Co-authored-by: Nick Bland <NickBland@users.noreply.github.com>

.vscode/launch.json

@@ -0,0 +1,16 @@
+{
+    // Use IntelliSense to learn about possible attributes.
+    // Hover to view descriptions of existing attributes.
+    // For more information, visit: https://go.microsoft.com/fwlink/?linkid=830387
+    "version": "0.2.0",
+    "configurations": [
+        {
+            "name": "Python: Current File",
+            "type": "python",
+            "request": "launch",
+            "program": "${file}",
+            "console": "integratedTerminal",
+            "justMyCode": true
+        }
+    ]
+}

asked_questions_log.txt

@@ -6,3 +6,65 @@ test
 /usr/local/bin/python3 "/Volumes/GoogleDrive-100976413726208790908/My Drive/UNI/INFS 2048 - System Design and Realisation/Assessment 2B/main.py"
 weather
 today
+birthday
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+#!@#!@#@!%#$@#$@%#$%$$#!$#@!#@weather
+weather
+Hello
+$$%
+$@$@$@$@$@$
+#$#$#$#$
+#$#$#$#
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather
+weather

main.py

@@ -1,43 +1,93 @@
 import json
+import string
 
-def jaccard_similarity(userQuestion, questionBank):
+def sanitiseInput(dirtyQuestion):
-    s1 = set(userQuestion)
+  """Sanitise a user inputted question to make it lower case, remove punctuation
-    s2 = set(questionBank)
-    return float(len(s1.intersection(s2)) / len(s1.union(s2))) 
 
-def most_likely(userQuestion):
+  Arguments:
+  dirtyQuestion -- a question that contains punctuation and different-cased letter
+
+  Returns:
+  String with the user's input without punctuation and all in lower cawse
+  """
+  # Translate the dirty input into a sanitised one by removing all punctuation characters
+  # maketrans creates a table of characters to change, change with, and remove. In this case, no changes except remove punctuation.
+  # Derived from: https://datagy.io/python-remove-punctuation-from-string/
+  newQuestion = dirtyQuestion.translate(str.maketrans('', '', string.punctuation))
+  newQuestion = newQuestion.lower().strip()
+  return newQuestion
+
+def jaccardSimilarity(userQuestion, questionBank):
+  """Check the similarity of a user defined question to the bank of questions available
+
+  Arguments:
+  userQuestion -- question that the user has entered
+  questionBank -- python json object of questions and answers
+
+  Returns:
+  Float between 0 and 1 of similarity (1 being exactly the same)
+  """
+  s1 = set(sanitiseInput(userQuestion))
+  s2 = set(sanitiseInput(questionBank))
+  return float(len(s1.intersection(s2)) / len(s1.union(s2))) 
+
+def mostLikely(userQuestion):
+  """Checks how likely a user entered question is to the list of known questions
+
+  Arguments:
+  userQuestion -- question that the user has entered
+
+  Returns:
+  String of the most likely question from the known user bank
+  """
   likelihoodScore = []
   fileObject = open("faq.json", "r")
   jsonContent = fileObject.read()
-  aList = json.loads(jsonContent)
+  questionBank = json.loads(jsonContent)
-  for json_object in aList:
+  # Using list of known questions, iterate through and determine similarity of all of them
-  	likelihoodScore.append(jaccard_similarity(userQuestion, json_object['question']))
+  for faqPair in questionBank:
-  #print(likelihoodScore)
+    likelihoodScore.append(jaccardSimilarity(userQuestion, faqPair['question']))
 
+  # Get the question that has the highest likelihood
   mostLikelyIndex = likelihoodScore.index(max(likelihoodScore))
-  mostLikely = aList[mostLikelyIndex]
+  mostLikely = questionBank[mostLikelyIndex]
     
