ban promotions, improve lolcow detection

autodrama.py

@@ -1,8 +1,10 @@
 TEST_MODE = True
 
+from audioop import avg
 from re import sub
 from time import sleep
 from typing import Tuple
+from numpy import average
 import praw
 from praw.models import Comment, Submission
 from vaderSentiment.vaderSentiment import SentimentIntensityAnalyzer
@@ -11,6 +13,8 @@ from os.path import exists, join, realpath, split
 
 from RDramaAPIInterface import RDramaAPIInterface
 
+BANNED_WORDS_IN_POST = ['comment', 'promotion']
+
 def get_real_filename(filename : str):
     path_to_script = realpath(__file__)
     path_to_script_directory, _ = split(path_to_script)
@@ -31,6 +35,10 @@ reddit = praw.Reddit(
 
 pushshift_api = PushshiftAPI(reddit)
 
+def contains_banned_words(sentence):
+    santitized_sentence = ''.join([i for i in sentence.lower() if i in 'abcdefghijklmnopqrstuvwxyz '])
+    return bool(set(BANNED_WORDS_IN_POST).intersection(santitized_sentence.split(" ")))
+
 def get_based_submissions(subreddit, time_frame, limit):
     subscriber_cache = {}
 
@@ -42,7 +50,7 @@ def get_based_submissions(subreddit, time_frame, limit):
     for submission in reddit.subreddit(subreddit).controversial(time_frame, limit=limit):
         try:
             basedness = (1-submission.upvote_ratio)*submission.num_comments
-            if (submission.author.name == "AutoModerator" or "comment" in submission.title.lower()):
+            if (submission.author.name == "AutoModerator" or contains_banned_words(submission.title)):
                 continue
             if (submission.subreddit not in subscriber_cache):
                 subscriber_cache[submission.subreddit] = submission.subreddit.subscribers
@@ -63,8 +71,9 @@ def get_based_submissions(subreddit, time_frame, limit):
 def analyze_comments(submission : 'Submission'):
     print(f"[{submission.id}]Retrieving Comments")
     comments = pushshift_api.search_comments(subreddit=submission.subreddit.display_name, link_id=submission.id)
+    comment_list = list(comments)
     print(f"[{submission.id}]Creating Network")
-    comment_map = {i.id:i for i in list(comments)}
+    comment_map = {i.id:i for i in comment_list}
     child_map = {}
     for comment in comment_map.values():
         try:
@@ -78,6 +87,7 @@ def analyze_comments(submission : 'Submission'):
     sid_obj = SentimentIntensityAnalyzer()
     print(f"[{submission.id}]Classifying Comments")
     user_to_total_anger = {}
+    redditors = {}
     ranked_comments = []
     angry_comments = []
     for comment in comment_map.values():
@@ -126,7 +136,14 @@ def analyze_comments(submission : 'Submission'):
             
             if comment.author not in user_to_total_anger:
                 user_to_total_anger[comment.author] = 0.0
+                redditors[comment.author] = {}
+                redditors[comment.author]['comments'] = []
+                redditors[comment.author]['angriness'] = 0
+                redditors[comment.author]['author'] = comment.author
+                
             user_to_total_anger[comment.author]+=score
+            redditors[comment.author]['comments'].append((score, comment_info))
+            redditors[comment.author]['angriness'] += score
         except Exception as e:
             print(f"Error while processing {comment}: {e}")
 
@@ -135,10 +152,13 @@ def analyze_comments(submission : 'Submission'):
     angry_comments.sort(key=lambda a:a[0])
     lolcows = [(v, k) for k, v in user_to_total_anger.items()]
     lolcows.sort(key=lambda a:a[0])
+    redditors_ranked = [(data['angriness'], data) for username, data in redditors.items()]
+    redditors_ranked.sort(key=lambda a:a[0])
     return {
         'based' : ranked_comments,
         'angry': angry_comments,
-        'lolcows': lolcows
+        'lolcows': lolcows,
+        'redditors': redditors_ranked
     }
 #get_based_submissions("all", "hour", 25, True)
 
@@ -202,18 +222,40 @@ def angriness_score_string(angriness):
     
     return get_score_string(score, "😡", "🔘")
 
