Add "view image" functionality (#268)

* add view image option

* prevent whoogle links from opening in a new tab.

* remove view image template on mobile requests

* change loop values to be more robust to the number of images

* Update app/templates/imageresults.html

* fix "Basically the .cvifge class needs width: 100%; in order to expand the search input to fit the form width."

* Update app/templates/imageresults.html

* remove hardcoded string from template

* Add view image config var to app.json

* Add view image config var to whoogle.env

Co-authored-by: jacr13 <ramos.joao@protonmail.com>
Co-authored-by: Ben Busby <benbusby@protonmail.com>

README.md

@@ -103,17 +103,25 @@ Sandboxed temporary instance:
 
 ```bash
 $ whoogle-search --help
-usage: whoogle-search [-h] [--port <port number>] [--host <ip address>] [--debug]
+usage: whoogle-search [-h] [--port <port number>] [--host <ip address>] [--debug] [--https-only] [--userpass <username:password>]
-                      [--https-only]
+                      [--proxyauth <username:password>] [--proxytype <socks4|socks5|http>] [--proxyloc <location:port>]
 
 Whoogle Search console runner
 
 optional arguments:
-  -h, --help            show this help message and exit
+  -h, --help            Show this help message and exit
   --port <port number>  Specifies a port to run on (default 5000)
   --host <ip address>   Specifies the host address to use (default 127.0.0.1)
   --debug               Activates debug mode for the server (default False)
-  --https-only          Enforces HTTPS redirects for all requests (default False)
+  --https-only          Enforces HTTPS redirects for all requests
+  --userpass <username:password>
+                        Sets a username/password basic auth combo (default None)
+  --proxyauth <username:password>
+                        Sets a username/password for a HTTP/SOCKS proxy (default None)
+  --proxytype <socks4|socks5|http>
+                        Sets a proxy type for all connections (default None)
+  --proxyloc <location:port>
+                        Sets a proxy location for all connections (default None)
 ```
 See the [available environment variables](#environment-variables) for additional configuration.
 
@@ -286,6 +294,7 @@ These environment variables allow setting default config values, but can be over
 | WHOOGLE_CONFIG_ALTS            | Use social media site alternatives (nitter, invidious, etc)     |
 | WHOOGLE_CONFIG_TOR             | Use Tor routing (if available)                                  |
 | WHOOGLE_CONFIG_NEW_TAB         | Always open results in new tab                                  |
+| WHOOGLE_CONFIG_VIEW_IMAGE      | Enable View Image option                                        |
 | WHOOGLE_CONFIG_GET_ONLY        | Search using GET requests only                                  |
 | WHOOGLE_CONFIG_URL             | The root url of the instance (`https://<your url>/`)            |
 | WHOOGLE_CONFIG_STYLE           | The custom CSS to use for styling (should be single line)       |

app.json

@@ -114,6 +114,11 @@
         "description": "[CONFIG] Always open results in new tab (set to 1 or leave blank)",
         "value": "",
         "required": false
+    },
+    "WHOOGLE_CONFIG_VIEW_IMAGE": {
+      "description": "[CONFIG] Enable View Image option (set to 1 or leave blank)",
+      "value": "",
+      "required": false
   },
     "WHOOGLE_CONFIG_GET_ONLY": {
         "description": "[CONFIG] Search using GET requests only (set to 1 or leave blank)",

app/filter.py

@@ -254,3 +254,65 @@ class Filter:
 
