1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
|
import xml.etree.ElementTree as etree
from dataclasses import dataclass
from functools import reduce
from operator import getitem
import subprocess
import importlib
import traceback
import tempfile
import inspect
import pkgutil
import shutil
import jinja2
import sys
import os
import re
@dataclass
class Reflect:
client_code_path:str
imported_modules = {}
def __post_init__(self):
self.client_code_path = os.path.normpath(self.client_code_path)
sys.path.insert(1, self.client_code_path)
self.client_modules = [p for p in pkgutil.iter_modules() if os.path.normpath(str(p[0])[12:-2]) == self.client_code_path]
# print("client moduules ", self.client_modules)
def import_module(self, module_name):
"""Imports a module. Before reflection can be conducted, a module
must be imported. WARNING: This will execute module code if it isn't
in a if __name__ == "__main__". Takes a module name (that the student made)
as the first argument.
Args:
module_name (str): The name of a student's module to import
"""
for module in self.client_modules:
if module.name == module_name:
try:
self.imported_modules[module_name] = importlib.import_module(module.name)
except ModuleNotFoundError as e:
print("Missing library dependency for client module:")
print(e)
exit()
# except Exception as e:
# print("CRITICAL ERROR IN CLIENT CODE - CANNOT CONTINUE")
# raise ClientCodeException(e)
def get_module_doc(self, module_name):
"""Gets the documentation provided for a module.
Args:
module_name (str): The student's module name to get documentation for
Returns:
str: Provided documentation
"""
return {
"comments": self.__format_doc(inspect.getcomments(self.imported_modules[module_name])),
"doc": self.__format_doc(inspect.getdoc(self.imported_modules[module_name]))
}
def get_classes(self, module_name):
"""Gets the classes in a given module. The module must be imported first.
Args:
module_name (str): The name of an imported module to get the name of.
Returns:
dict: Dictionary of classes. The name of the class is the index, followed by
a tuple containing the class object and the classes' documentation.
"""
return {
i[0]: (i[1], {"comments": self.__format_doc(inspect.getcomments(i[1])), "doc": self.__format_doc(inspect.getdoc(i[1]))})
for i in inspect.getmembers(self.imported_modules[module_name])
if inspect.isclass(i[1]) and self.get_class_full_name(i[1]).split(".")[0] in self.imported_modules.keys()
}
def get_class_methods(self, module_name, class_name):
"""Gets the user generated methods of a given class. The module must be imported first.
Args:
module_name (str): The name of the module in which the class is contained.
class_name (str): The name of the class.
Returns:
dict: A dictionary of the methods. The index is the function name, followed by a tuple
containing the function object, the documentation, and the args as a nicely formatted string.
"""
return {
i[0]: (
i[1],
{"comments": self.__format_doc(inspect.getcomments(i[1])), "doc": self.__format_doc(inspect.getdoc(i[1]))},
str(inspect.signature(i[1])),
inspect.getsource(i[1]).rstrip()
)
for i in inspect.getmembers(
self.get_classes(module_name)[class_name][0],
predicate=inspect.isfunction
)
}
def get_functions(self, module_name):
return {
i[0]: (
i[1],
{"comments": self.__format_doc(inspect.getcomments(i[1])), "doc": self.__format_doc(inspect.getdoc(i[1]))},
str(inspect.signature(i[1])),
inspect.getsource(i[1]).rstrip()
)
for i in inspect.getmembers(self.imported_modules[module_name])
if inspect.isfunction(i[1])
}
def get_class_full_name(self, class_):
"""Returns the name of a class object as a nice string. e.g. modulename.classname
except if it's a builtin there'll be no module name.
Args:
class_ (class): A class to get the name of
Returns:
str: A nicely formatted class name.
"""
if class_.__module__ in ['builtins', 'exceptions']:
return class_.__name__
return "%s.%s" % (class_.__module__, class_.__name__)
# classes that inherit from two classes doesn't print out nicely here.
# using this method is better https://pastebin.com/YuxkkTkv
def get_class_tree(self):
"""Generates a dictionary based tree structure showing inheritance of classes
of all the *imported modules*. WARNING: It doesn't deal well with multiple inheritance..
Read the comments.
