add is_gs_url(), split_gs_url() static methods to gs_utils.py

py/utils/gs_utils_manualtest.py

Index: py/utils/gs_utils_manualtest.py
diff --git a/py/utils/gs_utils_manualtest.py b/py/utils/gs_utils_manualtest.py
index 37621d1e70cdcf6e27fa456c3bd00bb4b80dc2af..773071eaaadbf26bc2e9444e4e51fb33ba64b7f7 100755
--- a/py/utils/gs_utils_manualtest.py
+++ b/py/utils/gs_utils_manualtest.py
@@ -39,6 +39,36 @@ def _get_unique_posix_dir():
   return 'gs_utils_manualtest/%d' % random.randint(0, sys.maxint)
 
 
+def _test_static_methods():
+  """Test all static methods."""
+  gs = gs_utils.GSUtils
+
+  # (input url, output bucket, output path) for each test case
+  testcases = [
+      (None,             None, None),
+      (5,                None, None),
+      ('',               None, None),
+      ('/one/two',       None, None),
+      ('http://one/two', None, None),
+      ('gs:',            None, None),
+      ('gs://',          None, None),
+      ('gs:///',         None, None),
+      ('gs://???',       None, None),
+      ('gs:///bucket',   None, None),
+      ('gs://bucket',             'bucket', ''),
+      ('gs://bucket/',            'bucket', ''),
+      ('gs://bucket/path1',       'bucket', 'path1'),

[rmistry, 2014/07/29 20:36:20]

Could you also please add gs://bucket/path1/ and maybe gs://bucket/path1//

[epoger, 2014/07/30 01:10:49]

Done.

+      ('gs://bucket/path1/path2', 'bucket', 'path1/path2'),

[rmistry, 2014/07/29 20:36:20]

Optional Nit: Instead of aligning some of these lets not align any of them. I
always prefer not aligning because if we need to add any new test cases that are
longer than the longest one then we will have to change all of them instead of
it being a one-line change.

[epoger, 2014/07/30 01:10:49]

Yeah, that makes sense, and the style guide agrees with you, so you have THAT
going for you, which is nice...

Done.

+  ]
+  for (url, bucket, path) in testcases:
+    is_legal_url = (bucket != None)
+    assert gs.is_gs_url(url) == is_legal_url, 'gs.is_gs_url("%s") == %s' % (
+        url, is_legal_url)
+    if is_legal_url:
+      assert gs.split_gs_url(url) == (bucket, path), (
+          'gs.split_gs_url("%s") == ("%s", "%s")' % (url, bucket, path))
+
+
 def _test_public_read():
   """Make sure we can read from public files without .boto file credentials."""
   gs = gs_utils.GSUtils()
@@ -355,6 +385,7 @@ def _test_dir_upload_and_download():
 
 
 if __name__ == '__main__':
+  _test_static_methods()
   _test_upload_if_multiple_files()
   _test_upload_if_one_file()
   _test_public_read()