9 from benderjab import rpc
11 from gaworkflow.pipeline.configure_run import *
12 from gaworkflow.pipeline.monitors import _percentCompleted
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 = "~/.gaworkflow"
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 fc,ft = status.statusFirecrest()
107 bc,bt = status.statusBustard()
108 gc,gt = status.statusGerald()
110 tc,tt = status.statusTotal()
112 fp = _percentCompleted(fc, ft)
113 bp = _percentCompleted(bc, bt)
114 gp = _percentCompleted(gc, gt)
115 tp = _percentCompleted(tc, tt)
119 output.append(u'Firecrest: %s%% (%s/%s)' % (fp, fc, ft))
120 output.append(u' Bustard: %s%% (%s/%s)' % (bp, bc, bt))
121 output.append(u' Gerald: %s%% (%s/%s)' % (gp, gc, gt))
122 output.append(u'-----------------------')
123 output.append(u' Total: %s%% (%s/%s)' % (tp, tc, tt))
125 return '\n'.join(output)
128 def sequencingFinished(self, run_dir):
130 Sequenceing (and copying) is finished, time to start pipeline
132 logging.debug("received sequencing finished message")
134 # Setup config info object
136 ci.base_analysis_dir = self.base_analysis_dir
137 ci.analysis_dir = os.path.join(self.base_analysis_dir, run_dir)
139 # get flowcell from run_dir name
140 flowcell = _get_flowcell_from_rundir(run_dir)
142 # Store ci object in dictionary
143 self.conf_info_dict[flowcell] = ci
146 # Launch the job in it's own thread and turn.
147 self.launchJob(run_dir, flowcell, ci)
150 def pipelineFinished(self, run_dir):
151 # need to strip off self.watch_dir from rundir I suspect.
152 logging.info("pipeline finished in" + str(run_dir))
153 #pattern = self.watch_dir
154 #if pattern[-1] != os.path.sep:
155 # pattern += os.path.sep
156 #stripped_run_dir = re.sub(pattern, "", run_dir)
157 #logging.debug("stripped to " + stripped_run_dir)
159 # Notify each user that the run has finished.
160 if self.notify_users is not None:
161 for u in self.notify_users:
162 self.send(u, 'Pipeline run %s finished' % (run_dir))
164 #if self.notify_runner is not None:
165 # for r in self.notify_runner:
166 # self.rpc_send(r, (stripped_run_dir,), 'sequencingFinished')
168 def reportMsg(self, msg):
170 if self.notify_users is not None:
171 for u in self.notify_users:
175 def _runner(self, run_dir, flowcell, conf_info):
177 # retrieve config step
178 cfg_filepath = os.path.join(conf_info.analysis_dir,
180 status_retrieve_cfg = retrieve_config(conf_info,
184 if status_retrieve_cfg:
185 logging.info("Runner: Retrieve config: success")
186 self.reportMsg("Retrieve config (%s): success" % (run_dir))
188 logging.error("Runner: Retrieve config: failed")
189 self.reportMsg("Retrieve config (%s): FAILED" % (run_dir))
193 if status_retrieve_cfg:
194 status = configure(conf_info)
196 logging.info("Runner: Configure: success")
197 self.reportMsg("Configure (%s): success" % (run_dir))
199 logging.error("Runner: Configure: failed")
200 self.reportMsg("Configure (%s): FAILED" % (run_dir))
202 #if successful, continue
204 # Setup status cmdline status monitor
205 #startCmdLineStatusMonitor(ci)
208 print 'Running pipeline now!'
209 run_status = run_pipeline(conf_info)
210 if run_status is True:
211 logging.info('Runner: Pipeline: success')
212 self.piplineFinished(run_dir)
214 logging.info('Runner: Pipeline: failed')
215 self.reportMsg("Pipeline run (%s): FAILED" % (run_dir))
218 def launchJob(self, run_dir, flowcell, conf_info):
220 Starts up a thread for running the pipeline
222 t = threading.Thread(target=self._runner,
223 args=[run_dir, flowcell, conf_info])
231 return bot.main(args)
233 if __name__ == "__main__":
234 sys.exit(main(sys.argv[1:]))