Can now view articles

robputt · Dec 19, 2017 · b27cb81 · b27cb81
1 parent f36fc73
commit b27cb81
Show file tree

Hide file tree

Showing 6 changed files with 34 additions and 4 deletions.
diff --git a/requirements.txt b/requirements.txt
@@ -1,4 +1,5 @@
 bs4
 requests
 sqlalchemy
-pymysql
+pymysql
+flask
diff --git a/setup.py b/setup.py
@@ -27,5 +27,6 @@
     # Adds dependencies
     install_requires = ['bs4',
                         'sqlalchemy',
-                        'pymysql']
+                        'pymysql',
+                        'flask']
 )
diff --git a/this_is_scraper/scraper.py b/this_is_scraper/scraper.py
@@ -90,7 +90,7 @@ def extract_content(link):
         paragraphs = div.findAll("p")
     ret_text = ""
     for paragraph in paragraphs:
-        ret_text += "%s \r\n\r\n" % paragraph.get_text().strip()
+        ret_text += "%s <br/><br/>" % paragraph.get_text().strip()
     return title, ret_text.encode('utf-8')
 
 

diff --git a/this_is_scraper/templates/article_list.html b/this_is_scraper/templates/article_list.html
@@ -27,7 +27,7 @@ <h2>ThisIsScraper &middot; Article Viewer</h2>
 					{% else %}
 						<td>{{ article.article_status }}</td>
 					{% endif %}
-					<td>{% if article.article_status == 'complete' %}<a href='/view'>View Summary</a> &middot; {% endif %}<a href='{{ article.article_link }}' target='_blank'>View Original</a></td>
+					<td>{% if article.article_status == 'complete' %}<a href='/view?url={{ article.article_link }}'>View Summary</a> &middot; {% endif %}<a href='{{ article.article_link }}' target='_blank'>View Original</a></td>
 				</tr>
 			{% endfor %}
 		</tbody>

diff --git a/this_is_scraper/templates/view_article.html b/this_is_scraper/templates/view_article.html
@@ -0,0 +1,12 @@
+<html>
+<head>
+	<title>ThisIsScraper &middot {{ article.article_title }}</title>
+</head>
+<body>
+	<h2>ThisIsScraper &middot; Article Viewer</h2>
+	<p><a href='/'>&lt; Back</a></p>
+	<h3>{{ article.article_title }}</h3>
+	<p>{{ article.article_dt.isoformat() }}</p>
+	<p>{{ article.article_content|safe }}</p>
+</body>
+</html>
diff --git a/this_is_scraper/viewer.py b/this_is_scraper/viewer.py
@@ -1,6 +1,7 @@
 from flask import Flask
 from flask import g
 from flask import render_template
+from flask import request
 from this_is_scraper.db import get_db_session
 from this_is_scraper.db import Articles
 
@@ -27,12 +28,27 @@ def get_articles():
     return articles
 
 
+def get_article(link):
+    article = g.db.query(Articles).filter(Articles.article_link == link).one()
+    return article
+
+
 @app.route('/')
 def display_article_list():
     articles = get_articles()
     return render_template('article_list.html',
                            articles=articles)
 
 
+@app.route('/view')
+def display_article():
+    url = request.args.get('url')
+    article = get_article(url)
+    content = article.article_content.replace('\r\n', '<br/>')
+    return render_template('view_article.html',
+                           article=article,
+                           content=content)
+
+
 if __name__ == "__main__":
     app.run('0.0.0.0', 5000, threaded=True, debug=True)