10 from benderjab import rpc
12 from htsworkflow.pipelines.configure_run import *
14 #s_fc = re.compile('FC[0-9]+')
15 s_fc = re.compile('_[0-9a-zA-Z]*$')
18 def _get_flowcell_from_rundir(run_dir):
20 Returns flowcell string based on run_dir.
21 Returns None and logs error if flowcell can't be found.
23 junk, dirname = os.path.split(run_dir)
24 mo = s_fc.search(dirname)
26 logging.error('RunDir 2 FlowCell error: %s' % (run_dir))
29 return dirname[mo.start()+1:]
33 class Runner(rpc.XmlRpcBot):
35 Manage running pipeline jobs.
37 def __init__(self, section=None, configfile=None):
38 #if configfile is None:
39 # self.configfile = "~/.htsworkflow"
40 super(Runner, self).__init__(section, configfile)
42 self.cfg['notify_users'] = None
43 self.cfg['genome_dir'] = None
44 self.cfg['base_analysis_dir'] = None
46 self.cfg['notify_users'] = None
47 self.cfg['notify_postanalysis'] = None
49 self.conf_info_dict = {}
51 self.register_function(self.sequencingFinished)
52 #self.eventTasks.append(self.update)
55 def read_config(self, section=None, configfile=None):
56 super(Runner, self).read_config(section, configfile)
58 self.genome_dir = self._check_required_option('genome_dir')
59 self.base_analysis_dir = self._check_required_option('base_analysis_dir')
61 self.notify_users = self._parse_user_list(self.cfg['notify_users'])
62 #FIXME: process notify_postpipeline cfg
65 def _parser(self, msg, who):
67 Parse xmpp chat messages
69 help = u"I can send [start] a run, or report [status]"
70 if re.match(u"help", msg):
72 elif re.match("status", msg):
75 reply = self.getStatusReport(words[1])
77 reply = u"Status available for: %s" \
78 % (', '.join([k for k in self.conf_info_dict.keys()]))
79 elif re.match(u"start", msg):
82 self.sequencingFinished(words[1])
83 reply = u"starting run for %s" % (words[1])
85 reply = u"need runfolder name"
87 reply = u"I didn't understand '%s'" %(msg)
89 logging.debug("reply: " + str(reply))
93 def getStatusReport(self, fc_num):
95 Returns text status report for flow cell number
97 if fc_num not in self.conf_info_dict:
98 return "No record of a %s run." % (fc_num)
100 status = self.conf_info_dict[fc_num].status
103 return "No status information for %s yet." \
104 " Probably still in configure step. Try again later." % (fc_num)
106 output = status.statusReport()
108 return '\n'.join(output)
111 def sequencingFinished(self, run_dir):
113 Sequenceing (and copying) is finished, time to start pipeline
115 logging.debug("received sequencing finished message")
117 # Setup config info object
119 ci.base_analysis_dir = self.base_analysis_dir
120 ci.analysis_dir = os.path.join(self.base_analysis_dir, run_dir)
122 # get flowcell from run_dir name
123 flowcell = _get_flowcell_from_rundir(run_dir)
125 # Store ci object in dictionary
126 self.conf_info_dict[flowcell] = ci
129 # Launch the job in it's own thread and turn.
130 self.launchJob(run_dir, flowcell, ci)
134 def pipelineFinished(self, run_dir):
135 # need to strip off self.watch_dir from rundir I suspect.
136 logging.info("pipeline finished in" + str(run_dir))
137 #pattern = self.watch_dir
138 #if pattern[-1] != os.path.sep:
139 # pattern += os.path.sep
140 #stripped_run_dir = re.sub(pattern, "", run_dir)
141 #logging.debug("stripped to " + stripped_run_dir)
143 # Notify each user that the run has finished.
144 if self.notify_users is not None:
145 for u in self.notify_users:
146 self.send(u, 'Pipeline run %s finished' % (run_dir))
148 #if self.notify_runner is not None:
149 # for r in self.notify_runner:
150 # self.rpc_send(r, (stripped_run_dir,), 'sequencingFinished')
152 def reportMsg(self, msg):
154 if self.notify_users is not None:
155 for u in self.notify_users:
159 def _runner(self, run_dir, flowcell, conf_info):
161 # retrieve config step
162 cfg_filepath = os.path.join(conf_info.analysis_dir,
164 status_retrieve_cfg = retrieve_config(conf_info,
168 if status_retrieve_cfg:
169 logging.info("Runner: Retrieve config: success")
170 self.reportMsg("Retrieve config (%s): success" % (run_dir))
172 logging.error("Runner: Retrieve config: failed")
173 self.reportMsg("Retrieve config (%s): FAILED" % (run_dir))
177 if status_retrieve_cfg:
178 status = configure(conf_info)
180 logging.info("Runner: Configure: success")
181 self.reportMsg("Configure (%s): success" % (run_dir))
183 os.linesep.join(glob(os.path.join(run_dir,'Data','C*')))
186 logging.error("Runner: Configure: failed")
187 self.reportMsg("Configure (%s): FAILED" % (run_dir))
189 #if successful, continue
191 # Setup status cmdline status monitor
192 #startCmdLineStatusMonitor(ci)
195 print 'Running pipeline now!'
196 run_status = run_pipeline(conf_info)
197 if run_status is True:
198 logging.info('Runner: Pipeline: success')
199 self.reportMsg("Pipeline run (%s): Finished" % (run_dir,))
201 logging.info('Runner: Pipeline: failed')
202 self.reportMsg("Pipeline run (%s): FAILED" % (run_dir))
205 def launchJob(self, run_dir, flowcell, conf_info):
207 Starts up a thread for running the pipeline
209 t = threading.Thread(target=self._runner,
210 args=[run_dir, flowcell, conf_info])
218 return bot.main(args)
220 if __name__ == "__main__":
221 sys.exit(main(sys.argv[1:]))