+def get_reddit_full_url(partial_url):
+    return f"https://reddit.com{partial_url}"
+
+def generate_lolcow_display_section(lolcows):
+    markdown_lines = []
+
+    
+    biggest_lolcow_info = lolcows[0]
+    biggest_lolcow_score = biggest_lolcow_info[0]
+    biggest_lolcow = biggest_lolcow_info[1]['author']
+    number_of_comments = len(biggest_lolcow_info[1]['comments'])
+    lolcow_comments = biggest_lolcow_info[1]['comments']
+    lolcow_score_string = get_score_string(-1*biggest_lolcow_score, "🐮", "🔘", allow_over=True)
+    markdown_lines.append(f"# Biggest Lolcow: /u/{biggest_lolcow.name}")
+    markdown_lines.append(f"Score: {lolcow_score_string}")
+    markdown_lines.append(f"Number of comments: {number_of_comments}")
+    comment_angryness_scores = [-1*a[0] for a in biggest_lolcow_info[1]['comments']]
+    average_angriness = average(comment_angryness_scores)
+    maximum_angry_info = max(lolcow_comments, key=lambda a : a[0])
+    minimum_angry_info = min(lolcow_comments, key=lambda a : a[0])
+    markdown_lines.append(f"Average angriness: {angriness_score_string(average_angriness)}")
+    markdown_lines.append(f"Maximum angriness: [{angriness_score_string(-1*maximum_angry_info[0])}]({get_reddit_full_url(maximum_angry_info[1]['comment'].permalink)})")
+    markdown_lines.append(f"Minimum angriness: [{angriness_score_string(-1*minimum_angry_info[0])}]({get_reddit_full_url(minimum_angry_info[1]['comment'].permalink)})")
+
+    return markdown_lines
+
 def generate_submission_report(submission : 'Submission', absolute: bool):
     markdown_lines = []
     comment_analysis_results = analyze_comments(submission)
     basedness_display_func = lambda a : get_comment_basedness_out_of_five(a, absolute)
     markdown_lines.extend(generate_comment_display_section(comment_analysis_results['based'], "Most Based Comments", "Basedness", 3, detail_func=basedness_display_func))
     markdown_lines.extend(generate_comment_display_section(comment_analysis_results['angry'], "Angriest Comments", "Angriness", 3, detail_func=angriness_score_string))
-    biggest_lolcow_info = comment_analysis_results['lolcows'][0]
+    markdown_lines.extend(generate_lolcow_display_section(comment_analysis_results['redditors']))
-    biggest_lolcow_score = biggest_lolcow_info[0]
+    markdown_lines.append(f"")
-    biggest_lolcow = biggest_lolcow_info[1]
-    markdown_lines.append(f"# Biggest lolcow")
-    lolcow_score_string = get_score_string(-1*biggest_lolcow_score, "🐮", "🔘")
-    markdown_lines.append(f"/u/{biggest_lolcow.name} {lolcow_score_string}")
     markdown_lines.append("*:marppy: autodrama: automating away the jobs of dramautists. :marseycapitalistmanlet: Ping HeyMoon if there are any problems or you have a suggestion :marseyjamming:*")
     return "\n\n".join(markdown_lines)
 
@@ -277,8 +319,12 @@ def get_comment_basedness_out_of_five(basedness: int, absolute : bool):
             score = 0
     return get_score_string(score, "🔥", "🔘")
 
-def get_score_string(score: int, filled_emoji, empty_emoji) -> str:
+def get_score_string(score: int, filled_emoji, empty_emoji, allow_over = False) -> str:
-    return "".join([filled_emoji if ((i+1) <= score) else empty_emoji for i in range(5)])
+    to_return = "".join([filled_emoji if ((i+1) <= score) else empty_emoji for i in range(5)])
+    if (allow_over):
+        if (score > 5):
+            to_return += f"(+{int(score)-5}{filled_emoji})"
+    return to_return
 
 def create_rdrama_report(rdrama : RDramaAPIInterface, submission : 'Submission', basedness: int, absolute_basedness: bool):
     score = get_basedness_score_out_of_five(basedness)
@@ -326,4 +372,5 @@ else:
     timeout = 10
 rdrama = RDramaAPIInterface(auth, website, timeout, https=https)
 
-daily_drama_post(rdrama)
+#daily_drama_post(rdrama)
+print(generate_submission_report(Submission(reddit, "uesnfd"), True))