diff --git a/BeautifulSoup/scrape.py b/BeautifulSoup/scrape.py new file mode 100644 index 0000000..42a191f --- /dev/null +++ b/BeautifulSoup/scrape.py @@ -0,0 +1,37 @@ +from bs4 import BeautifulSoup +import requests +import csv + +source = requests.get('http://coreyms.com').text + +soup = BeautifulSoup(source, 'lxml') + +csv_file = open('cms_scrape.csv', 'w') + +csv_writer = csv.writer(csv_file) +csv_writer.writerow(['headline', 'summary', 'video_link']) + +for article in soup.find_all('article'): + headline = article.h2.a.text + print(headline) + + summary = article.find('div', class_='entry-content').p.text + print(summary) + + try: + vid_src = article.find('iframe', class_='youtube-player')['src'] + + vid_id = vid_src.split('/')[4] + vid_id = vid_id.split('?')[0] + + yt_link = f'https://youtube.com/watch?v={vid_id}' + except Exception as e: + yt_link = None + + print(yt_link) + + print() + + csv_writer.writerow([headline, summary, yt_link]) + +csv_file.close() diff --git a/BeautifulSoup/simple.html b/BeautifulSoup/simple.html new file mode 100644 index 0000000..d7ee20c --- /dev/null +++ b/BeautifulSoup/simple.html @@ -0,0 +1,31 @@ + + + + Test - A Sample Website + + + + + +

Test Website

+
+
+

Article 1 Headline

+

This is a summary of article 1

+
+
+
+

Article 2 Headline

+

This is a summary of article 2

+
+
+ + + + + + + +