diff options
author | Prabhu Ramachandran | 2017-01-16 23:20:33 +0530 |
---|---|---|
committer | GitHub | 2017-01-16 23:20:33 +0530 |
commit | c3205b84ebd1796d98e140952802cdc0baad19a7 (patch) | |
tree | 4b63a64417776c95e12f2af75411075e6c0e7bd0 | |
parent | 899b2c4a347a6e7c4391dedce163a7ae0c4e8634 (diff) | |
parent | 987efbcba6e9976d1351a35454a62d6b8305009d (diff) | |
download | online_test-c3205b84ebd1796d98e140952802cdc0baad19a7.tar.gz online_test-c3205b84ebd1796d98e140952802cdc0baad19a7.tar.bz2 online_test-c3205b84ebd1796d98e140952802cdc0baad19a7.zip |
Merge pull request #176 from maheshgudi/hook_evaluator
Hook evaluator
-rw-r--r-- | yaksh/documentation/images/hook_testcase.jpg | bin | 0 -> 63419 bytes | |||
-rw-r--r-- | yaksh/documentation/moderator_docs/creating_question.rst | 18 | ||||
-rw-r--r-- | yaksh/evaluator_tests/test_bash_evaluation.py | 261 | ||||
-rw-r--r-- | yaksh/evaluator_tests/test_c_cpp_evaluation.py | 330 | ||||
-rw-r--r-- | yaksh/evaluator_tests/test_grader_evaluation.py (renamed from yaksh/evaluator_tests/test_code_evaluation.py) | 19 | ||||
-rw-r--r-- | yaksh/evaluator_tests/test_java_evaluation.py | 329 | ||||
-rw-r--r-- | yaksh/evaluator_tests/test_python_evaluation.py | 214 | ||||
-rw-r--r-- | yaksh/forms.py | 7 | ||||
-rw-r--r-- | yaksh/grader.py | 1 | ||||
-rw-r--r-- | yaksh/hook_evaluator.py | 67 | ||||
-rw-r--r-- | yaksh/models.py | 29 | ||||
-rw-r--r-- | yaksh/settings.py | 20 |
12 files changed, 1270 insertions, 25 deletions
diff --git a/yaksh/documentation/images/hook_testcase.jpg b/yaksh/documentation/images/hook_testcase.jpg Binary files differnew file mode 100644 index 0000000..3018050 --- /dev/null +++ b/yaksh/documentation/images/hook_testcase.jpg diff --git a/yaksh/documentation/moderator_docs/creating_question.rst b/yaksh/documentation/moderator_docs/creating_question.rst index ec273f5..f99bf7f 100644 --- a/yaksh/documentation/moderator_docs/creating_question.rst +++ b/yaksh/documentation/moderator_docs/creating_question.rst @@ -234,6 +234,24 @@ How to write Test cases For MCC based question, check the correct checkbox for multiple correct options. + * **Create Hook based Test Case** + + Select Hook from Add Test Case field. + + In Hook based test case type, moderator is provided with a evaluator function + called **check_answer** which is provided with a parameter called **user_answer**. + + **user_answer** is the code of the student in string format. + + A moderator can check the string for specific words in the user answer + and/or compile and execute the user answer (using standard python libraries) to + evaluate and hence return the mark fraction. + + + .. image:: ../images/hook_testcase.jpg + :width: 80% + + Features in Question -------------------- diff --git a/yaksh/evaluator_tests/test_bash_evaluation.py b/yaksh/evaluator_tests/test_bash_evaluation.py index 4b551d7..482d45e 100644 --- a/yaksh/evaluator_tests/test_bash_evaluation.py +++ b/yaksh/evaluator_tests/test_bash_evaluation.py @@ -269,5 +269,266 @@ class BashStdIOEvaluationTestCases(EvaluatorBaseTest): # Then self.assertTrue(result.get('success')) + +class BashHookEvaluationTestCases(EvaluatorBaseTest): + + def setUp(self): + self.f_path = os.path.join(tempfile.gettempdir(), "test.txt") + with open(self.f_path, 'wb') as f: + f.write('2'.encode('ascii')) + self.in_dir = tempfile.mkdtemp() + self.timeout_msg = ("Code took more than {0} seconds to run. " + "You probably have an infinite loop in your" + " code.").format(SERVER_TIMEOUT) + self.file_paths = None + + def tearDown(self): + os.remove(self.f_path) + shutil.rmtree(self.in_dir) + + def test_correct_answer(self): + # Given + user_answer = dedent(""" #!/bin/bash + echo -n Hello, world! + """ + ) + hook_code = dedent("""\ + def check_answer(user_answer): + import subprocess + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + proc = subprocess.Popen(user_answer, shell=True, + stdout=subprocess.PIPE, + stderr=subprocess.PIPE + ) + stdout,stderr = proc.communicate() + if stdout.decode("utf-8") == "Hello, world!": + success, err, mark_fraction = True, "", 1.0 + return success, err, mark_fraction + """ + ) + + test_case_data = [{"test_case_type": "hooktestcase", + "hook_code": hook_code,"weight": 1.0 + }] + kwargs = { + 'metadata': { + 'user_answer': user_answer, + 'file_paths': self.file_paths, + 'partial_grading': False, + 'language': 'bash' + }, + 'test_case_data': test_case_data, + } + + # When + grader = Grader(self.in_dir) + result = grader.evaluate(kwargs) + + # Then + self.assertTrue(result.get('success')) + + def test_incorrect_answer(self): + # Given + user_answer = dedent(""" #!