From 3c25310d760fb085e79703feab1c83a2abc3568f Mon Sep 17 00:00:00 2001
From: andrewso <9V5f1FkzI2LD>
Date: Wed, 14 Oct 2020 18:51:36 +0100
Subject: [PATCH] [14.10.20] Testing

---
 src/tweets/collector.py       |   4 +
 src/utils/whitelistedWords.py | 151 ++++++++++++++++++++++++++++++++++
 2 files changed, 155 insertions(+)
 create mode 100644 src/utils/whitelistedWords.py

diff --git a/src/tweets/collector.py b/src/tweets/collector.py
index c2ac37c..528d5ec 100644
--- a/src/tweets/collector.py
+++ b/src/tweets/collector.py
@@ -17,6 +17,7 @@ from src.utils.spamFilter import callSpamFilter
 from src.utils.sentimentAnalyser import callSentimentAnalyser
 from src.utils.activemqConnect import activeMQSender
 from src.utils.jsonLogger import log
+from src.utils.whitelistedWords import filterOutTweetsWithNoneWhitelistedWords
 
 from http.client import IncompleteRead
 from urllib3.exceptions import ProtocolError
@@ -94,15 +95,18 @@ class Listener(StreamListener):
                 if 'extended_tweet' in data['retweeted_status']:
                     #if tweet is over the 140 word limit
                     text = data['retweeted_status']['extended_tweet']['full_text']
+                    text = filterOutTweetsWithNoneWhitelistedWords(text)
                     dumpStack.append({'type': self.hashtag, 'tweet': text})
                 else:
                     text = data['retweeted_status']['text']
+                    text = filterOutTweetsWithNoneWhitelistedWords(text)
                     dumpStack.append({'type': self.hashtag, 'tweet': text})
             else:
                 # Else if a normal Tweeet
                 if 'extended_tweet' in data:
                     # If tweet is over 140 word limit
                     text = data['extended_tweet']['full_text']
+                    text = filterOutTweetsWithNoneWhitelistedWords(text)
                     dumpStack.append({'type': self.hashtag, 'tweet': text})
 
 def processTweet():
diff --git a/src/utils/whitelistedWords.py b/src/utils/whitelistedWords.py
new file mode 100644
index 0000000..8454e9f
--- /dev/null
+++ b/src/utils/whitelistedWords.py
@@ -0,0 +1,151 @@
+#!/usr/bin/env python
+
+from src.utils.jsonLogger import log
+
+whitelist = [
+    "bull",
+    "bear",
+    "bullish",
+    "bearish",
+    "up",
+    "down",
+    "high",
+    "low",
+    "higher",
+    "lower",
+    "absconded",
+    "maximalists",
+    "regulate",
+    "infamous",
+    "tradehigher",
+    "tradelower",
+    "revival",
+    "centralized",
+    "decentralized",
+    "centralised",
+    "decentralised",
+    "decentralization",
+    "decentralisation",
+    "centralization",
+    "centralisation",
+    "bans",
+    "hodl",
+    "ambiguity",
+    "revolutionize",
+    "revolutionise",
+    "consolidation",
+    "shorts",
+    "longs",
+    "long",
+    "short",
+    "shorting",
+    "grow",
+    "volatile",
+    "rally",
+    "rallying",
+    "noob",
+    "noobs",
+    "innovation",
+    "bottom",
+    "top",
+    "topped",
+    "bottomed",
+    "upwards",
+    "downwards",
+    "invest",
+    "raging",
+    "rocketing",
+    "swing",
+    "swinging",
+    "stake",
+    "whale",
+    "whales",
+    "lull",
+    "moon",
+    "choppy",
+    "buy",
+    "buying",
+    "sell",
+    "selling",
+    "startselling",
+    "stopselling",
+    "startbuying",
+    "stopbuying",
+    "bitcoin",
+    "btc",
+    "eth",
+    "xmr",
+    "xrp",
+    "ripple",
+    "block",
+    "reward",
+    "airdrop",
+    "drop",
+    "raise",
+    "stack",
+    "stake",
+    "invest",
+    "pull",
+    "push",
+    "token",
+    "sale",
+    "unhappy",
+    "happy",
+    "expert",
+    "novice"
+    "passed",
+    "mark",
+    "decline",
+    "incline",
+    "fees",
+    "crypto",
+    "wallet",
+    "price",
+    "history",
+    "reached",
+    "upward",
+    "downward",
+    "trading",
+    "mining",
+    "defi",
+    "finance",
+    "blockchain",
+    "interest",
+    "alt",
+    "alts",
+    "fiat",
+    "fiat",
+    "currency",
+    "currencies",
+    "wealth",
+    "hype",
+    "hyped",
+    "achievement",
+    "platform",
+    "incremental",
+    "increment",
+    "decrement",
+    "decremental",
+    "success",
+    "loss",
+    "win",
+    "lose",
+    "worth",
+    "strongest",
+    "weakest",
+    "strong",
+    "weak",
+    "trade",
+    "popping",
+    "sucking",
+    "shard",
+    "sharding",
+    "industry"
+]
+
+def filterOutTweetsWithNoneWhitelistedWords(text):
+    if any(x in text for x in whitelist):
+        return text
+    else:
+        log("Tweet [{}] did not contain any keywords for it to be considered crypto related".format(text), 'WARN')
+        return ""
\ No newline at end of file