Mark Roddy  committed 8404883

Added in depth code comments explaining how to define a test case, and comments explaining what the given test fixture is doing and why.

  • Participants
  • Parent commits bffda89

Comments (0)

Files changed (1)

File tests/

+The ExampleTest class in this module documents both how to define
+a test case for a Pig script, and common tasks you might like
+to perform in the course of writing tests.
+In order to create your own you'll need to define a class that
+inherits from squealer.PigTest (just like python's
+unittest.TestCase), specify the path to the script you're testing,
+and start writing tests.
+The squealer specific main() method is optional as PigTest cases
+will work with unittest's main().  However, it does include several
+nice features that make running test cases more enjoyable.  Also,
+the squealer main() will also run traditional unittest.TestCase
+based tests.
 from squealer import PigTest, main
+class ExampleTest(PigTest):
+    # Specify the path to the pig script that you're testing, this
+    # can be either an absolute path (if you hate your coworkers) or
+    # a path relative to where you'll be running your tests.
+    PigScript = "tests/pig-scripts/top_queries.pig"
-class ExampleTest(PigTest):
-    PigScript = "tests/pig-scripts/top_queries.pig"
+    # There are several values you can specify vai the Args dictionary,
+    # though the most common ones are parameters.  In this case the
+    # pig script under test uses parameters $input and $output.  This
+    # parameter is completely optional though.  
     Args = {
-        "n" : 3,
-        "reducers" : 1,
         "input" : "tests/data/top_queries_input_data.txt",
         "output" : "top_3_queries",
-    def testNoToN(self):
+    def testSimpleAssertionOfARelationsRecords(self):
+        """
+        Assert the expected set of records for one of the
+        relations in the pig script.
+        """
+        # Pig relations contain an interable of tuple objects,
+        # one for each record in the relation.  To construct
+        # the expected set of records, create a list of tuples
+        # with the expected data for each record.
         output = [
+        # To perform the actual assertion specify the name of the
+        # relation and pass in the expected records
         self.assertRelationEquals("queries_limit", output)