/bin/bash + echo -n Goodbye, world! + """ + ) + hook_code = dedent("""\ + def check_answer(user_answer): + import subprocess + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + proc = subprocess.Popen(user_answer, shell=True, + stdout=subprocess.PIPE, + stderr=subprocess.PIPE + ) + stdout,stderr = proc.communicate() + if stdout.decode("utf-8") == "Hello, world!": + success, err, mark_fraction = True, "", 1.0 + return success, err, mark_fraction + """ + ) + + test_case_data = [{"test_case_type": "hooktestcase", + "hook_code": hook_code,"weight": 1.0 + }] + + kwargs = { + 'metadata': { + 'user_answer': user_answer, + 'file_paths': self.file_paths, + 'partial_grading': False, + 'language': 'bash' + }, + 'test_case_data': test_case_data, + } + + # When + grader = Grader(self.in_dir) + result = grader.evaluate(kwargs) + + # Then + self.assertFalse(result.get('success')) + self.assert_correct_output('Incorrect Answer', result.get('error')) + + def test_assert_with_hook(self): + # Given + user_answer = ("#!/bin/bash\n[[ $# -eq 2 ]]" + " && echo $(( $1 + $2 )) && exit $(( $1 + $2 ))" + ) + assert_test_case = dedent(""" + #!/bin/bash + [[ $# -eq 2 ]] && echo $(( $1 + $2 )) && exit $(( $1 + $2 )) + """) + + assert_test_case_args = "1 2\n2 1" + + hook_code = dedent("""\ + def check_answer(user_answer): + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + if "echo $(( $1 + $2 ))" in user_answer: + success, err, mark_fraction = True, "", 1.0 + return success, err, mark_fraction + """ + ) + + + test_case_data = [{"test_case_type": "standardtestcase", + "test_case": assert_test_case, + "test_case_args":assert_test_case_args, + 'weight': 1.0 + }, + {"test_case_type": "hooktestcase", + "hook_code": hook_code, 'weight': 1.0}, + ] + kwargs = { + 'metadata': { + 'user_answer': user_answer, + 'file_paths': self.file_paths, + 'partial_grading': True, + 'language': 'bash' + }, + 'test_case_data': test_case_data, + } + + # When + grader = Grader(self.in_dir) + result = grader.evaluate(kwargs) + + # Then + self.assertTrue(result.get('success')) + self.assertEqual(result.get("weight"), 2.0) + + def test_multiple_hooks(self): + # Given + user_answer = dedent(""" #!/bin/bash + echo -n Hello, world! + """ + ) + + hook_code_1 = dedent("""\ + def check_answer(user_answer): + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + if "echo -n Hello, world!" in user_answer: + success, err, mark_fraction = True, "", 0.5 + return success, err, mark_fraction + """ + ) + hook_code_2 = dedent("""\ + def check_answer(user_answer): + import subprocess + import sys + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + proc = subprocess.Popen(user_answer, shell=True, + stdout=subprocess.PIPE, + stderr=subprocess.PIPE + ) + stdout,stderr = proc.communicate() + + if stdout.decode('utf-8') == "Hello, world!": + success, err, mark_fraction = True, "", 1.0 + return success, err, mark_fraction + """ + ) + + + test_case_data = [{"test_case_type": "hooktestcase", + "hook_code": hook_code_1, 'weight': 1.0}, + {"test_case_type": "hooktestcase", + "hook_code": hook_code_2, 'weight': 1.0}, + ] + kwargs = { + 'metadata': { + 'user_answer': user_answer, + 'file_paths': self.file_paths, + 'partial_grading': True, + 'language': 'bash' + }, + 'test_case_data': test_case_data, + } + + # When + grader = Grader(self.in_dir) + result = grader.evaluate(kwargs) + + # Then + self.assertTrue(result.get('success')) + self.assertEqual(result.get("weight"), 1.5) + + def test_infinite_loop(self): + # Given + user_answer = ("#!/bin/bash\nwhile [ 1 ] ;" + " do echo "" > /dev/null ; done") + + hook_code = dedent("""\ + def check_answer(user_answer): + import subprocess + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + proc = subprocess.Popen(user_answer, shell=True, + stdout=subprocess.PIPE, + stderr=subprocess.PIPE + ) + stdout,stderr = proc.communicate() + if stdout.decode("utf-8") == "Hello, world!": + success, err, mark_fraction = True, "", 1.0 + return success, err, mark_fraction + """ + ) + + + test_case_data = [{"test_case_type": "hooktestcase", + "hook_code": hook_code,"weight": 1.0 + }] + + kwargs = { + 'metadata': { + 'user_answer': user_answer, + 'file_paths': self.file_paths, + 'partial_grading': False, + 'language': 'bash' + }, + 'test_case_data': test_case_data, + } + + # When + grader = Grader(self.in_dir) + result = grader.evaluate(kwargs) + + # Then + self.assertFalse(result.get('success')) + self.assert_correct_output(self.timeout_msg, result.get('error')) + + if __name__ == '__main__': unittest.main() diff --git a/yaksh/evaluator_tests/test_c_cpp_evaluation.py b/yaksh/evaluator_tests/test_c_cpp_evaluation.py index d734cf2..304f1cb 100644 --- a/yaksh/evaluator_tests/test_c_cpp_evaluation.py +++ b/yaksh/evaluator_tests/test_c_cpp_evaluation.py @@ -639,5 +639,335 @@ class CppStdIOEvaluationTestCases(EvaluatorBaseTest): # Then self.assertTrue(result.get('success')) +class CppHookEvaluationTestCases(EvaluatorBaseTest): + + def setUp(self): + self.f_path = os.path.join(tempfile.gettempdir(), "test.txt") + with