kind of reader to work.

sllurp is distributed under version 3 of the GNU General Public License. See LICENSE.txt for details.

Quick Start

Install a version of Twisted appropriate for your Python installation. Non-Windows users can simply pip install -r .pip-requirements.txt. Windows users must choose the appropriate .exe installer at the Twisted website, and install zope.interface with the same parameters.

To connect to a reader and perform EPC Gen 2 inventory for 10 seconds:

  1. Figure out your reader's IP address ip.add.re.ss
  2. bin/inventory ip.add.re.ss

Run bin/inventory -h to see options.

If the reader gets into a funny state because you're debugging against it, you can stop all ROSpecs by running bin/reset ip.add.re.ss.

(On Windows, substitute bin\windows\inventory.bat for the inventory script, and do the same for the reset script.)

Reader API

sllurp relies on Twisted for network interaction with the reader. To make a connection, create an LLRPClientFactory and hand it to Twisted:

# Minimal example; see inventory.py for more.
from sllurp import llrp
from twisted.internet import reactor
import logging
logging.getLogger().setLevel(logging.INFO)

def cb (tagReport):
    tags = tagReport.msgdict['RO_ACCESS_REPORT']['TagReportData']
    print 'tags:', tags

factory = llrp.LLRPClientFactory()
factory.addTagReportCallback(cb)
reactor.connectTCP('myreader', llrp.LLRP_PORT, factory)
reactor.run()

Getting More Information From Tag Reports

When initializing LLRPClientFactory, pass in tag_content_selector:

llrp.LLRPClientFactory(tag_content_selector={
    'EnableROSpecID': False,
    'EnableSpecIndex': False,
    'EnableInventoryParameterSpecID': False,
    'EnableAntennaID': True,
    'EnableChannelIndex': False,
    'EnablePeakRRSI': True,
    'EnableFirstSeenTimestamp': False,
    'EnableLastSeenTimestamp': True,
    'EnableTagSeenCount': True,
    'EnableAccessSpecID': False,
}

Logging

sllurp logs under the name sllurp, so if you wish to log its output, you can do this the application that imports sllurp:

sllurp_logger = logging.getLogger('sllurp')
sllurp_logger.setLevel(logging.DEBUG)
sllurp_logger.setHandler(logging.FileHandler('sllurp.log'))
# or .setHandler(logging.StreamHandler()) to log to stderr...

Handy Reader Commands

To see what inventory settings an Impinj reader is currently using (i.e., to fetch the current ROSpec), ssh to the reader and

> show rfid llrp rospec 0

You can dump the reader's entire configuration, including the current ROSpec, to a set of files by running bin/get_reader_config.

The "nuclear option" for resetting a reader is:

> reboot

Decoding EPC Data:

sllurp.epc contains EPC decoding tools. Read here for example usage.

If You Find a Bug

Start an issue on this GitHub project!

Bug reports are most useful when they're accompanied by verbose error messages. Turn sllurp's log level up to DEBUG, which you can do by specifying the -d command-line option if you're using the inventory or reset scripts. You can log to a logfile with the -l [filename] option. Or simply put this at the beginning of your own code:

import logger
sllurp_logger = logging.getLogger('sllurp')
sllurp_logger.setLevel(logging.DEBUG)

Contributing

Want to contribute? Here are some areas that need improvement:

  • Reduce redundancy in the encode_* and decode_* functions in llrp_proto.py.
  • Support the AccessSpec primitive (basis for tag read and write).
  • Write tests for common encoding and decoding tasks.
  • Make get_reader_config use the fabric library to connect to readers via SSH.
  • Generalize LLRP support beyond Impinj readers. Remove Impinj-specific assumptions.