From 703df9c8e97f3e8e2df1b106fea616aa900a1950 2018-09-21 14:57:14
From: Brett Smith <brett@sfconservancy.org>
Date: 2018-09-21 14:57:14
Subject: [PATCH] Blogs/news only include a single OG image/video.

For now, this gives us more assurance that other sites will choose the
preview we want.

You can control the selection by adding data-ogpreview to image, video, and
source elements.  data-ogpreview=0 excludes the element from being included
in the preview.  Positive numbers set the preview priority.  The lowest
value found is chosen first.

---

diff --git a/www/conservancy/bsoup.py b/www/conservancy/bsoup.py
index fb0ef6cb3d2ad50322ecc1a17f4212f713727a91..28ccdd08826e017d0e1c0303efd5a4ea2c9d66f0 100644
--- a/www/conservancy/bsoup.py
+++ b/www/conservancy/bsoup.py
@@ -1,6 +1,7 @@
 # -*- encoding: utf-8 -*-
 
 import io
+import itertools
 import re
 
 import bs4
@@ -86,25 +87,22 @@ class BeautifulSoup(bs4.BeautifulSoup):
         except AttributeError:
             return False
 
-    def iter_attr(self, tag, attr_name, **kwargs):
-        kwargs[attr_name] = True
-        for elem in self.find_all(tag, **kwargs):
-            yield elem[attr_name]
+    def iter_images(self):
+        """Return an iterator of all image elements in this document.
 
-    def iter_image_urls(self):
-        """Return an iterator of source URL strings of all images in this document.
-
-        Images include <img> tags and <video> poster attributes.
+        Images include <img> and <video> with a poster attribute.
         """
         for elem in self.find_all(list(self.IMAGE_ATTRS.keys())):
             try:
-                yield elem[self.IMAGE_ATTRS[elem.name]]
+                elem[self.IMAGE_ATTRS[elem.name]]
             except KeyError:
                 pass
+            else:
+                yield elem
 
-    def iter_video_urls(self):
-        """Return an iterator of source URL strings of all videos in this document."""
-        return self.iter_attr(self.is_video_source, 'src')
+    def iter_videos(self):
+        """Return an iterator of all video source elements in this document."""
+        return self.find_all(self.is_video_source, src=True)
 
 
 class SoupModelMixin:
@@ -115,6 +113,7 @@ class SoupModelMixin:
     are usable.
     """
 
+    OG_PREVIEW_ATTR = 'data-ogpreview'
     SOUP_ATTRS = []
 
     def _get_soup(self):
@@ -128,17 +127,57 @@ class SoupModelMixin:
             self._soup = BeautifulSoup(html)
             return self._soup
 
+    def _elem_key(self, attr_name=OG_PREVIEW_ATTR, getvalue=int, fallback=999999):
+        def elem_sort_key(elem):
+            try:
+                sort_key = getvalue(elem[attr_name])
+            except (KeyError, ValueError):
+                sort_key = fallback
+            elem[attr_name] = sort_key
+            return sort_key
+        return elem_sort_key
+
+    def _elem_pred(self, attr_name=OG_PREVIEW_ATTR, test=lambda n: n > 0):
+        def elem_pred(elem):
+            return test(elem[attr_name])
+        return elem_pred
+
+    def _sort_and_slice_elems(self, elem_seq, elem_key, pred, *slice_args):
+        seq = itertools.ifilter(pred, sorted(elem_seq, key=elem_key))
+        if slice_args:
+            return itertools.islice(seq, *slice_args)
+        else:
+            return seq
+
     def get_description(self):
         """Return a string with a brief excerpt of body text from the HTML."""
         return u''.join(self._get_soup().some_body_text())
 
-    def get_image_urls(self):
+    def get_image_urls(self, *slice_args):
         """Return an iterator of source URL strings of all images in the HTML.
 
-        Images include <img> tags and <video> poster attributes.
+        Images include <img> sources and <video> poster attributes.
         """
-        return self._get_soup().iter_image_urls()
-
-    def get_video_urls(self):
+        for elem in self._sort_and_slice_elems(
+                self._get_soup().iter_images(),
+                self._elem_key(),
+                self._elem_pred(),
+                *slice_args
+        ):
+            yield elem[BeautifulSoup.IMAGE_ATTRS[elem.name]]
+
+    def get_one_image_url(self):
+        return self.get_image_urls(1)
+
+    def get_video_urls(self, *slice_args):
         """Return an iterator of source URL strings of all videos in the HTML."""
-        return self._get_soup().iter_video_urls()
+        for elem in self._sort_and_slice_elems(
+                self._get_soup().iter_videos(),
+                self._elem_key(),
+                self._elem_pred(),
+                *slice_args
+        ):
+            yield elem['src']
+
+    def get_one_video_url(self):
+        return self.get_video_urls(1)
diff --git a/www/conservancy/templates/blog/entry_detail.html b/www/conservancy/templates/blog/entry_detail.html
index 1949694ab5e086f1737279d00fcb906d7ae52d4b..36e9e1b05765afcb9b35d78c1eab46e22006fe41 100644
--- a/www/conservancy/templates/blog/entry_detail.html
+++ b/www/conservancy/templates/blog/entry_detail.html
@@ -2,8 +2,8 @@
 
 {% block head %}
 {% include "opengraph_partial.html" with url=object.get_absolute_url title=object.headline description=object.get_description %}
-{% include "opengraph_urllist_partial.html" with property='image' urls=object.get_image_urls fallback='/img/conservancy-logo.png' %}
-{% include "opengraph_urllist_partial.html" with property='video' urls=object.get_video_urls %}
+{% include "opengraph_urllist_partial.html" with property='image' urls=object.get_one_image_url fallback='/img/conservancy-logo.png' %}
+{% include "opengraph_urllist_partial.html" with property='video' urls=object.get_one_video_url %}
 {% endblock %}
 
 {% block subtitle %}{{ object.headline|striptags|safe }} - Conservancy Blog - {% endblock %}
diff --git a/www/conservancy/templates/news/pressrelease_detail.html b/www/conservancy/templates/news/pressrelease_detail.html
index bdde649d5836ef9be3e1e585a4a8abdd71154bc7..563eed3fb7230d4f3e57780752ddb58e11ba6032 100644
--- a/www/conservancy/templates/news/pressrelease_detail.html
+++ b/www/conservancy/templates/news/pressrelease_detail.html
@@ -2,8 +2,8 @@
 
 {% block head %}
 {% include "opengraph_partial.html" with url=object.get_absolute_url title=object.headline description=object.get_description %}
-{% include "opengraph_urllist_partial.html" with property='image' urls=object.get_image_urls fallback='/img/conservancy-logo.png' %}
-{% include "opengraph_urllist_partial.html" with property='video' urls=object.get_video_urls %}
+{% include "opengraph_urllist_partial.html" with property='image' urls=object.get_one_image_url fallback='/img/conservancy-logo.png' %}
+{% include "opengraph_urllist_partial.html" with property='video' urls=object.get_one_video_url %}
 {% endblock %}
 
 {% block subtitle %}{{ object.headline|striptags|safe }} - {% endblock %}