open(self.f_path, 'wb') as f: + f.write('2'.encode('ascii')) + tmp_in_dir_path = tempfile.mkdtemp() + self.in_dir = tmp_in_dir_path + self.timeout_msg = ("Code took more than {0} seconds to run. " + "You probably have an infinite loop in your" + " code.").format(SERVER_TIMEOUT) + self.file_paths = None + + def tearDown(self): + os.remove(self.f_path) + shutil.rmtree(self.in_dir) + + def test_correct_answer(self): + # Given + user_answer = dedent("""\ + #include<stdio.h> + main() + { + printf("Hello, world!"); + } + """) + hook_code = dedent("""\ + def check_answer(user_answer): + with open("Test.c", "w+") as f: + f.write(user_answer) + import subprocess + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + def _run_command(cmd): + proc = subprocess.Popen("{}".format(cmd), + shell=True, + stdout=subprocess.PIPE, + stderr=subprocess.PIPE + ) + stdout,stderr = proc.communicate() + return stdout,stderr + cmds = ["gcc Test.c", "./a.out"] + for cmd in cmds: + stdout, stderr = _run_command(cmd) + if stdout.decode("utf-8") == "Hello, world!": + success, err, mark_fraction = True, "", 1.0 + return success, err, mark_fraction + """ + ) + + test_case_data = [{"test_case_type": "hooktestcase", + "hook_code": hook_code,"weight": 1.0 + }] + kwargs = { + 'metadata': { + 'user_answer': user_answer, + 'file_paths': self.file_paths, + 'partial_grading': False, + 'language': 'cpp' + }, + 'test_case_data': test_case_data, + } + + # When + grader = Grader(self.in_dir) + result = grader.evaluate(kwargs) + + # Then + self.assertTrue(result.get('success')) + + def test_incorrect_answer(self): + # Given + user_answer = dedent("""\ + #include<stdio.h> + main() + { + printf("Goodbye, world!"); + } + """) + hook_code = dedent("""\ + def check_answer(user_answer): + with open("Test.c", "w+") as f: + f.write(user_answer) + import subprocess + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + def _run_command(cmd): + proc = subprocess.Popen("{}".format(cmd), + shell=True, + stdout=subprocess.PIPE, + stderr=subprocess.PIPE + ) + stdout,stderr = proc.communicate() + return stdout,stderr + cmds = ["gcc Test.c", "./a.out"] + for cmd in cmds: + stdout, stderr = _run_command(cmd) + if stdout.decode("utf-8") == "Hello, world!": + success, err, mark_fraction = True, "", 1.0 + return success, err, mark_fraction + """ + ) + + test_case_data = [{"test_case_type": "hooktestcase", + "hook_code": hook_code,"weight": 1.0 + }] + kwargs = { + 'metadata': { + 'user_answer': user_answer, + 'file_paths': self.file_paths, + 'partial_grading': False, + 'language': 'cpp' + }, + 'test_case_data': test_case_data, + } + + # When + grader = Grader(self.in_dir) + result = grader.evaluate(kwargs) + + # Then + self.assertFalse(result.get('success')) + self.assert_correct_output('Incorrect Answer', result.get('error')) + + def test_assert_with_hook(self): + # Given + user_answer = "int add(int a, int b)\n{return a+b;}" + + + assert_test_case = dedent("""\ + #include <stdio.h> + #include <stdlib.h> + + extern int add(int, int); + + template <class T> + + void check(T expect, T result) + { + if (expect == result) + { + printf("Correct: Expected %d got %d ",expect,result); + } + else + { + printf("Incorrect: Expected %d got %d ",expect,result); + exit (1); + } + } + + int main(void) + { + int result; + result = add(0,0); + printf("Input submitted to the function: 0, 0"); + check(0, result); + result = add(2,3); + printf("Input submitted to the function: 2 3"); + check(5,result); + printf("All Correct"); + return 0; + } + """) + + hook_code = dedent("""\ + def check_answer(user_answer): + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + if "return a+b;" in user_answer: + success, err, mark_fraction = True, "", 1.0 + return success, err, mark_fraction + """ + ) + + + test_case_data = [{"test_case_type": "standardtestcase", + "test_case": assert_test_case, + 'weight': 1.0 + }, + {"test_case_type": "hooktestcase", + "hook_code": hook_code, 'weight': 1.0}, + ] + kwargs = { + 'metadata': { + 'user_answer': user_answer, + 'file_paths': self.file_paths, + 'partial_grading': True, + 'language': 'cpp' + }, + 'test_case_data': test_case_data, + } + + # When + grader = Grader(self.in_dir) + result = grader.evaluate(kwargs) + + # Then + self.assertTrue(result.get('success')) + self.assertEqual(result.get("weight"), 2.0) + + def test_multiple_hooks(self): + # Given + user_answer = dedent("""\ + #include<stdio.h> + main() + { + printf("Hello, world!"); + } + """) + + hook_code_1 = dedent("""\ + def check_answer(user_answer): + with open("Test.c", "w+") as f: + f.write(user_answer) + import subprocess + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + def _run_command(cmd): + proc = subprocess.Popen("{}".format(cmd), + shell=True, + stdout=subprocess.PIPE, + stderr=subprocess.PIPE + ) + stdout,stderr = proc.communicate() + return