More test improvements.

.travis.yml

@@ -5,11 +5,16 @@ env:
     LD_PRELOAD=/lib/x86_64-linux-gnu/libSegFault.so
   matrix:
     - TOXENV=check
-    - TOXENV=py26
-    - TOXENV=py27
-    - TOXENV=py33
-    - TOXENV=py34
-    - TOXENV=pypy
+    - TOXENV=py26-pytest25
+    - TOXENV=py27-pytest25
+    - TOXENV=py33-pytest25
+    - TOXENV=py34-pytest25
+    - TOXENV=pypy-pytest25
+    - TOXENV=py26-pytest26
+    - TOXENV=py27-pytest26
+    - TOXENV=py33-pytest26
+    - TOXENV=py34-pytest26
+    - TOXENV=pypy-pytest26
 before_install:
   - python --version
   - uname -a

src/pytest_benchmark/stats.py

@@ -43,6 +43,6 @@ def stddev(self):
         return math.sqrt(self.variance)
 
     def __str__(self):
-        return "Stats[%s runs in %.4fsec, min=%.4fsec, max=%.4fsec, avg=%.4fsec, mean=%.4fsec, stddev=%.4fsec]" % (
-            self.runs, self.total, self.min, self.max, self.avg, self.mean, self.stddev
+        return "Stats[%s runs in %.4fsec, min=%.4fsec, max=%.4fsec, mean=%.4fsec, stddev=%.4fsec]" % (
+            self.runs, self.total, self.min, self.max, self.mean, self.stddev
         )

tests/test_no_run_broken.t

@@ -80,15 +80,15 @@ Test that we don't benchmark code that raises exceptions:
           def result():
               raise Exception()
   \s* (re)
-  tests.py:8:
-  _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
+  tests.py:.* (re)
+  (_ )+ (re)
   .*pytest_benchmark/plugin.py:.*: in __call__ (re)
       duration, scale = self._calibrate_timer(runner)
   .*pytest_benchmark/plugin.py:.*: in _calibrate_timer (re)
       duration = runner(loops)
   .*pytest_benchmark/plugin.py:.*: in runner (re)
       function_to_benchmark()
-  _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
+  (_ )+ (re)
   \s* (re)
       @benchmark
       def result():

tests/test_stats.py

@@ -5,14 +5,11 @@ def test_1():
     stats = RunningStats()
     for i in 4., 36., 45., 50., 75.:
         stats.update(i)
-    print(stats)
-    fail
+    assert stats.mean == 42.0
 
 def test_2():
     stats = RunningStats()
-
     stats.update(17.0)
     stats.update(19.0)
     stats.update(24.0)
-    print(stats)
-    fail
+    assert stats.mean == 20.0

tox.ini

@@ -1,7 +1,7 @@
 [tox]
 envlist =
     check,
-    {py26,py27,py33,py34,pypy},
+    {py26,py27,py33,py34,pypy}-{pytest25,pytest26},
     docs
 
 [testenv]
@@ -17,11 +17,14 @@ setenv =
     PYTHONUNBUFFERED=yes
 deps =
     cram==0.6
-    pytest==2.6.4
+    pytest25: pytest==2.5.2
+    pytest26: pytest==2.6.4
     pytest-cov==1.8.1
 
 testspath = {toxinidir}/tests/test_pytest_benchmark.py
-commands = {posargs:cram --verbose tests}
+commands =
+    {posargs:py.test tests}
+    cram --verbose tests
 usedevelop = true
 
 [testenv:spell]