-  fileObject.close()
+  fileObject.close() # Close file to prevent IO errors
 
   return mostLikely
 
+def continueQuestions(userQuestion):
+  """Check whether question has any content in it.
+
+  Arguments:
+  userQuestion -- question that the user has entered
+
+  Returns:
+  False if input is empty or True if required to re-loop
+  """
+  if sanitiseInput(userQuestion):
+    likelyQuestion = mostLikely(userQuestion)
+    print ("I think that you asked " + "'" + (likelyQuestion['question']) + "'" + " and conclude that the answer is " + "'" + (likelyQuestion['answer']) + "'.\n")
+    # Open the file in append & read mode ('a+')
+    with open("asked_questions_log.txt", "a+") as file_object:
+      file_object.write(userQuestion + "\n")
+    return True
+  else:
+    # return False if there is no input to exit the program
+    return False
+
 def main():
   print("Hello, I am a question answering bot.\n")
+
   userQuestion = ''
-  while True:
+  userQuestion = input('Please enter a question, and press the ENTER key: \n')
-    userQuestion = input('Please enter a question, and press the ENTER key: \n').strip()
-    if userQuestion:
-      likelyQuestion = most_likely(userQuestion)
-      print ("I think that you asked " + "'" + (likelyQuestion['question']) + "'" + " and conclude that the answer is " + "'" + (likelyQuestion['answer']) + "'.\n")
-      # Open the file in append & read mode ('a+')
-      with open("asked_questions_log.txt", "a+") as file_object:
-          file_object.write("\n")
-          file_object.write(userQuestion)
 
+  # Loop again and again when user enters a question and provide user with results
+  while continueQuestions(userQuestion):
+    userQuestion = input('Please enter a question, and press the ENTER key: \n')
 
-    else :
+  # Exit program gracefully with a message
-      print ("Goodbye!")
+  print("Goodbye!")
-      exit()
 
 if __name__ == "__main__":
   main()

test_questionbot.py

@@ -1,25 +1,55 @@
 import main
 import pytest
-
+from os.path import exists
-#def test_CanInstantiatejaccard_similarity():
-#    co = main.jaccard_similarity(userQuestion="Weather", questionBank="Same as yesterday.")
-#    assert main.jaccard_similarity(float)
 
 def test_likelyQuestion():
-  likelyQuestion = main.most_likely("birthday")
+  likelyQuestion = main.mostLikely("birthday")
   assert likelyQuestion['question'] == 'What is your birthday?'
 
 def test_likelyAnswer():
-  likelyAnswer = main.most_likely("weather today")
+  likelyAnswer = main.mostLikely("weather today")
   assert likelyAnswer['answer'] == 'Same as yesterday.'
 
 def test_detection_of_no_user_input():
-        # Override the Python built-in input method
+  userQuestion = ''
-        main.main.input = lambda: ''
+  assert not main.continueQuestions(userQuestion)
-        # Call the function you would like to test (which uses input)
+
-        output = main.main()
+def testDetectionOfUserInput():
-        assert output == [
+  userQuestion = 'weather'
-  'Hello, I am a question answering bot.',
+  assert main.continueQuestions(userQuestion)
-  'Please enter a question, and press the ENTER key: ',
+
-  'Goodbye!'
+def testFaqJsonFileExists():
-]
+  fileName = exists('./faq.json')
+  assert fileName
+
+#  Check whether log file gets made after a successful run
+
+def testLogfileExists():
+  logName = exists('./asked_questions_log.txt')
+  assert logName
+
+def readFileLines(fileName):
+  file = open(fileName, "r")
+  # Create a count of non-empty lines
+  nonemptyLines = [line.strip("\n") for line in file if line != "\n"]
+  lineCount = len(nonemptyLines)
+  file.close()
+  return lineCount
+
+def testLogfileLinecountIncreases():
+  oldLineCount = readFileLines("./asked_questions_log.txt")
+
+  userQuestion = 'weather'
+  main.continueQuestions(userQuestion)
+
+  newLineCount = readFileLines("./asked_questions_log.txt")
+
+  assert newLineCount == oldLineCount + 1
+
+def testDetectionIfUserEntersSymbolsOnly():
+  userQuestion = '#$%^&*'
+  assert not main.continueQuestions(userQuestion)
+
+def testDetectionIfUserEntersSpaceOnly():
+  userQuestion = '    '
+  assert not main.continueQuestions(userQuestion)