Experimental

qtop_py/plugins/pbs.py

@@ -13,7 +13,7 @@ class PBSStatExtractor(StatExtractor):
     def __init__(self, config, options):
         StatExtractor.__init__(self, config, options)
         self.user_q_search = r'^(?P<host_name>(?P<job_id>[0-9\[\]-]+)\.(?P<domain>[\w-]+))\s+' \
-                             r'(?P<name>[\w@%.=+/{}-]+)\s+' \
+                             r'(?P<name>[\w@%:.=+/{}-]+)\s+' \
                              r'(?P<user>[A-Za-z0-9.]+)\s+' \
                              r'(?P<time>\d+:\d+:?\d*|0)\s+' \
                              r'(?P<state>[BCEFHMQRSTUWX])\s+' \
@@ -37,29 +37,21 @@ def extract_qstat(self, orig_file):
         except fileutils.FileEmptyError:
             logging.error('File %s seems to be empty.' % orig_file)
             all_qstat_values = []
-        else:
-            all_qstat_values = list()
-            with open(orig_file, 'r') as fin:
-                _ = fin.readline()  # header
-                fin.readline()
-                line = fin.readline()
-                re_match_positions = ('job_id', 'user', 'state', 'queue_name')  # was: (1, 5, 7, 8), (1, 4, 5, 8)
-                try:  # first qstat line determines which format qstat follows.
-                    re_search = self.user_q_search
-                    qstat_values = self._process_qstat_line(re_search, line, re_match_positions)
-                    # unused: _job_nr, _ce_name, _name, _time_use = m.group(2), m.group(3), m.group(4), m.group(6)
-                except AttributeError:  # this means 'prior' exists in qstat, it's another format
-                    re_search = self.user_q_search_prior
-                    qstat_values = self._process_qstat_line(re_search, line, re_match_positions)
-                    # unused:  _prior, _name, _submit, _start_at, _queue_domain, _slots, _ja_taskID =
-                    # m.group(2), m.group(3), m.group(6), m.group(7), m.group(9), m.group(10), m.group(11)
-                finally:
-                    all_qstat_values.append(qstat_values)
-
-                # hence the rest of the lines should follow either try's or except's same format
-                for line in fin:
-                    qstat_values = self._process_qstat_line(re_search, line, re_match_positions)
-                    all_qstat_values.append(qstat_values)
+
+        all_qstat_values = list()
+        with open(orig_file, 'r') as fin:
+            _ = fin.readline()  # header
+            fin.readline()  # unimportant
+            line = fin.readline()  # any qstat line determines which format qstat follows.
+            self.re_match_positions = ('job_id', 'user', 'state', 'queue_name')  # was: (1, 5, 7, 8), (1, 4, 5, 8)
+            re_search = self.decide_format(line)
+
+            qstat_values = self._process_qstat_line(re_search, line)
+            all_qstat_values.append(qstat_values)
+
+            for line in fin:
+                qstat_values = self._process_qstat_line(re_search, line)
+                all_qstat_values.append(qstat_values)
 
         return all_qstat_values
 

qtop_py/serialiser.py

@@ -15,7 +15,16 @@ def __init__(self, config, options):
         self.options = options
         self.anonymize = self.anonymize_func() if self.options.ANONYMIZE else self.eponymize_func()
 
-    def _process_qstat_line(self, re_search, line, re_match_positions):
+    def decide_format(self, line):
+        m1 = re.search(self.user_q_search, line.strip())
+        m2 = re.search(self.user_q_search_prior, line.strip())
+        if not any([m1, m2]):
+            logging.critical("Line: %s not matched to any known PBS format or unexpected character. Exiting...' % line.strip()")
+            sys.exit(1)
+
+        return self.user_q_search if m1 else self.user_q_search_prior  # else being m2
+
+    def _process_qstat_line(self, re_search, line):
         """
         extracts data from a tabular qstat-like file
         returns a list
@@ -24,15 +33,16 @@ def _process_qstat_line(self, re_search, line, re_match_positions):
         m = re.search(re_search, line.strip())
 
         try:
-            job_id, user, job_state, queue = [m.group(x) for x in re_match_positions]
+            job_id, user, job_state, queue = [m.group(x) for x in self.re_match_positions]
         except AttributeError:
-            logging.warn('Line: %s not properly parsed by regex expression. Assuming alternative qstat format.' % line.strip())
-            raise
-        job_id = job_id.split('.')[0]
-        user = self.anonymize(user, 'users')
-        for key, value in [('JobId', job_id), ('UnixAccount', user), ('S', job_state), ('Queue', queue)]:
-            qstat_values[key] = value
-        return qstat_values
+            logging.critical('Line:\n%s\n not properly parsed by regex expression. Unexpected character.' % line.strip())
+            sys.exit(1)
+        else:
+            job_id = job_id.split('.')[0]
+            user = self.anonymize(user, 'users')
+            for key, value in [('JobId', job_id), ('UnixAccount', user), ('S', job_state), ('Queue', queue)]:
+                qstat_values[key] = value
+            return qstat_values
 
     def anonymize_func(self):
         """