add unittest for render_pictures binary

tools/tests/base_unittest.py

Index: tools/tests/base_unittest.py
diff --git a/tools/tests/base_unittest.py b/tools/tests/base_unittest.py
new file mode 100755
index 0000000000000000000000000000000000000000..fe47a0c3f05366a886e3f1b2ca764360a68c3f52
--- /dev/null
+++ b/tools/tests/base_unittest.py
@@ -0,0 +1,81 @@
+#!/usr/bin/python
+
+"""
+Copyright 2014 Google Inc.
+
+Use of this source code is governed by a BSD-style license that can be
+found in the LICENSE file.
+
+A wrapper around the standard Python unittest library, adding features we need
+for various unittests within this directory.
+"""
+
+import os
+import subprocess
+import unittest
+
+
+class TestCase(unittest.TestCase):
+
+  def run_command(self, args):
+    """Runs a program from the command line and returns stdout.
+
+    Args:
+      args: Command line to run, as a list of string parameters. args[0] is the
+            binary to run.
+
+    Returns:
+      stdout from the program, as a single string.
+
+    Raises:
+      Exception: the program exited with a nonzero return code.
+    """
+    proc = subprocess.Popen(args,
+                            stdout=subprocess.PIPE,
+                            stderr=subprocess.PIPE)
+    (stdout, stderr) = proc.communicate()
+    if proc.returncode is not 0:
+      raise Exception('command "%s" failed: %s' % (args, stderr))
+    return stdout
+
+  def find_path_to_program(self, program):

[epoger, 2014/01/03 21:46:12]

I copied this function over from tools/test_rendering.py .  As I have noted
within that file, at some point the two files should be combined.

+    """Returns path to an existing program binary.
+
+    Args:
+      program: Basename of the program to find (e.g., 'render_pictures').
+
+    Returns:
+      Absolute path to the program binary, as a string.
+
+    Raises:
+      Exception: unable to find the program binary.
+    """
+    trunk_path = os.path.abspath(os.path.join(os.path.dirname(__file__),
+                                              os.pardir, os.pardir))
+    possible_paths = [os.path.join(trunk_path, 'out', 'Release', program),
+                      os.path.join(trunk_path, 'out', 'Debug', program),
+                      os.path.join(trunk_path, 'out', 'Release',
+                                   program + '.exe'),
+                      os.path.join(trunk_path, 'out', 'Debug',
+                                   program + '.exe')]
+    for try_path in possible_paths:
+      if os.path.isfile(try_path):
+        return try_path
+    raise Exception('cannot find %s in paths %s; maybe you need to '
+                    'build %s?' % (program, possible_paths, program))
+
+
+def main(test_case_class):
+  """Run the unit tests within the given class.
+
+  Raises an Exception if any of those tests fail (in case we are running in the
+  context of run_all.py, which depends on that Exception to signal failures).
+
+  TODO(epoger): Make all of our unit tests use the Python unittest framework,
+  so we can leverage its ability to run *all* the tests and report failures at
+  the end.
+  """
+  suite = unittest.TestLoader().loadTestsFromTestCase(test_case_class)
+  results = unittest.TextTestRunner(verbosity=2).run(suite)
+  if not results.wasSuccessful():
+    raise Exception('failed unittest %s' % test_case_class)