             # Replace link destination
             link_desc[0].replace_with(get_site_alt(link_desc[0]))
+
+    def view_image(self, soup) -> BeautifulSoup:
+        """Replaces the soup with a new one that handles mobile results and
+        adds the link of the image full res to the results.
+
+        Args:
+            soup: A BeautifulSoup object containing the image mobile results.
+
+        Returns:
+            BeautifulSoup: The new BeautifulSoup object
+        """
+
+        # get some tags that are unchanged between mobile and pc versions
+        search_input = soup.find_all('td', attrs={'class': "O4cRJf"})[0]
+        search_options = soup.find_all('div', attrs={'class': "M7pB2"})[0]
+        cor_suggested = soup.find_all('table', attrs={'class': "By0U9"})
+        next_pages = soup.find_all('table', attrs={'class': "uZgmoc"})[0]
+        information = soup.find_all('div', attrs={'class': "TuS8Ad"})[0]
+
+        results = []
+        # find results div
+        results_div = soup.find_all('div', attrs={'class': "nQvrDb"})[0]
+        # find all the results
+        results_all = results_div.find_all('div', attrs={'class': "lIMUZd"})
+
+        for item in results_all:
+            urls = item.find('a')['href'].split('&imgrefurl=')
+
+            img_url = urlparse.unquote(urls[0].replace('/imgres?imgurl=', ''))
+            webpage = urlparse.unquote(urls[1].split('&')[0])
+            img_tbn = urlparse.unquote(item.find('a').find('img')['src'])
+            results.append({
+                'domain': urlparse.urlparse(webpage).netloc,
+                'img_url': img_url,
+                'webpage': webpage,
+                'img_tbn': img_tbn
+            })
+
+        soup = BeautifulSoup(render_template('imageresults.html',
+                                             length=len(results),
+                                             results=results,
+                                             view_label="View Image"),
+                             features='html.parser')
+        # replace search input object
+        soup.find_all('td',
+                      attrs={'class': "O4cRJf"})[0].replaceWith(search_input)
+        # replace search options object (All, Images, Videos, etc.)
+        soup.find_all('div',
+                      attrs={'class': "M7pB2"})[0].replaceWith(search_options)
+        # replace correction suggested by google object if exists
+        if len(cor_suggested):
+            soup.find_all(
+                'table',
+                attrs={'class': "By0U9"}
+            )[0].replaceWith(cor_suggested[0])
+        # replace next page object at the bottom of the page
+        soup.find_all('table',
+                      attrs={'class': "uZgmoc"})[0].replaceWith(next_pages)
+        # replace information about user connection at the bottom of the page
+        soup.find_all('div',
+                      attrs={'class': "TuS8Ad"})[0].replaceWith(information)
+        return soup

app/models/config.py

@@ -27,7 +27,9 @@ class Config:
         self.tor = read_config_bool('WHOOGLE_CONFIG_TOR')
         self.near = os.getenv('WHOOGLE_CONFIG_NEAR', '')
         self.new_tab = read_config_bool('WHOOGLE_CONFIG_NEW_TAB')
+        self.view_image = read_config_bool('WHOOGLE_CONFIG_VIEW_IMAGE')
         self.get_only = read_config_bool('WHOOGLE_CONFIG_GET_ONLY')
+
         self.safe_keys = [
             'lang_search',
             'lang_interface',

app/request.py

@@ -151,6 +151,8 @@ class Request:
         self.language = config.lang_search
         self.mobile = 'Android' in normal_ua or 'iPhone' in normal_ua
         self.modified_user_agent = gen_user_agent(self.mobile)
+        if not self.mobile:
+            self.modified_user_agent_mobile = gen_user_agent(True)
 
         # Set up proxy, if previously configured
         if os.environ.get('WHOOGLE_PROXY_LOC'):
@@ -197,7 +199,8 @@ class Request:
         return [_.attrib['data'] for _ in
                 root.findall('.//suggestion/[@data]')]
 
-    def send(self, base_url=SEARCH_URL, query='', attempt=0) -> Response:
+    def send(self, base_url=SEARCH_URL, query='', attempt=0,
+             force_mobile=False) -> Response:
         """Sends an outbound request to a URL. Optionally sends the request
         using Tor, if enabled by the user.
 
@@ -211,8 +214,13 @@ class Request:
             Response: The Response object returned by the requests call
 
         """
+        if force_mobile and not self.mobile:
+            modified_user_agent = self.modified_user_agent_mobile
+        else:
+            modified_user_agent = self.modified_user_agent
+
         headers = {
-            'User-Agent': self.modified_user_agent
+            'User-Agent': modified_user_agent
         }
 
