search/test_match.py

   1 from translate.search import match
   2 from translate.storage import csvl10n
   3
   4 class TestMatch:
   5     """Test the matching class"""
   6     def candidatestrings(self, units):
   7         """returns only the candidate strings out of the list with (score, string) tuples"""
   8         return [unit.source for unit in units]
   9
  10     def buildcsv(self, sources, targets=None):
  11         """Build a csvfile store with the given source and target strings"""
  12         if targets is None:
  13             targets = sources
  14         else:
  15             assert len(sources) == len(targets)
  16         csvfile = csvl10n.csvfile()
  17         for source, target in zip(sources, targets):
  18             unit = csvfile.addsourceunit(source)
  19             unit.target = target
  20         return csvfile
  21
  22     def test_matching(self):
  23         """Test basic matching"""
  24         csvfile = self.buildcsv(["hand", "asdf", "fdas", "haas", "pond"])
  25         matcher = match.matcher(csvfile)
  26         candidates = self.candidatestrings(matcher.matches("hond"))
  27         candidates.sort()
  28         assert candidates == ["hand", "pond"]
  29         message = "Ek skop die bal"
  30         csvfile = self.buildcsv(
  31             ["Hy skop die bal",
  32             message,
  33             "Jannie skop die bal",
  34             "Ek skop die balle",
  35             "Niemand skop die bal nie"])
  36         matcher = match.matcher(csvfile)
  37         candidates = self.candidatestrings(matcher.matches(message))
  38         assert len(candidates) == 3
  39         #test that the 100% match is indeed first:
  40         assert candidates[0] == message
  41         candidates.sort()
  42         assert candidates[1:] == ["Ek skop die balle", "Hy skop die bal"]
  43
  44     def test_multiple_store(self):
  45         """Test using multiple datastores"""
  46         csvfile1 = self.buildcsv(["hand", "asdf", "fdas"])
  47         csvfile2 = self.buildcsv(["haas", "pond"])
  48         matcher = match.matcher([csvfile1, csvfile2])
  49         candidates = self.candidatestrings(matcher.matches("hond"))
  50         candidates.sort()
  51         assert candidates == ["hand", "pond"]
  52         message = "Ek skop die bal"
  53         csvfile1 = self.buildcsv(
  54             ["Hy skop die bal",
  55             message,
  56             "Jannie skop die bal"])
  57         csvfile2 = self.buildcsv(
  58             ["Ek skop die balle",
  59             "Niemand skop die bal nie"])
  60         matcher = match.matcher([csvfile1, csvfile2])
  61         candidates = self.candidatestrings(matcher.matches(message))
  62         assert len(candidates) == 3
  63         #test that the 100% match is indeed first:
  64         assert candidates[0] == message
  65         candidates.sort()
  66         assert candidates[1:] == ["Ek skop die balle", "Hy skop die bal"]
  67
  68     def test_extendtm(self):
  69         """Test that we can extend the TM after creation."""
  70         message = "Open file..."
  71         csvfile1 = self.buildcsv(["Close application", "Do something"])
  72         matcher = match.matcher([csvfile1])
  73         candidates = self.candidatestrings(matcher.matches(message))
  74         assert len(candidates) == 0
  75         csvfile2 = self.buildcsv(["Open file"])
  76         matcher.extendtm(csvfile2.units, store=csvfile2)
  77         candidates = self.candidatestrings(matcher.matches(message))
  78         assert len(candidates) == 1
  79         assert candidates[0] == "Open file"
  80
  81     def test_terminology(self):
  82         csvfile = self.buildcsv(["file", "computer", "directory"])
  83         matcher = match.terminologymatcher(csvfile)
  84         candidates = self.candidatestrings(matcher.matches("Copy the files from your computer"))
  85         candidates.sort()
  86         assert candidates == ["computer", "file"]
  87