stdout,stderr + cmds = ["gcc Test.c", "./a.out"] + for cmd in cmds: + stdout, stderr = _run_command(cmd) + if stdout.decode("utf-8") == "Hello, world!": + success, err, mark_fraction = True, "", 1.0 + return success, err, mark_fraction + """ + ) + hook_code_2 = dedent("""\ + def check_answer(user_answer): + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + if 'printf("Hello, world!");' in user_answer: + success, err, mark_fraction = True, "", 0.5 + return success, err, mark_fraction + """ + ) + + + + test_case_data = [{"test_case_type": "hooktestcase", + "hook_code": hook_code_1, 'weight': 1.0}, + {"test_case_type": "hooktestcase", + "hook_code": hook_code_2, 'weight': 1.0}, + ] + kwargs = { + 'metadata': { + 'user_answer': user_answer, + 'file_paths': self.file_paths, + 'partial_grading': True, + 'language': 'cpp' + }, + 'test_case_data': test_case_data, + } + + # When + grader = Grader(self.in_dir) + result = grader.evaluate(kwargs) + + # Then + self.assertTrue(result.get('success')) + self.assertEqual(result.get("weight"), 1.5) + + def test_infinite_loop(self): + # Given + user_answer = dedent("""\ + #include<stdio.h> + int main(void){ + while(0==0){ + printf("abc");} + }""") + + hook_code= dedent("""\ + def check_answer(user_answer): + with open("Test.c", "w+") as f: + f.write(user_answer) + import subprocess + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + def _run_command(cmd): + proc = subprocess.Popen("{}".format(cmd), + shell=True, + stdout=subprocess.PIPE, + stderr=subprocess.PIPE + ) + stdout,stderr = proc.communicate() + return stdout,stderr + cmds = ["gcc Test.c", "./a.out"] + for cmd in cmds: + stdout, stderr = _run_command(cmd) + if stdout.decode("utf-8") == "Hello, world!": + success, err, mark_fraction = True, "", 1.0 + return success, err, mark_fraction + """ + ) + + test_case_data = [{"test_case_type": "hooktestcase", + "hook_code": hook_code,"weight": 1.0 + }] + + kwargs = { + 'metadata': { + 'user_answer': user_answer, + 'file_paths': self.file_paths, + 'partial_grading': False, + 'language': 'cpp' + }, + 'test_case_data': test_case_data, + } + + # When + grader = Grader(self.in_dir) + result = grader.evaluate(kwargs) + + # Then + self.assertFalse(result.get('success')) + self.assert_correct_output(self.timeout_msg, result.get('error')) + + if __name__ == '__main__': unittest.main() diff --git a/yaksh/evaluator_tests/test_code_evaluation.py b/yaksh/evaluator_tests/test_grader_evaluation.py index cb783b0..d11f4a0 100644 --- a/yaksh/evaluator_tests/test_code_evaluation.py +++ b/yaksh/evaluator_tests/test_grader_evaluation.py @@ -16,34 +16,27 @@ class RegistryTestCase(unittest.TestCase): stdio_evaluator_path = ("yaksh.python_stdio_evaluator." "PythonStdIOEvaluator" ) + + hook_evaluator_path = ("yaksh.hook_evaluator." + "HookEvaluator" + ) code_evaluators['python'] = \ {"standardtestcase": assertion_evaluator_path, - "stdiobasedtestcase": stdio_evaluator_path + "stdiobasedtestcase": stdio_evaluator_path, + "hooktestcase": hook_evaluator_path } def test_set_register(self): evaluator_class = self.registry_object.get_class("python", "standardtestcase" ) - assertion_evaluator_path = ("yaksh.python_assertion_evaluator" - ".PythonAssertionEvaluator" - ) - stdio_evaluator_path = ("yaksh.python_stdio_evaluator." - "PythonStdIOEvaluator" - ) class_name = getattr(python_assertion_evaluator, 'PythonAssertionEvaluator' ) - self.registry_object.register("python", - {"standardtestcase": assertion_evaluator_path, - "stdiobasedtestcase": stdio_evaluator_path - } - ) self.assertEqual(evaluator_class, class_name) def tearDown(self): self.registry_object = None - if __name__ == '__main__': unittest.main() diff --git a/yaksh/evaluator_tests/test_java_evaluation.py b/yaksh/evaluator_tests/test_java_evaluation.py index b53d8aa..3d127af 100644 --- a/yaksh/evaluator_tests/test_java_evaluation.py +++ b/yaksh/evaluator_tests/test_java_evaluation.py @@ -507,5 +507,334 @@ class JavaStdIOEvaluationTestCases(EvaluatorBaseTest): self.assertTrue(result.get("success")) +class JavaHookEvaluationTestCases(EvaluatorBaseTest): + + def setUp(self): + self.f_path = os.path.join(tempfile.gettempdir(), "test.txt") + with open(self.f_path, 'wb') as f: + f.write('2'.encode('ascii')) + tmp_in_dir_path = tempfile.mkdtemp() + self.in_dir = tmp_in_dir_path + self.file_paths = None + gd.SERVER_TIMEOUT = 9 + self.timeout_msg = ("Code took more than {0} seconds to run. " + "You probably have an infinite loop in" + " your code.").format(gd.SERVER_TIMEOUT) + + def tearDown(self): + gd.SERVER_TIMEOUT = 4 + os.remove(self.f_path) + shutil.rmtree(self.in_dir) + + def test_correct_answer(self): + # Given + user_answer = dedent("""\ + class Test + {public static void main(String[] args){ + System.out.print("Hello, world!"); + }} + """) + hook_code = dedent("""\ + def check_answer(user_answer): + with