Merge pull request #88 from life4/small-fixes

Small fixes

Author: orsinium

tests/test_compression/test_entropy_ncd.py

@@ -31,7 +31,7 @@ def test_simmetry_compressor(text):
 @hypothesis.given(text=hypothesis.strategies.text(min_size=1))
 def test_idempotency_compressor(text):
     # I've modified idempotency to some kind of distributivity for constant.
-    # Now it indicates that compressor really compress.
+    # Now it indicates that compressor actually does compression.
     assert ALG._get_size(text * 2) < ALG._get_size(text) * 2
 
 

textdistance/algorithms/base.py

@@ -156,7 +156,7 @@ def _sum_counters(self, *sequences: Counter[T]) -> Counter[T]:
             result += s
         return result
 
-    def _count_counters(self, counter: Counter) -> float:
+    def _count_counters(self, counter: Counter) -> int:
         """Return all elements count from Counter
         """
         if getattr(self, 'as_set', False):

textdistance/algorithms/compression_based.py

@@ -108,8 +108,7 @@ def _make_probs(self, *sequences) -> dict[str, tuple[Fraction, Fraction]]:
 
         prob_pairs = {}
         cumulative_count = 0
-        counts = sorted(counts.items(), key=lambda x: (x[1], x[0]), reverse=True)
-        for char, current_count in counts:
+        for char, current_count in counts.most_common():
             prob_pairs[char] = (
                 Fraction(cumulative_count, total_letters),
                 Fraction(current_count, total_letters),
@@ -216,7 +215,7 @@ def _get_size(self, data: Sequence) -> float:
 class EntropyNCD(_NCDBase):
     """Entropy based NCD
 
-    Get Entropy of input secueance as a size of compressed data.
+    Get Entropy of input sequence as a size of compressed data.
 
     https://en.wikipedia.org/wiki/Entropy_(information_theory)
     https://en.wikipedia.org/wiki/Entropy_encoding

textdistance/algorithms/token_based.py

@@ -123,7 +123,7 @@ def __call__(self, *sequences: Sequence) -> float:
         sequences = [self._count_counters(s) for s in sequences]  # ints
         ks = list(islice(self.ks, len(sequences)))
 
-        if len(sequences) == 2 or self.bias is None:
+        if len(sequences) != 2 or self.bias is None:
             result = intersection
             for k, s in zip(ks, sequences):
                 result += k * (s - intersection)