Merge pull request #15 from hayd/metrics

Metrics

Author: hayd

.gitignore

@@ -2,6 +2,7 @@
 *.dev*
 *.nja
 *.egg-info
+*.db
 
 .tox
 build
@@ -13,4 +14,4 @@ test/pattern_unittest_db
 pattern_unittest_db
 examples/06-graph/test/
 
-.DS_Store
+.DS_Store

.travis.yml

@@ -10,7 +10,6 @@ python:
 matrix:
     allow_failures:
         - python: "3.3"
-        - python: "3.4"
         - python: "pypy"
 
 install:
@@ -23,7 +22,11 @@ script:
     # test_05vector_07slp takes too long (so travis errors the build), it also fails on py2!
     # TODO make test_05vector_07slp run faster (or do slightly less).
     # TODO perhaps split build into tests and examples?
-    - nosetests --exclude=test_05vector_07slp --with-coverage --cover-package=pattern
+    # For now we only run the passing python 3 tests are run on the 3.4 build
+    - if [ "$TRAVIS_PYTHON_VERSION" == "3.4" ]; then
+      nosetests test/test_graph.py test/test_metrics.py; else
+      nosetests --exclude=test_05vector_07slp --with-coverage --cover-package=pattern;
+      fi
 
 after_success:
     - pip install --quiet coveralls

pattern/db/__init__.py

@@ -1552,7 +1552,7 @@ def SQL(self, **kwargs):
             >>>     FilterChain(
             >>>         filter("tail", True),
             >>>         filter("wing", True), operator=OR))
-            Yields: 
+            Yields:
             "type='pet' and weight between 4 and 6 and (tail=1 or wing=1)"
         """
         # Remember to pass the right escape() function as optional parameter.
@@ -1836,7 +1836,7 @@ def render(self, *path, **query):
         """ This method should be overwritten to return formatted table output (XML, HTML, RSS, ...)
             For web apps, the given path should list all parts in the relative URL path,
             and query is a dictionary of all POST and GET variables sent from the client.
-            For example: http://books.com/science/new 
+            For example: http://books.com/science/new
             => ["science", "new"]
             => render() data from db.books.filter(ALL, category="science", new=True).
         """
@@ -2226,7 +2226,8 @@ def load(cls, path, separator=",", decoder=lambda v: v, headers=False, preproces
         data = data if not password else StringIO(
             data.replace("\r\n", "\n").replace("\r", "\n"))
         data = data if not preprocess else StringIO(preprocess(data.read()))
-        data.seek(data.readline().startswith(BOM_UTF8) and len(BOM_UTF8) or 0)
+        bom_utf8 = BOM_UTF8 if sys.version < "3" else BOM_UTF8.decode("utf-8")
+        data.seek(data.readline().startswith(bom_utf8) and len(BOM_UTF8) or 0)
         data = csvlib.reader(data, delimiter=separator)
         i, n = kwargs.get("start"), kwargs.get("count")
         if i is not None and n is not None:

test/test_metrics.py

@@ -164,14 +164,18 @@ def test_suffixes(self):
         # Assert base => inflected and reversed inflected => base suffixes.
         s = [("beau", "beaux"), ("jeune", "jeunes"), ("hautain", "hautaines")]
         v = metrics.suffixes(s, n=3)
-        self.assertEqual(v, [
-            (2, "nes", [("ne", 0.5), ("n", 0.5)]),
-            (1, "aux", [("au", 1.0)])])
+        v[0][2].sort() # order is not well-defined in python 3
+        expected = [
+            (2, "nes", [("n", 0.5), ("ne", 0.5)]),
+            (1, "aux", [("au", 1.0)])]
+        self.assertEqual(v, expected)
+
         v = metrics.suffixes(s, n=2, reverse=False)
-        self.assertEqual(v, [
+        expected = [
             (1, "ne", [("nes", 1.0)]),
             (1, "in", [("ines", 1.0)]),
-            (1, "au", [("aux", 1.0)])])
+            (1, "au", [("aux", 1.0)])]
+        self.assertEqual(v, expected)
         print("pattern.metrics.suffixes()")
 
     def test_isplit(self):