project evaluation statistics

src/unitgrade_private/pipelines/dtulearn.py

@@ -139,6 +139,8 @@ def docker_stagewise_evaluation(base_directory, Dockerfile=None, instructor_grad
                                 unmute_docker = True,
                                 plagiarism_check=False,
                                 accept_problems=False, # No!
+                                copydetect_check=False,
+                                slim_rs=False, # Slim the rs data structure that is returned.
                                 ):
     """
     This is the main verification scripts. It is the main entry point for project verifications as downloaded from DTU Learn.
@@ -708,6 +710,8 @@ def docker_stagewise_evaluation(base_directory, Dockerfile=None, instructor_grad
             else:
                 t, _ = load_token(tkns[0])
             rs[id] = {**rs[id], **t}
+            if slim_rs and 'sources' in rs[id]:
+                rs[id]['sources'] = "Sources have been removed from this token because slim_rs=True (see dtulearn.py)."
         return rs
 
     rs = _stage_report()
@@ -730,65 +734,87 @@ def docker_stagewise_evaluation(base_directory, Dockerfile=None, instructor_grad
     with open(base_directory +"/log.txt", "w") as f:
         f.write("\n".join(all_msgs))
 
-    # rs['messages'] = messages
-
-    # with open()
-
-    if plagiarism_check and True:
+    if plagiarism_check or copydetect_check:
         from unitgrade_private.plagiarism.mossit import moss_it2023
         moss_it2023(submissions_base_dir=stage4_dir, submissions_pattern="*-token", instructor_grade_script=instructor_grade_script,
-                    student_files_dir=student_handout_folder)
+                    student_files_dir=student_handout_folder, submit_to_server=not copydetect_check)
+        # Write the moss files.
 
-    if plagiarism_check and False: # This check is based on detector and is deprecated. I don't like detector.
+    if plagiarism_check and copydetect_check: # This check is based on detector and is deprecated. I don't like detector.
         from coursebox.core.info_paths import get_paths
         paths = get_paths()
         from copydetect import CopyDetector
-        detector = CopyDetector(extensions=["py"], display_t=0.7)
-        # rs['Group10']['sources'][0]['report_relative_location']
-        relatives = []
-        for id in rs:
-            v = [int(s.split("_")[-3]) for s in rs[id]['token_produced']]
-            token = rs[id]['token_produced'][v.index(max(v))]
-            tk, _ = load_token(token)
-            rl = tk['sources'][0]['report_relative_location']
-            bd = f"{stage3_dir}/{os.path.basename(os.path.dirname(token))}/{os.path.dirname(rl)}"
-            chk_files = []
-            for q in tk['details']:
-                # print(q)
-                with open(f"{bd}/unitgrade_data/{tk['details'][q]['name']}.pkl", 'rb') as f:
-                    pk = pickle.load(f)
-
-                for item in tk['details'][q]['items']:
-                    key = (item, 'coverage')
-                    if key in pk:
-                        for f in pk[key]:
-                            relatives.append(f)
-                            chk_files.append( f"{stage3_dir}/{os.path.basename(os.path.dirname(token))}/{f}")
-            # print(chk_files)
-            chk_files = list(set(chk_files))
-
-            for f in chk_files:
-                detector.add_file(f)
-            # Fuck this shit just add the solutions.
-            for f in set(relatives):
-                ff = paths['02450private']+"/Exam/exam2023spring/handout/" + f
-                detector.add_file(ff, type="boilerplate")
-
+        working_dir = os.path.dirname(stage4_dir) + "/moss"
+
+        # Use the plagiarism checker.
+        def read_all_chunk(bdir):
+            py_file = []
+            for f in sorted(glob.glob(bdir + "/*.py")):
+                with open(f, "r") as f:
+                    py_file.append(f.read())
+            return "\n\n".join(py_file)
+
+        copydetect_submissions_dir = working_dir + "/copydetect_submissions"
+        if os.path.isdir(copydetect_submissions_dir):
+            shutil.rmtree(copydetect_submissions_dir)
+        os.makedirs(copydetect_submissions_dir)
+
+        for bdir in glob.glob(working_dir + "/moss_submissions/*"):
+            os.makedirs(odir := copydetect_submissions_dir + "/" + os.path.basename(bdir))
+            with open(odir +"/student_code.py", "w") as f:
+                f.write(read_all_chunk(bdir))
+
+        copydetect_handout_dir = working_dir + "/copydetect_handout"
+        if os.path.isdir(copydetect_handout_dir):
+            shutil.rmtree(copydetect_handout_dir)
+        os.makedirs(copydetect_handout_dir)
+        with open(copydetect_handout_dir + "/student_code.py", "w") as f:
+            f.write(read_all_chunk(working_dir + "/handouts"))
+
+        test_dir_list = list(glob.glob(copydetect_submissions_dir + "/*"))
+
+        detector = CopyDetector(extensions=["py"], display_t=0.7, boilerplate_dirs=[copydetect_handout_dir], test_dirs=test_dir_list, same_name_only=True)
+        detector.out_file = working_dir + "/copydetect_report.html"
         detector.run()
         detector.generate_html_report()
-
-
-
-
-
-
-
-
-        # 799171 89568609
-        # _stage_plagiarism()
-
-        pass
-
+        # """
+        # file:///home/tuhe/Documents/02002instructors/project_evaluations_2023fall/project1/moss/handouts
+        # file:///home/tuhe/Documents/02002instructors/project_evaluations_2023fall/project1/moss/moss_submissions
+        #
+        # """
+        #
+        #
+        # detector = CopyDetector(same_name_only=True, extensions=["py"], display_t=0.7)
+        #
+        # relatives = []
+        # for id in rs:
+        #     v = [int(s.split("_")[-3]) for s in rs[id]['token_produced']]
+        #     token = rs[id]['token_produced'][v.index(max(v))]
+        #     tk, _ = load_token(token)
+        #     rl = tk['sources'][0]['report_relative_location']
+        #     bd = f"{stage3_dir}/{os.path.basename(os.path.dirname(token))}/{os.path.dirname(rl)}"
+        #     chk_files = []
+        #     for q in tk['details']:
+        #         # print(q)
+        #         with open(f"{bd}/unitgrade_data/{tk['details'][q]['name']}.pkl", 'rb') as f:
+        #             pk = pickle.load(f)
+        #
+        #         for item in tk['details'][q]['items']:
+        #             key = (item, 'coverage')
+        #             if key in pk:
+        #                 for f in pk[key]:
+        #                     relatives.append(f)
+        #                     chk_files.append( f"{stage3_dir}/{os.path.basename(os.path.dirname(token))}/{f}")
+        #
+        #     chk_files = list(set(chk_files))
+        #     for f in chk_files:
+        #         detector.add_file(f)
+        #     for f in set(relatives):
+        #         ff = paths['02450private']+"/Exam/exam2023spring/handout/" + f
+        #         detector.add_file(ff, type="boilerplate")
+        #
+        #     detector.run()
+        #     detector.generate_html_report()
 
