Source Code for Module jazzparser.shell.gstools

  1  """Gold standard comparison shell tools 
  2   
  3  Some additional tools for the Jazz Parser interactive shell for  
  4  comparing results to a gold standard. 
  5   
  6  """ 
  7  """ 
  8  ============================== License ======================================== 
  9   Copyright (C) 2008, 2010-12 University of Edinburgh, Mark Granroth-Wilding 
 10    
 11   This file is part of The Jazz Parser. 
 12    
 13   The Jazz Parser is free software: you can redistribute it and/or modify 
 14   it under the terms of the GNU General Public License as published by 
 15   the Free Software Foundation, either version 3 of the License, or 
 16   (at your option) any later version. 
 17    
 18   The Jazz Parser is distributed in the hope that it will be useful, 
 19   but WITHOUT ANY WARRANTY; without even the implied warranty of 
 20   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the 
 21   GNU General Public License for more details. 
 22    
 23   You should have received a copy of the GNU General Public License 
 24   along with The Jazz Parser.  If not, see <http://www.gnu.org/licenses/>. 
 25   
 26  ============================ End license ====================================== 
 27   
 28  """ 
 29  __author__ = "Mark Granroth-Wilding <mark.granroth-wilding@ed.ac.uk>"  
 30   
 31  from .tools import Tool 
 32   
 33 -class LoadGoldStandardTool(Tool): 
 34      """ 
 35      Load an annotated chord sequence file. 
 36      """ 
 37      name = "Gold Standard" 
 38      commands = ['loadgs', 'lgs'] 
 39      usage = ("loadgs <filename>", "load an annotated data file for use by other GS tools") 
 40      help = """ 
 41  Given a filename of a db mirror data file, loads the data ready for  
 42  use by other GS comparison tools. 
 43  """ 
 44       
 45 -    def run(self, args, state): 
 46          from jazzparser.data.db_mirrors import SequenceIndex 
 47          from .shell import ShellError 
 48          if len(args) < 1: 
 49              raise ShellError, "Please specify a file to load" 
 50          filename = args[0] 
 51          # Load the data file 
 52          si = SequenceIndex.from_file(filename) 
 53          # Store it in the state 
 54          state.gs_sequences = si 
 55          print "Loaded %d gold standard sequences from %s" % (len(si),filename) 
 56   
 57 -class MarkCorrectTool(Tool): 
 58      name = "Mark Correct" 
 59      commands = ['markcorrect', 'mark'] 
 60      usage = ("markcorrect", "marks all the lexical entries in the "\ 
 61                  "chart that are correct according to the gold standard") 
 62      help = """ 
 63  Goes through the chart and marks those lexical signs that were the  
 64  correct choice according to the gold standard. Each correct sign has its  
 65  __str__ wrapped so it will display a marker of its correctness when  
 66  printed. 
 67  """ 
 68       
 69 -    def run(self, args, state): 
 70          from .shell import ShellError 
 71          # Only allow this to be run once 
 72          if hasattr(state, '_markcorrect_run'): 
 73              raise ShellError, "Correct entries have already been marked" 
 74          id = _get_seq_id(state) 
 75          _check_gs_has_sequence(state, id) 
 76          # We've got an id and the right GS sequence 
 77          gs_seq = state.gs_sequences.sequence_by_id(id) 
 78           
 79          # Decorate the str method of Sign so that we can put in an extra marker 
 80          def _mark_str(str_fn): 
 81              def _correct_str(obj): 
 82                  prepend = getattr(obj, '_str_prepend', '') 
 83                  return "%s%s" % (prepend, str_fn(obj)) 
 84              return _correct_str 
 85          state.formalism.Syntax.Sign.__str__ = _mark_str(state.formalism.Syntax.Sign.__str__) 
 86           
 87          # Go through each chord in the gold standard 
 88          for i,chord in enumerate(gs_seq.iterator()): 
 89              tag = chord.category 
 90              found = False 
 91              for sign in state.parser.chart.get_signs(i, i+1): 
 92                  # If the sign stores a tag we can check whether it was right 
 93                  if hasattr(sign, 'tag') and sign.tag == tag: 
 94                      found = True 
 95                      print "Found correct tag for position %d: %s" % (i,sign) 
 96                      sign._str_prepend = "***" 
 97              if not found: 
 98                  print "Correct tag not found for position %d" % i 
 99          state._markcorrect_run = True 
100   
101 -def _check_gs_loaded(state): 
102      from .shell import ShellError 
103      if not hasattr(state, 'gs_sequences'): 
104          raise ShellError, "The no gold standard has been loaded. Use the loadgs command to load a file." 
105       
106 -def _check_gs_has_sequence(state, id): 
107      _check_gs_loaded(state) 
108      if id not in state.gs_sequences.ids: 
109          raise ShellError, "The loaded gold standard doesn't include the sequence with id %s" % id 
110   
111 -def _get_seq_id(state): 
112      from .shell import ShellError 
113      if state.seq_id is None: 
114          raise ShellError, "Not sequence id available for parsed sequence: cannot run this tool" 
115      return state.seq_id 
116