"""
def expand(a:list):
out = []
for l in a:
for i in reversed(range(0, len(l))):
out.append(l[:len(l) - i])
return out
# https://www.geeksforgeeks.org/python-convert-a-list-of-lists-into-tree-like-dict/
def getTree(tree, mappings):
return reduce(getitem, mappings, tree)
# https://www.geeksforgeeks.org/python-convert-a-list-of-lists-into-tree-like-dict/
def setTree(tree, mappings):
getTree(tree, mappings[:-1])[mappings[-1]] = dict()
unexpanded_class_paths = []
for module in self.imported_modules.keys():
for class_ in self.get_classes(module).values():
unexpanded_class_paths.append([
self.get_class_full_name(i)
for i in reversed(list(inspect.getmro(class_[0])))
])
tree = {}
added = [] # the expander makes duplicates. keep a list to remove them
# sadly a collections.Counter doesnt work with lists of lists
for s in expand(unexpanded_class_paths):
if s not in added:
setTree(tree, [i for i in reversed(s)][::-1])
added.append(s)
# return inspect.getclasstree(classes)
return tree
def run_tests(self, tests, run_colourful = False):
"""Build and run pytests from the configuration yaml. Indentation needs to
be four spaces only otherwise it won't work. We recommend running this last
so all modules are already imported.
Args:
tests (dict): dict with the filename as the key followed by a list of
python code to make the test
run_colourful (bool, optional): Run pytest again, printing out the
exact output of pytest as soon as it's ready. Has the advantage that
colours are preserved, but is only useful for when the user wants to
print out the report to stdout. Defaults to False.
Returns:
[dict]: A dictionary consisting of the pytest output string, junit xml
output (which might be useful for rendering nicely in some output formats)
and some nice meta information.
"""
test_results = {}
test_results["pytest_report"] = ""
with open("pytest_template.jinja2", "r") as f:
jinja_template = jinja2.Template(f.read())
for filename, filestests in tests.items():
with open(os.path.join(self.client_code_path, "test_" + filename), "w") as f:
f.write(jinja_template.render(
module = os.path.splitext(filename)[0],
filestests = filestests,
enumerate = enumerate # a function thats needed
))
with tempfile.TemporaryDirectory() as tmp:
junitxmlpath = os.path.join(tmp, "report.xml")
test_files = [os.path.join(self.client_code_path, "test_%s" % f) for f in tests.keys()]
cmd = ["pytest", "-v"] + test_files + ["--junitxml=%s" % junitxmlpath]
# print("cmd: ", " ".join(cmd))
if test_files == []:
test_results["pytest_report"] = "*** No Tests ***"
return test_results
proc = subprocess.Popen(cmd, stdout = subprocess.PIPE)
while True:
line = proc.stdout.readline()
if not line:
break
test_results["pytest_report"] += line.decode()
with open(junitxmlpath, "r") as f:
test_results["junitxml"] = f.read()
root = etree.fromstring(test_results["junitxml"])
test_results["meta"] = root.findall("./testsuite")[0].attrib
if run_colourful:
subprocess.run(cmd)
return test_results
def __format_doc(*doc):
return str(doc[1]).rstrip()
def gen_reflection_report(client_code_path, assessment_struct, student_no, configuration):
# print(configuration)
reflection = Reflect(client_code_path)
present_module_names = [i.name for i in reflection.client_modules]
out = assessment_struct
out["student_no"] = student_no
tests_to_run = {}
for i, required_file in enumerate(assessment_struct["files"], 0):
required_file = list(required_file.keys())[0]
module_name = os.path.splitext(required_file)[0]
if module_name in present_module_names:
out["files"][i][required_file]["present"] = True
else:
out["files"][i][required_file]["present"] = False
continue
try:
reflection.import_module(module_name)
except Exception as e:
out["files"][i][required_file]["has_exception"] = True
out["files"][i][required_file]["exception"] = {}
out["files"][i][required_file]["exception"]["type"] = str(type(e))
out["files"][i][required_file]["exception"]["str"] = str(e)
# TODO: work out how to only get the exception stack of the client code
out["files"][i][required_file]["exception"]["traceback"] = ''.join(traceback.format_exception(None, e, e.__traceback__))
continue
required_files_features = assessment_struct["files"][i][required_file]
out["files"][i][required_file]["has_exception"] = False
out["files"][i][required_file]["documentation"] = reflection.get_module_doc(module_name)
if "classes" in required_files_features.keys():
present_classes = reflection.get_classes(module_name)
for j, class_name in enumerate(required_files_features["classes"], 0):
class_name = list(class_name.keys())[0]