open("Test.java", "w+") as f: + f.write(user_answer) + import subprocess + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + def _run_command(cmd): + proc = subprocess.Popen("{}".format(cmd), + shell=True, + stdout=subprocess.PIPE, + stderr=subprocess.PIPE + ) + stdout,stderr = proc.communicate() + return stdout,stderr + cmds = ["javac Test.java", "java Test"] + for cmd in cmds: + stdout, stderr = _run_command(cmd) + if stdout.decode("utf-8") == "Hello, world!": + success, err, mark_fraction = True, "", 1.0 + return success, err, mark_fraction + """ + ) + + test_case_data = [{"test_case_type": "hooktestcase", + "hook_code": hook_code,"weight": 1.0 + }] + kwargs = { + 'metadata': { + 'user_answer': user_answer, + 'file_paths': self.file_paths, + 'partial_grading': False, + 'language': 'java' + }, + 'test_case_data': test_case_data, + } + + # When + grader = Grader(self.in_dir) + result = grader.evaluate(kwargs) + + # Then + self.assertTrue(result.get('success')) + + def test_incorrect_answer(self): + # Given + user_answer = dedent("""\ + class Test + {public static void main(String[] args){ + System.out.print("Goodbye, world!"); + }} + """) + hook_code = dedent("""\ + def check_answer(user_answer): + with open("Test.java", "w+") as f: + f.write(user_answer) + import subprocess + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + def _run_command(cmd): + proc = subprocess.Popen("{}".format(cmd), + shell=True, + stdout=subprocess.PIPE, + stderr=subprocess.PIPE + ) + stdout,stderr = proc.communicate() + return stdout,stderr + cmds = ["javac Test.java", "java Test"] + for cmd in cmds: + stdout, stderr = _run_command(cmd) + if stdout.decode("utf-8") == "Hello, world!": + success, err, mark_fraction = True, "", 1.0 + return success, err, mark_fraction + """ + ) + + test_case_data = [{"test_case_type": "hooktestcase", + "hook_code": hook_code,"weight": 1.0 + }] + kwargs = { + 'metadata': { + 'user_answer': user_answer, + 'file_paths': self.file_paths, + 'partial_grading': False, + 'language': 'java' + }, + 'test_case_data': test_case_data, + } + + # When + grader = Grader(self.in_dir) + result = grader.evaluate(kwargs) + + # Then + self.assertFalse(result.get('success')) + self.assert_correct_output('Incorrect Answer', result.get('error')) + + def test_assert_with_hook(self): + # Given + user_answer = "class Test {\n\tint square_num(int a) {\n\treturn a*a;\n\t}\n}" + assert_test_case = dedent(""" + class main + { + public static <E> void check(E expect, E result) + { + if(result.equals(expect)) + { + System.out.println("Correct:Output expected "+expect+" and got "+result); + } + else + { + System.out.println("Incorrect:Output expected "+expect+" but got "+result); + System.exit(1); + } + } + public static void main(String arg[]) + { + Test t = new Test(); + int result, input, output; + input = 0; output = 0; + result = t.square_num(input); + System.out.println("Input submitted to the function: "+input); + check(output, result); + input = 5; output = 25; + result = t.square_num(input); + System.out.println("Input submitted to the function: "+input); + check(output, result); + input = 6; output = 36; + result = t.square_num(input); + System.out.println("Input submitted to the function: "+input); + check(output, result); + } + } + """) + + hook_code = dedent("""\ + def check_answer(user_answer): + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + if "return a*a" in user_answer: + success, err, mark_fraction = True, "", 1.0 + return success, err, mark_fraction + """ + ) + + + test_case_data = [{"test_case_type": "standardtestcase", + "test_case": assert_test_case, + 'weight': 1.0 + }, + {"test_case_type": "hooktestcase", + "hook_code": hook_code, 'weight': 1.0}, + ] + kwargs = { + 'metadata': { + 'user_answer': user_answer, + 'file_paths': self.file_paths, + 'partial_grading': True, + 'language': 'java' + }, + 'test_case_data': test_case_data, + } + + # When + grader = Grader(self.in_dir) + result = grader.evaluate(kwargs) + + # Then + self.assertTrue(result.get('success')) + self.assertEqual(result.get("weight"), 2.0) + + def test_multiple_hooks(self): + # Given + user_answer = dedent("""\ + class Test + {public static void main(String[] args){ + System.out.print("Hello, world!"); + }} + """) + + hook_code_1 = dedent("""\ + def check_answer(user_answer): + with open("Test.java", "w+") as f: + f.write(user_answer) + import subprocess + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + def _run_command(cmd): + proc = subprocess.Popen("{}".format(cmd), + shell=True, + stdout=subprocess.PIPE, + stderr=subprocess.PIPE + ) + stdout,stderr = proc.communicate() + return stdout,stderr + cmds = ["javac Test.java", "java Test"] + for cmd in cmds: + stdout, stderr = _run_command(cmd) + if stdout.decode("utf-8") == "Hello, world!": + success, err, mark_fraction = True, "", 1.0 + return success, err, mark_fraction + """ + ) + + hook_code_2 = dedent("""\ + def check_answer(user_answer): + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + if 'System.out.print("Hello, world!");' in user_answer: + success, err, mark_fraction = True, "", 0.5 + return success, err, mark_fraction + """ + ) + + + test_case_data = [{"test_case_type": "hooktestcase", + "hook_code": hook_code_1, 'weight': 1.0}, + {"test_case_type": "hooktestcase", + "hook_code": hook_code_2, 'weight': 1.0}, + ] + kwargs = { + 'metadata': { + 'user_answer': user_answer, + 'file_paths': self.file_paths, + 'partial_grading': True, + 'language': 'java' + }, + 'test_case_data': test_case_data, + } + + # When + grader = Grader(self.in_dir) + result = grader.evaluate(kwargs) + + # Then + self.assertTrue(result.get('success')) + self.assertEqual(result.get("weight"), 1.5) + + def test_infinite_loop(self): + # Given + user_answer = dedent("""\ + class Test + {public static void main(String[] args){ + while(0==0) + { + System.out.print("a");} + }}""") + + hook_code = dedent("""\ + def check_answer(user_answer): + with open("Test.java", "w+") as f: + f.write(user_answer) + import subprocess + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + def _run_command(cmd): + proc = subprocess.Popen("{}".format(cmd), + shell=True, + stdout=subprocess.PIPE, + stderr=subprocess.PIPE + ) + stdout,stderr = proc.communicate() + return stdout,stderr + cmds = ["javac Test.java", "java Test"] + for cmd in cmds: + stdout, stderr = _run_command(cmd) + if stdout.decode("utf-8") == "Hello, world!": + success, err, mark_fraction = True, "", 1.0 + return success, err, mark_fraction + """ + ) + + + test_case_data = [{"test_case_type": "hooktestcase", + "hook_code": hook_code,"weight": 1.0 + }] + + kwargs = { + 'metadata': { + 'user_answer': user_answer, + 'file_paths': self.file_paths, + 'partial_grading': False, + 'language': 'java' + }, + 'test_case_data': test_case_data, + } + + # When + grader = Grader(self.in_dir) + result = grader.evaluate(kwargs) + + # Then + self.assertFalse(result.get('success')) + self.assert_correct_output(self.timeout_msg, result.get('error')) + + if __name__ == '__main__': unittest.main() diff --git a/yaksh/evaluator_tests/test_python_evaluation.py b/yaksh/evaluator_tests/test_python_evaluation.py index 43dfe6b..1fba73e 100644 --- a/yaksh/evaluator_tests/test_python_evaluation.py +++ b/yaksh/evaluator_tests/test_python_evaluation.py @@ -7,8 +7,6 @@ from textwrap import dedent # Local import from yaksh.grader import Grader -from yaksh.python_assertion_evaluator import PythonAssertionEvaluator -from yaksh.python_stdio_evaluator import PythonStdIOEvaluator from yaksh.settings import SERVER_TIMEOUT @@ -643,5 +641,217 @@ class PythonStdIOEvaluationTestCases(EvaluatorBaseTest): self.assertFalse(result.get('success')) +class PythonHookEvaluationTestCases(EvaluatorBaseTest): + + def setUp(self): + with open('/tmp/test.txt', 'wb') as f: + f.write('2'.encode('ascii')) + tmp_in_dir_path = tempfile.mkdtemp() + self.in_dir = tmp_in_dir_path + self.timeout_msg = ("Code took more than {0} seconds to run. " + "You probably have an infinite loop in" + " your code.").format(SERVER_TIMEOUT) + self.file_paths = None + + def tearDown(self): + os.remove('/tmp/test.txt') + shutil.rmtree(self.in_dir) + + def test_correct_answer(self): + # Given + user_answer = "def add(a,b):\n\treturn a + b" + hook_code = dedent("""\ + def check_answer(user_answer): + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + exec(user_answer, globals()) + if add(1,2) == 3: + success, err, mark_fraction = True, "", 1.0 + return success, err, mark_fraction + """ + ) + + test_case_data = [{"test_case_type": "hooktestcase", + "hook_code": hook_code,"weight": 1.0 + }] + kwargs = { + 'metadata': { + 'user_answer': user_answer, + 'file_paths': self.file_paths, + 'partial_grading': True, + 'language': 'python' + }, + 'test_case_data': test_case_data, + } + + # When + grader = Grader(self.in_dir) + result = grader.evaluate(kwargs) + + # Then + self.assertTrue(result.get('success')) + + def test_incorrect_answer(self): + # Given + user_answer = "def add(a,b):\n\treturn a - b" + hook_code = dedent("""\ + def check_answer(user_answer): + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + exec(user_answer, globals()) + if add(1,2) == 3: + success, err, mark_fraction = True, "", 1.0 + return success, err, mark_fraction + """ + ) + + test_case_data = [{"test_case_type": "hooktestcase", + "hook_code": hook_code,"weight": 1.0 + }] + + kwargs = { + 'metadata': { + 'user_answer': user_answer, + 'file_paths': self.file_paths, + 'partial_grading': False, + 'language': 'python' + }, + 'test_case_data': test_case_data, + } + + # When + grader = Grader(self.in_dir) + result = grader.evaluate(kwargs) + + # Then + self.assertFalse(result.get('success')) + self.assert_correct_output('Incorrect Answer', result.get('error')) + + def test_assert_with_hook(self): + # Given + user_answer = "def add(a,b):\n\treturn a + b" + assert_test_case = "assert add(1,2) == 3" + hook_code = dedent("""\ + def check_answer(user_answer): + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + if "return a + b" in user_answer: + success, err, mark_fraction = True, "", 1.0 + return success, err, mark_fraction + """ + ) + + test_case_data = [{"test_case_type": "standardtestcase", + "test_case": assert_test_case, 'weight': 1.0}, + {"test_case_type": "hooktestcase", + "hook_code": hook_code, 'weight': 1.0}, + ] + kwargs = { + 'metadata': { + 'user_answer': user_answer, + 'file_paths': self.file_paths, + 'partial_grading': True, + 'language': 'python' + }, + 'test_case_data': test_case_data, + } + + # When + grader = Grader(self.in_dir) + result = grader.evaluate(kwargs) + + # Then + self.assertTrue(result.get('success')) + self.assertEqual(result.get("weight"), 2.0) + + def test_multiple_hooks(self): + # Given + user_answer = "def add(a,b):\n\treturn a + b" + hook_code_1 = dedent("""\ + def check_answer(user_answer): + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + if "return a + b" in user_answer: + success, err, mark_fraction = True, "", 0.5 + return success, err, mark_fraction + """ + ) + hook_code_2 = dedent("""\ + def check_answer(user_answer): + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + exec(user_answer, globals()) + if add(1,2) == 3: + success, err, mark_fraction = True, "", 1.0 + return success, err, mark_fraction + """ + ) + + + test_case_data = [{"test_case_type": "hooktestcase", + "hook_code": hook_code_1, 'weight': 1.0}, + {"test_case_type": "hooktestcase", + "hook_code": hook_code_2, 'weight': 1.0}, + ] + kwargs = { + 'metadata': { + 'user_answer': user_answer, + 'file_paths': self.file_paths, + 'partial_grading': True, + 'language': 'python' + }, + 'test_case_data': test_case_data, + } + + # When + grader = Grader(self.in_dir) + result = grader.evaluate(kwargs) + + # Then + self.assertTrue(result.get('success')) + self.assertEqual(result.get("weight"), 1.5) + + def test_infinite_loop(self): + # Given + user_answer = "def add(a, b):\n\twhile True:\n\t\tpass" + hook_code = dedent("""\ + def check_answer(user_answer): + success = False + err = "Incorrect Answer" + mark_fraction = 0.0 + exec(user_answer, globals()) + if add(1,2) == 3: + success, err, mark_fraction = True, "", 1.0 + return success, err, mark_fraction + """ + ) + test_case_data = [{"test_case_type": "hooktestcase", + "hook_code": hook_code,"weight": 1.0 + }] + + kwargs = { + 'metadata': { + 'user_answer': user_answer, + 'file_paths': self.file_paths, + 'partial_grading': False, + 'language': 'python' + }, + 'test_case_data': test_case_data, + } + + # When + grader = Grader(self.in_dir) + result = grader.evaluate(kwargs) + + # Then + self.assertFalse(result.get('success')) + self.assert_correct_output(self.timeout_msg, result.get('error')) + + if __name__ == '__main__': unittest.main() diff --git a/yaksh/forms.py b/yaksh/forms.py index 1d18d29..8a90dee 100644 --- a/yaksh/forms.py +++ b/yaksh/forms.py @@ -1,6 +1,7 @@ from django import forms from yaksh.models import get_model_class, Profile, Quiz, Question, TestCase, Course,\ - QuestionPaper, StandardTestCase, StdIOBasedTestCase + QuestionPaper, StandardTestCase, StdIOBasedTestCase, \ + HookTestCase from django.contrib.auth import authenticate from django.contrib.auth.models import User from django.contrib.contenttypes.models import ContentType @@ -40,7 +41,8 @@ test_case_types = ( ("standardtestcase", "Standard Testcase"), ("stdiobasedtestcase", "StdIO Based Testcase"), ("mcqtestcase", "MCQ Testcase"), - ) + ("hooktestcase", "Hook Testcase"), + ) UNAME_CHARS = letters + "._" + digits PWD_CHARS = letters + punctuation + digits @@ -296,3 +298,4 @@ class QuestionPaperForm(forms.ModelForm): class Meta: model = QuestionPaper fields = ['shuffle_questions'] + diff --git a/yaksh/grader.py b/yaksh/grader.py index 086abb7..a9a3738 100644 --- a/yaksh/grader.py +++ b/yaksh/grader.py @@ -120,7 +120,6 @@ class Grader(object): for test_case in test_case_data: test_case_instance = create_evaluator_instance(metadata, test_case) test_case_instances.append(test_case_instance) - return test_case_instances diff --git a/yaksh/hook_evaluator.py b/yaksh/hook_evaluator.py new file mode 100644 index 0000000..3956da1 --- /dev/null +++ b/yaksh/hook_evaluator.py @@ -0,0 +1,67 @@ +#!