         # FIXME: Should investigate this further to ensure the consent

app/templates/imageresults.html

@@ -0,0 +1,116 @@
+<!DOCTYPE html>
+<html>
+ <head>
+  <meta content="application/xhtml+xml; charset=utf-8" http-equiv="Content-Type"/>
+  <meta content="no-cache" name="Cache-Control"/>
+  <title>
+  </title>
+  <style>
+   a{text-decoration:none;color:inherit}a:hover{text-decoration:underline}a img{border:0}body{font-family:Roboto,Helvetica,Arial,sans-serif;padding:8px;margin:0 auto;max-width:700px;min-width:240px;}.FbhRzb{border-left:thin solid #dadce0;border-right:thin solid #dadce0;border-top:thin solid #dadce0;height:40px;overflow:hidden}.n692Zd{margin-bottom:10px}.cvifge{height:40px;border-spacing:0;width:100%;}.QvGUP{height:40px;padding:0 8px 0 8px;vertical-align:top}.O4cRJf{height:40px;width:100%;padding:0;padding-right:16px}.O1ePr{height:40px;padding:0;vertical-align:top}.kgJEQe{height:36px;width:98px;vertical-align:top;margin-top:4px}.lXLRf{vertical-align:top}.MhzMZd{border:0;vertical-align:middle;font-size:14px;height:40px;padding:0;width:100%;padding-left:16px}.xB0fq{height:40px;border:none;font-size:14px;background-color:#4285f4;color:#fff;padding:0 16px;margin:0;vertical-align:top;cursor:pointer}.xB0fq:focus{border:1px solid #000}.M7pB2{border:thin solid #dadce0;margin:0 0 3px 0;font-size:13px;font-weight:500;height:40px}.euZec{width:100%;height:40px;text-align:center;border-spacing:0}table.euZec td{padding:0;width:25%}.QIqI7{display:inline-block;padding-top:4px;font-weight:bold;color:#4285f4}.EY24We{border-bottom:2px solid #4285f4}.CsQyDc{display:inline-block;color:#70757a}.TuS8Ad{font-size:14px}.HddGcc{padding:8px;color:#70757a}.dzp8ae{font-weight:bold;color:#3c4043}.rEM8G{color:#70757a}.bookcf{table-layout:fixed;width:100%;border-spacing:0}.InWNIe{text-align:center}.uZgmoc{border:thin solid #dadce0;color:#70757a;font-size:14px;text-align:center;table-layout:fixed;width:100%}.frGj1b{display:block;padding:12px 0 12px 0;width:100%}.BnJWBc{text-align:center;padding:6px 0 13px 0;height:35px}.e3goi{vertical-align:top;padding:0;height:180px}.GpQGbf{margin:auto;border-collapse:collapse;border-spacing:0;width:100%}
+  </style>
+ </head>
+ <body>
+  <style>
+   .X6ZCif{color:#202124;font-size:11px;line-height:16px;display:inline-block;padding-top:2px;overflow:hidden;padding-bottom:4px;width:100%}.TwVfHd{border-radius:16px;border:thin solid #dadce0;display:inline-block;padding:8px 8px;margin-right:8px;margin-bottom:4px}.yekiAe{background-color:#dadce0}.svla5d{width:100%}.ezO2md{border:thin solid #dadce0;padding:12px 16px 12px 16px;margin-bottom:10px;font-family:Roboto,Helvetica,Arial,sans-serif}.lIMUZd{font-family:Roboto,Helvetica,Arial,sans-serif}.TxbwNb{border-spacing:0}.K35ahc{width:100%}.owohpf{text-align:center}.RAyV4b{width:162px;height:140px;line-height:140px;overflow:'hidden';text-align:center;}.t0fcAb{text-align:center;margin:auto;vertical-align:middle;width:100%;height:100%;object-fit: contain}.Tor4Ec{padding-top:2px;padding-bottom:8px;}.fYyStc{word-break:break-word}.ynsChf{display:block;white-space:nowrap;overflow:hidden;text-overflow:ellipsis}.Fj3V3b{color:#1967D2;font-size:14px;line-height:20px}.FrIlee{color:#202124;font-size:11px;line-height:16px}.F9iS2e{color:#70757a;font-size:11px;line-height:16px}.WMQ2Le{color:#70757a;font-size:12px;line-height:16px}.x3G5ab{color:#202124;font-size:12px;line-height:16px}.fuLhoc{color:#1967D2;font-size:18px;line-height:24px}.epoveb{font-size:32px;line-height:40px;font-weight:400;color:#202124}.dXDvrc{color:#0d652d;font-size:14px;line-height:20px;word-wrap:break-word}.dloBPe{font-weight:bold}.YVIcad{color:#70757a}.JkVVdd{color:#ea4335}.oXZRFd{color:#ea4335}.MQHtg{color:#fbbc04}.pyMRrb{color:#1e8e3e}.EtTZid{color:#1e8e3e}.M3vVJe{color:#1967D2}.qXLe6d{display:block}.NHQNef{font-style:italic}.Cb8Z7c{white-space:pre}a.ZWRArf{text-decoration:none}a .CVA68e:hover{text-decoration:underline}
+  </style>
+  <div class="n692Zd">
+   <div class="BnJWBc">
+    <a class="lXLRf" href="/?safe=off&amp;gbv=1&amp;output=images&amp;ie=UTF-8&amp;tbm=isch&amp;sa=X&amp;ved=0ahUKEwjhh7TZyd_vAhWShf0HHeYzCmsQPAgC">
+     <img alt="Google" class="kgJEQe" src="/images/branding/searchlogo/1x/googlelogo_desk_heirloom_color_150x55dp.gif"/>
+    </a>
+   </div>
+   <div class="FbhRzb">
+    <form action="/search">
+     <input name="safe" type="hidden" value="off"/>
+     <input name="gbv" type="hidden" value="1"/>
+     <input name="ie" type="hidden" value="ISO-8859-1"/>
+     <input name="tbm" type="hidden" value="isch"/>
+     <input name="oq" type="hidden"/>
+     <input name="aqs" type="hidden"/>
+     <table class="cvifge">
+      <tr>
+       <td class="O4cRJf">
+         <!-- search input -->
+       </td>
+      </tr>
+     </table>
+    </form>
+   </div>
+   <div class="M7pB2">
+    <!-- search options -->
+   </div>
+  </div>
+  <!-- <div class="X6ZCif"> Not present in mobile
+  </div> -->
+  <div>
+    <div>
+      <div>
+       <div class="lIMUZd">
+        <table class="By0U9">
+          <!-- correction suggested -->
+        </table>
+       </div>
+      </div>
+    </div>
+    <table class="GpQGbf">
+      {% for i in range((length // 4) + 1) %}
+      <tr>
+        {% for j in range([length - (i*4), 4]|min) %}
+          <td align="center" class="e3goi">
+          <div class="svla5d">
+            <div>
+            <div class="lIMUZd">
+              <div>
+              <table class="TxbwNb">
+                <tr>
+                <td>
+                  <a href="{{ results[(i*4)+j].webpage }}">
+                  <div class="RAyV4b">
+                    <img alt="" class="t0fcAb" src="{{ results[(i*4)+j].img_tbn }}"/>
+                  </div>
+                  </a>
+                </td>
+                </tr>
+                <tr>
+                <td>
+                  <a href="{{ results[(i*4)+j].webpage }}">
+                  <div class="Tor4Ec">
+                    <span class="qXLe6d x3G5ab">
+                    <span class="fYyStc">
+                      {{ results[(i*4)+j].domain }}
+                    </span>
+                    </span>
+                  </div>
+                  </a>
+                  <a href="{{ results[(i*4)+j].img_url }}">
+                    <div class="Tor4Ec">
+                      <span class="qXLe6d F9iS2e">
+                        <span class="fYyStc">
+                          {{ view_label }}
+                        </span>
+                        </span>
+                    </div>
+                  </a>
+                </td>
+                </tr>
+              </table>
+              </div>
+            </div>
+            </div>
+          </div>
+          </td>
+          {% endfor %}
+        </tr>
+      {% endfor %}
+    </table>
+  </div>
+  <table class="uZgmoc">
+    <!-- next page object -->
+  </table>
+  <br/>
+  <div class="TuS8Ad">
+    <!-- information about user connection  -->
+  <div>
+  </div>
+ </body>
+</html>

