Refactor and create __search_data to pay attention to gaps

dasscripts/tdms.py

@@ -20,6 +20,14 @@ def tup2time(fraction, seconds):
     return result
 
 
+class PotentialGap(Exception):
+    pass
+
+
+class NoData(Exception):
+    pass
+
+
 class TDMS(object):
     """Class to read and export seismic waveforms in TDMS format"""
 
@@ -30,7 +38,6 @@ class TDMS(object):
 
     def __enter__(self):
         """Method which allows to use the syntax 'with object:' and use it inside"""
-        # self.__select_file()
 
         # Create a buffer space to store the signal coefficients to be
         # convoluted during the decimation
@@ -190,19 +197,29 @@ class TDMS(object):
         self.__sampleend = None
         self.__samplecur = None
 
-        self.__select_file()
+        try:
+            self.__search_data()
+        except IndexError:
+            raise NoData()
+
 
     def __select_file(self):
         logs = logging.getLogger('Select file')
         logs.setLevel(self.__loglevel)
 
+        # print('select', self.__twstart, self.starttime, self.__currentfile)
+
         if self.__currentfile is None:
             for idx, fi in enumerate(self.__available):
+                # print(self.__twstart, fi['dt'])
                 if self.__twstart < fi['dt']:
                     if not idx:
                         raise Exception('Data not available in the specified time window')
                     filename = os.path.join(self.__directory, self.__available[idx-1]['name'])
                     self.__currentfile = idx-1
+                    # print(self.__currentfile, filename)
+                    logs.debug(
+                        'Opening %s; Starttime: %s' % (self.__available[self.__currentfile]['name'], self.__twstart))
                     break
             else:
                 raise Exception('Data not available in the specified time window')
@@ -218,8 +235,10 @@ class TDMS(object):
                 raise IndexError
             logs.debug('Opening %s; Starttime: %s' % (self.__available[self.__currentfile]['name'], self.__twstart))
 
+        # print(self.__twstart)
         # Reset some properties before opening the new file
         self.starttime = self.__available[self.__currentfile]['dt']
+        # print(self.starttime)
         self.endtime = None
         self.metadata = dict()
 
@@ -281,13 +300,33 @@ class TDMS(object):
         logs.debug('BigEndian: ' + ('yes' if self.__endian == '<' else 'no'))
         logs.debug('DAQmx raw data: ' + ('yes' if hasDAQmxRawData else 'no'))
 
+        # self.__readmetadata()
+
+    def __search_data(self):
+        """
+        :raise: IndexError
+        """
+
+        while True:
+            # Loop through files until there is nothing else (IndexError)
+            self.__select_file()
+
+            # Read the metadata and calculate samples to read
+            # Skip to the next file if there is a gap
+            try:
                 self.__readmetadata()
+                return
+            except PotentialGap:
+                logging.debug('Potential gap detected!')
+                self.__currentfile += 1
 
     def __resetcurrenttime(self):
         self.__twstart = self.__origstarttime
         self.__twend = self.__origendtime
         self.__currentfile = None
-        self.__select_file()
+        self.__search_data()
+
+        # print('reset', self.__twstart, self.starttime)
 
     def __readmetadata(self):
         # Metadata
@@ -375,10 +414,15 @@ class TDMS(object):
         self.endtime = self.starttime + datetime.timedelta(seconds=(self.__samples-1)/self.sampling_rate)
 
         # Sample to start extraction from based on the initial datetime of the file (__twstart)
+        # print(self.__twstart, self.starttime)
         self.__samplestart = max(floor((self.__twstart - self.starttime).total_seconds() * self.sampling_rate), 0)
+        if self.__samplestart >= self.__samples:
+            raise PotentialGap('Start reading at %s, but only %s samples' % (self.__samplestart, self.__samples))
+
         # Should I readjust __twstart to align it exactly with the time of the samples?
         # print(self.__twstart, self.starttime + datetime.timedelta(seconds=self.__samplestart/self.sampling_rate))
         self.__twstart = self.starttime + datetime.timedelta(seconds=self.__samplestart/self.sampling_rate)
+        # print(self.__twstart, self.starttime, self.__samplestart)
 
         self.__samplecur = self.__samplestart
 
@@ -546,6 +590,7 @@ class TDMS(object):
         # Data
         logs = logging.getLogger('Iterate Data')
 
+        # print(self.__currentfile, self.__twstart, self.starttime, self.__samplestart, self.__samplecur, self.__sampleend)
         # Data is stored with values from one channel in a continuous array
         if not self.__hasInterleavedData:
             for ch in range(self.__chstart, self.__chstop + 1, self.__chstep):
@@ -571,7 +616,7 @@ class TDMS(object):
                     # No more data in this file. Skip to the next one.
                     self.__currentfile += 1
                     try:
-                        self.__select_file()
+                        self.__search_data()
                     except IndexError:
                         break
         else:
@@ -600,7 +645,7 @@ class TDMS(object):
                 self.__currentfile += 1
                 try:
                     logs.debug('Moving to next file...')
-                    self.__select_file()
+                    self.__search_data()
                 except IndexError:
                     break
 
@@ -619,7 +664,7 @@ class TDMS(object):
             # No more data in this file. Skip to the next one.
             self.__currentfile += 1
             try:
-                self.__select_file()
+                self.__search_data()
             except IndexError:
                 break