/usr/bin/env python +from __future__ import unicode_literals +import sys +import traceback +import os + +# Local imports +from .file_utils import copy_files, delete_files +from .base_evaluator import BaseEvaluator +from .grader import TimeoutException + + +class HookEvaluator(BaseEvaluator): + def __init__(self, metadata, test_case_data): + self.files = [] + + # Set metadata values + self.user_answer = metadata.get('user_answer') + self.file_paths = metadata.get('file_paths') + self.partial_grading = metadata.get('partial_grading') + + # Set test case data values + self.hook_code = test_case_data.get('hook_code') + self.weight = test_case_data.get('weight') + + def teardown(self): + # Delete the created file. + if self.files: + delete_files(self.files) + + def check_code(self): + """ Function evaluates user answer by running a python based hook code + against it. + Returns + -------- + Returns a tuple (success, error, test_case_weight) + + success - Boolean, indicating if code was executed successfully, correctly + mark_fraction - Float, indicating fraction of the weight to a test case + error - String, error message if success is false + + returns (True, "Correct answer", 1.0) : If the student script passes all + test cases/have same output, when compared to the instructor script + + returns (False, error_msg, 0.0): If the student script fails a single + test/have dissimilar output, when compared to the instructor script. + + Returns (False, error_msg, 0.0): If mandatory arguments are not files or if + the required permissions are not given to the file(s). + """ + success = False + mark_fraction = 0.0 + try: + tb = None + _tests = compile(self.hook_code, '<string>', mode='exec') + hook_scope = {} + exec(_tests, hook_scope) + check = hook_scope["check_answer"] + success, err, mark_fraction = check(self.user_answer) + except TimeoutException: + raise + except Exception: + msg = traceback.format_exc(limit=0) + err = "Error in Hook code: {0}".format(msg) + del tb + return success, err, mark_fraction +
\ No newline at end of file diff --git a/yaksh/models.py b/yaksh/models.py index ca41885..ad61872 100644 --- a/yaksh/models.py +++ b/yaksh/models.py @@ -4,6 +4,7 @@ import json from random import sample, shuffle from itertools import islice, cycle from collections import Counter +from textwrap import dedent from django.db import models from django.db.models import Q from django.contrib.auth.models import User @@ -1219,5 +1220,31 @@ class McqTestCase(TestCase): class HookTestCase(TestCase): - code = models.TextField() + hook_code = models.TextField(default=dedent\ + ("""\ + def check_answer(user_answer): + ''' Evaluates user answer to return - + success - Boolean, indicating if code was executed correctly + mark_fraction - Float, indicating fraction of the + weight to a test case + error - String, error message if success is false''' + success = False + err = "Incorrect Answer" # Please make the error message more specific + mark_fraction = 0.0 + + # write your code here + + return success, err, mark_fraction + + """) + + ) weight = models.FloatField(default=1.0) + + def get_field_value(self): + return {"test_case_type": "hooktestcase", "hook_code": self.hook_code, + "weight": self.weight} + + def __str__(self): + return u'Hook Testcase | Correct: {0}'.format(self.hook_code) + diff --git a/yaksh/settings.py b/yaksh/settings.py index 0e432cf..72f9fda 100644 --- a/yaksh/settings.py +++ b/yaksh/settings.py @@ -21,18 +21,26 @@ URL_ROOT = '' code_evaluators = { "python": {"standardtestcase": "yaksh.python_assertion_evaluator.PythonAssertionEvaluator", - "stdiobasedtestcase": "yaksh.python_stdio_evaluator.PythonStdIOEvaluator" + "stdiobasedtestcase": "yaksh.python_stdio_evaluator.PythonStdIOEvaluator", + "hooktestcase": "yaksh.hook_evaluator.HookEvaluator" }, "c": {"standardtestcase": "yaksh.cpp_code_evaluator.CppCodeEvaluator", - "stdiobasedtestcase": "yaksh.cpp_stdio_evaluator.CppStdIOEvaluator" + "stdiobasedtestcase": "yaksh.cpp_stdio_evaluator.CppStdIOEvaluator", + "hooktestcase": "yaksh.hook_evaluator.HookEvaluator" }, "cpp": {"standardtestcase": "yaksh.cpp_code_evaluator.CppCodeEvaluator", - "stdiobasedtestcase": "yaksh.cpp_stdio_evaluator.CppStdIOEvaluator" + "stdiobasedtestcase": "yaksh.cpp_stdio_evaluator.CppStdIOEvaluator", + "hooktestcase": "yaksh.hook_evaluator.HookEvaluator" }, "java": {"standardtestcase": "yaksh.java_code_evaluator.JavaCodeEvaluator", - "stdiobasedtestcase": "yaksh.java_stdio_evaluator.JavaStdIOEvaluator"}, + "stdiobasedtestcase": "yaksh.java_stdio_evaluator.JavaStdIOEvaluator", + "hooktestcase": "yaksh.hook_evaluator.HookEvaluator" + }, "bash": {"standardtestcase": "yaksh.bash_code_evaluator.BashCodeEvaluator", - "stdiobasedtestcase": "yaksh.bash_stdio_evaluator.BashStdIOEvaluator" + "stdiobasedtestcase": "yaksh.bash_stdio_evaluator.BashStdIOEvaluator", + "hooktestcase": "yaksh.hook_evaluator.HookEvaluator" }, - "scilab": {"standardtestcase": "yaksh.scilab_code_evaluator.ScilabCodeEvaluator"}, + "scilab": {"standardtestcase": "yaksh.scilab_code_evaluator.ScilabCodeEvaluator", + "hooktestcase": "yaksh.hook_evaluator.HookEvaluator" + }, } |