app/templates/index.html

@@ -132,6 +132,12 @@
 					<label for="config-new-tab">Open Links in New Tab: </label>
 					<input type="checkbox" name="new_tab" id="config-new-tab" {{ 'checked' if config.new_tab else '' }}>
 				</div>
+				<div class="config-div config-div-view-image">
+					<label for="config-view-image">Full Size Image Search: </label>
+					<input type="checkbox" name="view_image" id="config-view-image" {{ 'checked' if config.view_image else '' }}>
+					<div><span class="info-text"> — (Experimental) Adds the "View Image" option on desktop to view full size images in search results.
+					This will cause image result thumbnails to be lower resolution.</span></div>
+				</div>
 				<div class="config-div config-div-tor">
 					<label for="config-tor">Use Tor: {{ '' if tor_available else 'Unavailable' }}</label>
 					<input type="checkbox" name="tor" id="config-tor" {{ '' if tor_available else 'hidden' }} {{ 'checked' if config.tor else '' }}>

app/utils/search.py

@@ -119,11 +119,23 @@ class Search:
                                self.request_params,
                                self.config,
                                content_filter.near)
-        get_body = g.user_request.send(query=full_query)
+
+        # force mobile search when view image is true and
+        # the request is not already made by a mobile
+        view_image = ('tbm=isch' in full_query
+                      and self.config.view_image
+                      and not g.user_request.mobile)
+
+        get_body = g.user_request.send(query=full_query,
+                                       force_mobile=view_image)
 
         # Produce cleanable html soup from response
         html_soup = bsoup(content_filter.reskin(get_body.text), 'html.parser')
 
+        # Replace current soup if view_image is active
+        if view_image:
+            html_soup = content_filter.view_image(html_soup)
+
         # Indicate whether or not a Tor connection is active
         tor_banner = bsoup('', 'html.parser')
         if g.user_request.tor_valid:

whoogle.env

@@ -46,6 +46,9 @@
 # Open results in new tab
 #WHOOGLE_CONFIG_NEW_TAB=1
 
+# Enable View Image option
+#WHOOGLE_CONFIG_VIEW_IMAGE=1 
+
 # Search using GET requests only (exposes query in logs)
 #WHOOGLE_CONFIG_GET_ONLY=1