args = parser.parse_args() # ---------------------------------------------------------------------------- if not args.quiet: setup_logging(1,None) # ---------------------------------------------------------------------------- # Automatic alignment is here: # ---------------------------------------------------------------------------- # Fix resources modelText = args.r # Acoustic model of the language of the text (required) modelSpk = args.R # Acoustic model of the mother language of the speaker (optional) a = sppasAlign( modelText, modelSpk ) # Fix options a.set_clean( True ) if args.noclean: a.set_clean( False ) a.set_basic( False ) if args.basic: a.set_basic( True ) a.set_activity_tier( True ) if args.noactivity: a.set_activity_tier( False ) a.set_phntokalign_tier( True )
def run_alignment(self, stepidx): """ Execute the SPPAS-Alignment program. """ # Initializations step = self.parameters.get_step(stepidx) stepname = self.parameters.get_step_name(stepidx) files_processed_success = 0 self._progress.set_header(stepname) self._progress.update(0,"") # Get the list of input file names, with the ".wav" (or ".wave") extension filelist = self.set_filelist(".wav")#,not_start=["track_"]) if len(filelist) == 0: return 0 total = len(filelist) # Create annotation instance try: a = sppasAlign( step.get_langresource(), logfile=self._logfile ) except Exception as e: if self._logfile is not None: self._logfile.print_message( "%s\n"%str(e), indent=1,status=4 ) return 0 # Execute the annotation for each file in the list for i,f in enumerate(filelist): # fix the default values a.fix_options( step.get_options() ) # Indicate the file to be processed self._progress.set_text( os.path.basename(f)+" ("+str(i+1)+"/"+str(total)+")" ) if self._logfile is not None: self._logfile.print_message(stepname+" of file " + f, indent=1 ) # Get the input file extt = ['-token'+self.parameters.get_output_format()] extp = ['-phon'+self.parameters.get_output_format()] for e in annotationdata.io.extensions_out: extt.append( '-token'+e ) extp.append( '-phon'+e ) extt.append('-chunks'+self.parameters.get_output_format()) extp.append('-chunks'+self.parameters.get_output_format()) inname = self._get_filename(f, extp) intok = self._get_filename(f, extt) if inname is not None: # Fix output file name outname = os.path.splitext(f)[0] + '-palign' + self.parameters.get_output_format() # Execute annotation try: a.run( inname, intok, f, outname ) except Exception as e: if self._logfile is not None: stre = unicode(e.message).encode("utf-8") self._logfile.print_message( "%s for file %s\n"%(stre,outname), indent=2,status=-1 ) else: files_processed_success += 1 if self._logfile is not None: self._logfile.print_message(outname, indent=2,status=0 ) else: if self._logfile is not None: self._logfile.print_message("Failed to find a file with phonetization. Read the documentation for details.",indent=2,status=2) # Indicate progress self._progress.set_fraction(float((i+1))/float(total)) if self._logfile is not None: self._logfile.print_newline() # Indicate completed! self._progress.update(1,"Completed (%d files successfully over %d files).\n"%(files_processed_success,total)) self._progress.set_header("") return files_processed_success