From e5b24ac102ede510d7f4f6b882d8617db1733dc3 Mon Sep 17 00:00:00 2001
From: Paul Rothrock <paul@movetoiceland.com>
Date: Thu, 14 May 2020 22:26:22 -0400
Subject: [PATCH] Putting '! ' at the beginning of the query now redirects to
 the first search result

Signed-off-by: Paul Rothrock <paul@movetoiceland.com>
---
 app/filter.py       | 34 ++++++++++++++++++++++++++++++++++
 app/routes.py       | 17 +++++++++++++++--
 test/test_routes.py |  4 ++++
 3 files changed, 53 insertions(+), 2 deletions(-)

diff --git a/app/filter.py b/app/filter.py
index a77116c..93cb812 100644
--- a/app/filter.py
+++ b/app/filter.py
@@ -58,6 +58,40 @@ class Filter:
 
         return soup
 
+    def get_first_url(self, soup):
+        # Replace hrefs with only the intended destination (no "utm" type tags)
+        for a in soup.find_all('a', href=True):
+            href = a['href'].replace('https://www.google.com', '')
+            
+            result_link = urlparse.urlparse(href)
+            query_link = parse_qs(result_link.query)['q'][0] if '?q=' in href else ''
+
+            # Return the first search result URL
+            if 'url?q=' in href:
+                parsed_link = urlparse.urlparse(query_link)
+                link_args = parse_qs(parsed_link.query)
+                safe_args = {}
+
+                if len(link_args) == 0 and len(parsed_link) > 0:
+                    return query_link
+                    
+
+                for arg in link_args.keys():
+                    if arg in SKIP_ARGS:
+                        continue
+
+                    safe_args[arg] = link_args[arg]
+
+                # Remove original link query and replace with filtered args
+                query_link = query_link.replace(parsed_link.query, '')
+                if len(safe_args) > 0:
+                    query_link = query_link + urlparse.urlencode(safe_args, doseq=True)
+                else:
+                    query_link = query_link.replace('?', '')
+
+                print(query_link)
+                return query_link
+
     def remove_ads(self, soup):
         main_divs = soup.find('div', {'id': 'main'})
         if main_divs is None:
diff --git a/app/routes.py b/app/routes.py
index 4fa3c93..958efd3 100644
--- a/app/routes.py
+++ b/app/routes.py
@@ -64,7 +64,7 @@ def opensearch():
 def search():
     request_params = request.args if request.method == 'GET' else request.form
     q = request_params.get('q')
-
+    
     if q is None or len(q) == 0:
         return redirect('/')
     else:
@@ -74,6 +74,11 @@ def search():
         except InvalidToken:
             pass
 
+    feeling_lucky = q.startswith("! ")
+
+    if feeling_lucky:
+        q = q[2:]
+
     user_agent = request.headers.get('User-Agent')
     mobile = 'Android' in user_agent or 'iPhone' in user_agent
 
@@ -82,7 +87,15 @@ def search():
     get_body = g.user_request.send(query=full_query)
 
     results = content_filter.reskin(get_body)
-    formatted_results = content_filter.clean(BeautifulSoup(results, 'html.parser'))
+    dirty_soup = BeautifulSoup(results, 'html.parser')
+
+    if feeling_lucky:
+        redirect_url = content_filter.get_first_url(dirty_soup)
+        return redirect(redirect_url, 303) # Using 303 so the browser performs a GET request for the URL
+    else:
+        formatted_results = content_filter.clean(dirty_soup)
+
+
 
     return render_template('display.html', query=urlparse.unquote(q), response=formatted_results)
 
diff --git a/test/test_routes.py b/test/test_routes.py
index 1124f5f..91e17be 100644
--- a/test/test_routes.py
+++ b/test/test_routes.py
@@ -17,6 +17,10 @@ def test_search(client):
     rv = client.get('/search?q=test')
     assert rv._status_code == 200
 
+def test_feeling_lucky(client):
+    rv = client.get('/search?q=!%20test')
+    assert rv._status_code == 303
+
 
 def test_config(client):
     rv = client.post('/config', data=demo_config)