stop_here_flag = False
# surprised the yaml parser doesnt do this automatically...
if out["files"][i][required_file]["classes"][j][class_name] is None:
out["files"][i][required_file]["classes"][j][class_name] = {}
stop_here_flag = True
if class_name in present_classes.keys():
out["files"][i][required_file]["classes"][j][class_name]["present"] = True
else:
out["files"][i][required_file]["classes"][j][class_name]["present"] = False
continue
# print( present_classes[class_name][1])
out["files"][i][required_file]["classes"][j][class_name]["documentation"] = present_classes[class_name][1]
if stop_here_flag:
continue
present_methods = reflection.get_class_methods(module_name, class_name)
for k, required_method in enumerate(assessment_struct["files"][i][required_file]["classes"][j][class_name]["methods"], 0):
out["files"][i][required_file]["classes"][j][class_name]["methods"][k] = {required_method: {}}
method_name = re.sub(r"\(\d+\)", "", required_method)
if method_name in present_methods.keys():
out["files"][i][required_file]["classes"][j][class_name]["methods"][k][required_method]["present"] = True
else:
out["files"][i][required_file]["classes"][j][class_name]["methods"][k][required_method]["present"] = False
continue
out["files"][i][required_file]["classes"][j][class_name]["methods"][k][required_method]["arguments"] = present_methods[method_name][-2]
out["files"][i][required_file]["classes"][j][class_name]["methods"][k][required_method]["minimum_arguments"] = present_methods[method_name][-2].count(",") + 1
out["files"][i][required_file]["classes"][j][class_name]["methods"][k][required_method]["documentation"] = present_methods[method_name][-3]
out["files"][i][required_file]["classes"][j][class_name]["methods"][k][required_method]["source_code"] = present_methods[method_name][-1]
if "functions" in required_files_features.keys():
present_functions = reflection.get_functions(module_name)
for j, required_function in enumerate(assessment_struct["files"][i][required_file]["functions"], 0):
function_name = re.sub(r"\(\d+\)", "", required_function)
out["files"][i][required_file]["functions"][j] = {required_function: {}}
if function_name in present_functions.keys():
out["files"][i][required_file]["functions"][j][required_function]["present"] = True
else:
out["files"][i][required_file]["functions"][j][required_function]["present"] = False
continue
out["files"][i][required_file]["functions"][j][required_function]["documentation"] = present_functions[function_name][-3]
out["files"][i][required_file]["functions"][j][required_function]["arguments"] = present_functions[function_name][-2]
out["files"][i][required_file]["functions"][j][required_function]["minimum_arguments"] = present_functions[function_name][-2].count(",") + 1
out["files"][i][required_file]["functions"][j][required_function]["source_code"] = present_functions[function_name][-1]
if "tests" in required_files_features.keys():
filename = list(assessment_struct["files"][i].keys())[0]
if not out["files"][i][filename]["has_exception"]:
for j, test in enumerate(assessment_struct["files"][i][required_file]["tests"], 0):
try:
tests_to_run[filename].append(test)
except KeyError:
tests_to_run[filename] = [test]
out["test_results"] = reflection.run_tests(tests_to_run, configuration["out"] == "stdout" and configuration["format"] in ["text", "txt"])
out["class_tree"] = reflection.get_class_tree()
return out
if __name__ == "__main__":
# user_code_path = "D:\\Edencloud\\UniStuff\\3.0 - CMP 3rd Year Project\\Smarker\\../ExampleSubmissions/Submission_A"
# reflect = Reflect(user_code_path)
# reflect.import_module("pjtool")
# # for c, v in reflect.get_classes(("pjtool")).items():
# # print(c, v)
# for k, v in reflect.get_functions("pjtool").items():
# print(k, v)
reflect = Reflect(os.getcwd())
print(reflect.client_modules)
reflect.import_module("jinja_helpers")
print({k: v for k, v in reflect.get_functions("jinja_helpers").items()})
|