Merge pull request #797 from anderspitman/move_find_features

Move find_features method into NucleotideSequence

CHANGELOG.md

@@ -5,6 +5,10 @@
 ### Features
 * Modified ``skbio.stats.distance.pwmantel`` to accept a list of filepaths. This is useful as it allows for a smaller amount of memory consumption as it only loads two matrices at a time as opposed to requiring that all distance matrices are loaded into memory.
 
+### Backward-incompatible changes
+* Removed ``feature_types`` attribute from ``BiologicalSequence`` and all subclasses ([#797](https://github.com/biocore/scikit-bio/pull/797)).
+* Removed ``find_features`` method from ``BiologicalSequence`` and ``ProteinSequence`` ([#797](https://github.com/biocore/scikit-bio/pull/797)).
+
 ## Version 0.2.2 (2014-12-04)
 
 ### Features

skbio/sequence/_sequence.py

@@ -88,8 +88,6 @@ class BiologicalSequence(Sequence, SkbioObject):
     """
     default_write_format = 'fasta'
 
-    feature_types = set([])
-
     @classmethod
     def alphabet(cls):
         """Return the set of characters allowed in a `BiologicalSequence`.
@@ -1503,46 +1501,6 @@ def regex_iter(self, regex, retrieve_group_0=False):
             for g in range(start, len(match.groups())+1):
                 yield (match.start(g), match.end(g), match.group(g))
 
-    def find_features(self, feature_type, min_length=1, allow_gaps=False):
-        """Search the sequence for features
-
-        Parameters
-        ----------
-        feature_type : str
-            The type of feature to find
-        min_length : int, optional
-            Defaults to 1. Only features at least as long as this will be
-            returned
-        allow_gaps : bool, optional
-            Defaults to ``False``. If ``True``, then gaps will not be
-            considered to disrupt a feature
-
-        Returns
-        -------
-        generator
-            Yields tuples of the start of the feature, the end of the feature,
-            and the subsequence that composes the feature
-        """
-        if feature_type not in self.feature_types:
-            raise ValueError("Unknown feature type: %s" % feature_type)
-
-        acceptable = '-' if allow_gaps else ''
-
-        if isinstance(self, NucleotideSequence):
-            if feature_type == 'purine_run':
-                pat_str = '([AGag%s]{%d,})' % (acceptable, min_length)
-            if feature_type == 'pyrimidine_run':
-                pat_str = '([CTUctu%s]{%d,})' % (acceptable, min_length)
-
-        pat = re.compile(pat_str)
-
-        for hits in self.regex_iter(pat):
-            if allow_gaps:
-                if len(hits[2].replace('-', '')) >= min_length:
-                    yield hits
-            else:
-                yield hits
-
 
 class NucleotideSequence(BiologicalSequence):
     """Base class for nucleotide sequences.
@@ -1561,8 +1519,6 @@ class NucleotideSequence(BiologicalSequence):
 
     """
 
-    feature_types = set(['purine_run', 'pyrimidine_run'])
-
     @classmethod
     def complement_map(cls):
         """Return the mapping of characters to their complements.
@@ -1743,6 +1699,62 @@ def reverse_complement(self):
         return self._complement(reverse=True)
     rc = reverse_complement
 
+    def find_features(self, feature_type, min_length=1, allow_gaps=False):
+        """Search the sequence for features
+
+        Parameters
+        ----------
+        feature_type : {'purine_run', 'pyrimidine_run'}
+            The type of feature to find
+        min_length : int, optional
+            Defaults to 1. Only features at least as long as this will be
+            returned
+        allow_gaps : bool, optional
+            Defaults to ``False``. If ``True``, then gaps will not be
+            considered to disrupt a feature
+
+        Returns
+        -------
+        generator
+            Yields tuples of the start of the feature, the end of the feature,
+            and the subsequence that composes the feature
+
+        Examples
+        --------
+        >>> from skbio.sequence import NucleotideSequence
+        >>> s = NucleotideSequence('G-AT.T')
+        >>> list(s.find_features('purine_run'))
+        [(0, 1, 'G'), (2, 3, 'A')]
+        >>> list(s.find_features('purine_run', 2))
+        []
+        >>> list(s.find_features('purine_run', 2, allow_gaps=True))
+        [(0, 3, 'G-A')]
+        >>> list(s.find_features('pyrimidine_run', 2, allow_gaps=True))
+        [(3, 6, 'T.T')]
+
+        """
+        gaps = re.escape(''.join(self.gap_alphabet()))
+        acceptable = gaps if allow_gaps else ''
+
+        if feature_type == 'purine_run':
+            pat_str = '([AGag%s]{%d,})' % (acceptable, min_length)
+        elif feature_type == 'pyrimidine_run':
+            pat_str = '([CTUctu%s]{%d,})' % (acceptable, min_length)
+        else:
+            raise ValueError("Unknown feature type: %s" % feature_type)
+
+        pat = re.compile(pat_str)
+
+        for hits in self.regex_iter(pat):
+            if allow_gaps:
+                degapped = hits[2]
+                for gap_char in self.gap_alphabet():
+                    degapped = degapped.replace(gap_char, '')
+                if len(degapped) >= min_length:
+                    yield hits
+            else:
+                yield hits
+
 
 class DNASequence(NucleotideSequence):
     """Base class for DNA sequences.

skbio/sequence/tests/test_sequence.py

@@ -779,10 +779,6 @@ def test_regex_iter(self):
         exp = [(2, 7, 'TTACA'), (2, 5, 'TTA'), (5, 7, 'CA')]
         self.assertEqual(obs, exp)
 
-    def test_find_features_nonexistent_feature_type(self):
-        with self.assertRaises(ValueError):
-            list(self.b1.find_features('purine_run'))
-
 
 class NucelotideSequenceTests(TestCase):
 
@@ -792,6 +788,7 @@ def setUp(self):
         self.b2 = NucleotideSequence(
             'ACCGGUACC', id="test-seq-2",
             description="A test sequence")
+        self.b3 = NucleotideSequence('G-AT-TG.AT.T')
 
     def test_alphabet(self):
         exp = {
@@ -951,6 +948,15 @@ def test_find_features_no_feature_type(self):
         with self.assertRaises(ValueError):
             list(self.b1.find_features('nonexistent_feature_type'))
 
+    def test_find_features_allow_gaps(self):
+        exp = [(0, 3, 'G-A'), (6, 9, 'G.A')]
+        obs = list(self.b3.find_features('purine_run', 2, True))
+        self.assertEqual(obs, exp)
+
+        exp = [(3, 6, 'T-T'), (9, 12, 'T.T')]
+        obs = list(self.b3.find_features('pyrimidine_run', 2, True))
+        self.assertEqual(obs, exp)
+
     def test_nondegenerates_propagate_optional_properties(self):
         seq = NucleotideSequence('RS', id='foo', description='bar',
                                  quality=[42, 999])