feed: support the parsing of OPDS type feeds

lumbunglib/feed.py

@@ -50,37 +50,60 @@ def create_frontmatter(entry):
     """
     parse RSS metadata and return as frontmatter 
     """
-    if "published" in entry:
+    if 'published' in entry:
         published = entry.published_parsed
-    if "updated" in entry:
+    if 'updated' in entry:
         published = entry.updated_parsed
 
     published = arrow.get(published)
 
-    if "author" in entry:
+    if 'author' in entry:
         author = entry.author
     else:
-        author = ""
+        author = ''
     
-    if "title" in entry:
+    if 'authors' in entry:
-        title = entry.title
+        authors = []
+        for a in entry.authors:
+            authors.append(a['name'])
+
+    if 'summary' in entry:
+        summary = entry.summary
     else:
-        title = ""
+        summary = ''
+
+    if 'publisher' in entry:
+        publisher = entry.publisher
+    else:
+        publisher = ''
 
     tags = []
-    if "tags" in entry:
+    if 'tags' in entry:
-        # TODO finish categories
+        #TODO finish categories
         for t in entry.tags:
-            tags.append(t["term"])
+            tags.append(t['term'])
 
+    if "opds" in entry:
         frontmatter = {
-        "title": title,
+        'title':entry.title,
-        "date": published.format(),
+        'date': published.format(),
-        "summary": "",
+        'summary': summary,
-        "author": author,
+        'author': ",".join(authors),
-        "original_link": entry.link,
+        'publisher': publisher,
-        "feed_name": entry["feed_name"],
+        'original_link': entry.links[0]['href'].replace('opds/cover/','books/'),
-        "tags": str(tags),
+        'feed_name': entry['feed_name'],
+        'tags': str(tags),
+        'category': "books"
+        }
+    else:
+        frontmatter = {
+        'title':entry.title,
+        'date': published.format(),
+        'summary': '',
+        'author': author,
+        'original_link': entry.link,
+        'feed_name': entry['feed_name'],
+        'tags': str(tags)
         }
 
     return frontmatter
@@ -132,28 +155,31 @@ def create_post(post_dir, entry):
         print("created post for", entry.title, "({})".format(entry.link))
 
 
-def grab_media(post_directory, url):
+def grab_media(post_directory, url, prefered_name=None):
     """
     download media linked in post to have local copy
     if download succeeds return new local path otherwise return url
     """
-    image = urlparse(url).path.split("/")[-1]
+    media_item = urlparse(url).path.split('/')[-1]
+
+    if prefered_name:
+        media_item = prefered_name
 
     try:
-        if not os.path.exists(os.path.join(post_directory, image)):
+        if not os.path.exists(os.path.join(post_directory, media_item)):
-            # TODO: stream is true is a conditional so we could check the headers for things, mimetype etc
+            #TODO: stream is true is a conditional so we could check the headers for things, mimetype etc
             response = requests.get(url, stream=True)
             if response.ok:
-                with open(os.path.join(post_directory, image), "wb") as img_file:
+                with open(os.path.join(post_directory, media_item), 'wb') as media_file:
-                    shutil.copyfileobj(response.raw, img_file)
+                    shutil.copyfileobj(response.raw, media_file)
-                    print("Downloaded cover image", image)
+                    print('Downloaded media item', media_item)
-                    return image
+                    return media_item
-            return image
+            return media_item
-        elif os.path.exists(os.path.join(post_directory, image)):
+        elif os.path.exists(os.path.join(post_directory, media_item)):
-            return image
+            return media_item
 
     except Exception as e:
-        print("Failed to download image", url)
+        print('Failed to download image', url)
         print(e)
     return url
 
@@ -208,6 +234,58 @@ def grab_feed(feed_url):
         return data
     return False
 
+def create_opds_post(post_dir, entry):
+    """
+    create a HUGO post based on OPDS entry
+    or update it if the timestamp is newer
+    Downloads the cover & file
+    """
+
+    frontmatter = create_frontmatter(entry)
+
+    template_dir = os.path.join(Path(__file__).parent.resolve(), "templates")
+    env = jinja2.Environment(loader=jinja2.FileSystemLoader(template_dir))
+    template = env.get_template("feed.md")
+
+    if not os.path.exists(post_dir):
+        os.makedirs(post_dir)
+
+    if os.path.exists(os.path.join(post_dir, '.timestamp')):
+        old_timestamp = open(os.path.join(post_dir, '.timestamp')).read()
+        old_timestamp = arrow.get(float(old_timestamp))
+        current_timestamp = arrow.get(entry['updated_parsed'])
+
+        if current_timestamp > old_timestamp:
+            pass
+        else:
+            print('Book "{}..." already up to date'.format(entry['title'][:32]))
+            return
+
+    for item in entry.links:
+        ft = item['type'].split('/')[-1]
+        fn = item['rel'].split('/')[-1]
+
+        if fn == "acquisition":
+            fn = "publication" #calling the publications acquisition is weird
+
+        prefered_name = "{}-{}.{}".format(fn, slugify(entry['title']), ft)
+
+        grab_media(post_dir, item['href'], prefered_name)
+
+        if "summary" in entry:
+            summary = entry.summary
+        else:
+            summary = ""
+
+    with open(os.path.join(post_dir,'index.md'),'w') as f:
+        post = template.render(frontmatter=sanitize_yaml(frontmatter), content=summary)
+        f.write(post)
+        print('created post for Book', entry.title)
+
+    with open(os.path.join(post_dir, '.timestamp'), 'w') as f:
+        timestamp = arrow.get(entry['updated_parsed'])
+        f.write(timestamp.format('X'))
+
 
 def main():
     feed_urls = open("feeds_list.txt", "r").read().splitlines()
@@ -252,6 +330,15 @@ def main():
         data = grab_feed(feed_url)
 
         if data:
+
+            opds_feed = False
+            for i in data.feed['links']:
+                if i['rel'] == 'self':
+                    if 'opds' in i['type']:
+                        opds_feed = True
+                        print("OPDS type feed!")
+
+
             for entry in data.entries:
                 # if 'tags' in entry:
                 #     for tag in entry.tags:
@@ -261,6 +348,7 @@ def main():
                 entry["feed_name"] = feed_name
 
                 post_name = slugify(entry.title)
+                
                 # pixelfed returns the whole post text as the post name. max
                 # filename length is 255 on many systems. here we're shortening
                 # the name and adding a hash to it to avoid a conflict in a
@@ -268,14 +356,26 @@ def main():
                 if len(post_name) > 150:
                     post_hash = md5(bytes(post_name, "utf-8"))
                     post_name = post_name[:150] + "-" + post_hash.hexdigest()
+                
+                if opds_feed:
+                    entry['opds'] = True
+                    #format: Beyond-Debiasing-Report_Online-75535a4886e3
+                    post_name = slugify(entry['title'])+'-'+entry['id'].split('-')[-1]
+
                 post_dir = os.path.join(output_dir, feed_name, post_name)
 
                 if post_name not in existing_posts:
                     # if there is a blog entry we dont already have, make it
+                    if opds_feed:
+                        create_opds_post(post_dir, entry)
+                    else:
                         create_post(post_dir, entry)
 
                 elif post_name in existing_posts:
                     # if we already have it, update it
+                    if opds_feed:
+                        create_opds_post(post_dir, entry)
+                    else:
                         create_post(post_dir, entry)   
                     existing_posts.remove(
                         post_name