Merge pull request #19 from Bonfire/tom-backend-changes

Tom backend changes

Backend/app_interface.py

@@ -1,6 +1,6 @@
 '''
-Authors: Thomas Serrano, Curtis Helsel
-Last Updated: OCT-29-2019
+Authors: Thomas Serrano, Curtis Helsel, Baran Barut
+Last Updated: NOV-20-2019
 '''
 from config import *   #Application stuff
 from predictor import *                    #ML stuff
@@ -10,60 +10,114 @@
 from github import Github
 from os import walk, path         #For parsing through directories
 
+'''
+Gets a list of filenames for a given file palth.
+base_dir is used to exclude the common base director
+'''
+def get_filenames(base_dir, file_path, file_names):
+
+	for cur in listdir(file_path):
+
+		#Full path is needed to test whether cur is a directory
+		#full_path = path.join(file_path, cur)
+		full_path = file_path + '/' + cur
+
+		#Gets the name of the current directory, then joins with the current filename before appending
+		directory = path.split(file_path)[-1]
+		final_name = full_path.replace(base_dir, '')[1:]
+
+		#If file...
+		if not path.isdir(full_path):
+			#Filter out stuff like .gitignore etc
+			if final_name[0] != '.':
+				file_names.append(final_name)
+		#If directory, recursively got through it
+		else:
+			get_filenames(base_dir, full_path, file_names)
+
 '''
 Provides an interface to connect the application with the machine learning model
 '''
 def app_interface():
-    #Get command-line arguments
-    username = sys.argv[1]
-    password = sys.argv[2]
-    repo_name = sys.argv[3]
-    repo_dir = sys.argv[4]
-
-    #print("Starting interface...")
-    #print("Arguments:\n" + username + " / " + password + " / " + repo_name + " / " + repo_dir)
-
-    #Get Github access
-    git = Github(username, password)
 
-    #Grabs repo object for data collection
-    github_repo = git.get_repo(repo_name, lazy=False)
-    repo = Repo(repo_dir)
-
-    #Retrieve repository data
-    repo_data = list(repository_data(github_repo, repo, repo_name, repo_dir))
-
-    #Retrieve all filenames for all directories
-    files_path = []
-    for root, dirs, files in walk(repo_dir):
-        f_path = root.split(path.basename(repo_name) + "\\")[-1]
-        if f_path is not "":
-            f_path += "/"
-        files_path += [str(f_path + f) for f in files if not f[0] == '.']
-        dirs[:] = [d for d in dirs if not d[0] == '.']
+	#Get number of command-line arguments
+	argc = len(sys.argv)
+
+	'''
+	print("Starting interface...")
+	print(argc)
+	for a in sys.argv:
+		print(a)
+	'''
+
+	#Must have 1, 2, or 4 arguments (program name is the first argument)
+	if argc not in [2, 3, 5]:
+		print("\n\tERROR: INCORRECT NUMBER OF ARGUMENTS PASSED")
+		exit()
+
+	#Used for flow control for different types of repos
+	repo_remote = False
+	repo_private = False
+
+	#The first argument is the repo directory. This is required for all repos, and is the only
+	#argument for local repos.
+	repo_dir = sys.argv[1]
+	repo = Repo(repo_dir)
+
+	#Repo name is required for remote & public repos.
+	if argc > 2:
+		repo_remote = True
+		repo_name = sys.argv[2]
+
+		#Login information is required for remote & private repos.
+		if argc > 3:
+			repo_private = True
+			username = sys.argv[3]
+			password = sys.argv[4]
+
+		#Get Github access. Use login if applicable.
+		if repo_private:
+			git = Github(username, password)
+		else:
+			git = Github(None, None)
+
+		#Grabs repo object for data collection
+		github_repo = git.get_repo(repo_name, lazy=False)
+
+		#Retrieve repository data
+		repo_data = list(repository_data(github_repo, repo, repo_name, repo_dir))
+
+	#Retrieve all filenames for all directories
+	files_path = []
+	get_filenames(repo_dir, repo_dir, files_path)
+
+	#Get a list of data points for each file
+	avgs = get_averages(files_path, repo.head.commit.hexsha, repo)
+
+	#Separate file names from avgs
+	file_names = []
+	avgs_final = []
+	for a in avgs:
+		file_names.append(a[0])
+		avgs_final.append(a[1:])
+
+	#Collect data together into a single list
+	final_data = []
 
-    #Get a list of data points for each file
-    avgs = get_averages(files_path, repo.head.commit.hexsha, repo)
+	#If this is a remote repo, we need to add the repo data, otherwise use averages
+	if repo_remote:
+		for af in avgs_final:
+			final_data.append(af + repo_data[1:])
+	else:
+		final_data = avgs_final
 
-    #Separate file names from avgs
-    file_names = []
-    avgs_final = []
-    for a in avgs:
-        file_names.append(a[0])
-        avgs_final.append(a[1:])
-
-    #Collect data together into a single list
-    final_data = []
-    for af in avgs_final:
-        final_data.append(af + repo_data[1:])
-
-    #Predict and print results data to std out
-    results = predict(final_data)
-
-    #Final results output
-    for i, f in enumerate(file_names):
-        print(f + "," + str(results[i][0]))
+	#Predict and print results data to std out
+	results = predict(final_data)
+
+	#Final results output
+	for i, f in enumerate(file_names):
+		print(f + "," + str(results[i][0]))
 
 #Main - delete later?
 if __name__ == "__main__":
-    app_interface()
+	app_interface()