From 8e9f97e60717dbb19fc9e2e810a080e5897bf495 Mon Sep 17 00:00:00 2001
From: laptop <laptop@DESKTOP-68EPQJ4>
Date: Thu, 21 May 2020 14:30:41 +0500
Subject: [PATCH] 1st

---
 get_subs_from_one_channel.py | 34 ++++++++++++++++
 output.txt                   | 79 ++++++++++++++++++++++++++++++++++++
 requirements.txt             |  2 +
 3 files changed, 115 insertions(+)
 create mode 100644 get_subs_from_one_channel.py
 create mode 100644 output.txt
 create mode 100644 requirements.txt

diff --git a/get_subs_from_one_channel.py b/get_subs_from_one_channel.py
new file mode 100644
index 0000000..dc03a34
--- /dev/null
+++ b/get_subs_from_one_channel.py
@@ -0,0 +1,34 @@
+from youtube_transcript_api import YouTubeTranscriptApi
+import urllib3
+from collections import Counter
+def get_videoids(channel_url):
+    #https://stackoverflow.com/questions/15512239/python-get-all-youtube-video-urls-of-a-channel
+    r=http.request("GET",channel_url)
+    url_data=str(r.data.decode("utf-8")).split()
+    item= 'href="/watch?'
+    vids = [line.replace('href="', 'youtube.com') for line in url_data if item in line]
+    vids=list(set([i[i.find("=")+1:i.find('"')] for i in vids]))
+    return vids
+
+def get_subs(video_id):
+    text=""
+    subs=YouTubeTranscriptApi.get_transcript(video_id)
+    for i in subs:
+        text+=i["text"]
+    return text
+user_agent={'user-agent':'Mozilla/5.0 (compatible; Googlebot/2.1; +http://www.google.com/bot.html)'}
+http=urllib3.PoolManager(num_pools=1,headers=user_agent)
+channel_url=input("Введите ссылку на канал: ")
+#получаем все VideoID
+vids=get_videoids(channel_url)
+print(vids)
+print("Получили {} ссылок на видео".format(len(vids)))
+text=""
+for i in vids:
+    text=text+get_subs(i)
+text=text.encode("ascii", "ignore").decode("utf-8")
+text=text.split()
+with open("output.txt","w") as f:
+    for i,j in Counter(text).most_common():
+        print(i,j,file=f)
+print("Сохранили в output.txt")
\ No newline at end of file
diff --git a/output.txt b/output.txt
new file mode 100644
index 0000000..fd40cc6
--- /dev/null
+++ b/output.txt
@@ -0,0 +1,79 @@
+(' ', 54861)
+('e', 30580)
+('t', 27850)
+('o', 22359)
+('a', 20689)
+('i', 18014)
+('n', 17932)
+('s', 15942)
+('h', 13499)
+('r', 13296)
+('l', 11357)
+('d', 9334)
+('u', 8771)
+('y', 7526)
+('m', 7368)
+('c', 6496)
+('g', 6270)
+('w', 5875)
+('f', 5328)
+('p', 4838)
+('b', 3912)
+('k', 3503)
+('I', 2964)
+('v', 2716)
+("'", 2667)
+(',', 1563)
+('.', 1401)
+('\n', 1302)
+('j', 587)
+('x', 489)
+('-', 432)
+('A', 307)
+('T', 283)
+('0', 264)
+('S', 256)
+('z', 217)
+('q', 193)
+('M', 189)
+('N', 165)
+('1', 140)
+('W', 125)
+('B', 108)
+('Y', 105)
+('(', 104)
+(')', 104)
+('3', 97)
+('?', 88)
+('G', 85)
+('[', 83)
+(']', 83)
+('O', 81)
+('C', 74)
+('2', 72)
+('D', 68)
+('F', 60)
+('L', 59)
+('5', 57)
+('J', 57)
+('"', 53)
+('H', 48)
+('P', 48)
+('R', 41)
+('!', 35)
+('E', 34)
+('K', 32)
+(':', 29)
+('9', 28)
+('V', 27)
+('%', 23)
+('4', 23)
+('6', 18)
+('8', 17)
+('7', 15)
+('U', 12)
+('/', 8)
+('$', 8)
+(';', 5)
+('Z', 4)
+('X', 4)
diff --git a/requirements.txt b/requirements.txt
new file mode 100644
index 0000000..9b9ac0a
--- /dev/null
+++ b/requirements.txt
@@ -0,0 +1,2 @@
+youtube_transcript_api
+urllib3
\ No newline at end of file