     return rs
 

src/unitgrade_private/plagiarism/mossit.py

@@ -57,7 +57,7 @@ def get_id(moss_pl):
     return pl.split("=")[1][:-1]
 
 def moss_it2023(submissions_base_dir=None, submissions_pattern="*-token", whitelisted_tokens="", instructor_grade_script=None, moss_id=None,
-                student_files_dir=None):
+                student_files_dir=None, submit_to_server=True):
 
     a = 234
 
@@ -101,6 +101,8 @@ def moss_it2023(submissions_base_dir=None, submissions_pattern="*-token", whitel
                             for g in i:
                                 if os.path.isfile(student_file := f"{tmpdirname}/{g}"):
                                     shutil.copy(student_file, f"{sdir}/{os.path.basename(g)}")
+    if cov_files is None:
+        return
     if student_files_dir is not None:
 
         for q in cov_files:
@@ -128,14 +130,25 @@ def moss_it2023(submissions_base_dir=None, submissions_pattern="*-token", whitel
         print("You need to specify a moss id. You can do that by putting the moss.pl script at:", os.path.expanduser('~') + "/Documents/moss.pl")
         return
 
+    if submit_to_server:
         m = mosspy.Moss(moss_id, "python")
         for f in glob.glob(working_dir + "/handouts/**/*.py", recursive=True):
-        print(f)
+            print("Moss adding base file>", f)
             m.addBaseFile(f)
 
         m.addFilesByWildcard(working_dir + "/moss_submissions/*/*.py")
         print("> Calling moss with id", moss_id)
-    url = m.send(lambda file_path, display_name: print('*', end='', flush=True))
+        d = dict()
+        d['count'] = 0
+        def gcount():
+            d['count'] = d['count'] + 1
+            return d['count']
+        verbose = False
+        if verbose:
+            status_fun = lambda file_path, display_name: print("moss> " + str(gcount()) + ": " + file_path + " - " + display_name, flush=True)
+        else:
+            status_fun = lambda file_path, display_name: print("*", end='', flush=True)
+        url = m.send(status_fun)
         print()
         print("Report Url: " + url)
         r = working_dir + "/report/report.html"