handle OPDS feeds from calibre-web

rss_aggregator.py

@@ -1,7 +1,7 @@
 #!/bin/python3
 
-#lumbung.space rss feed aggregator
-#© 2021 roel roscam abbing gplv3 etc
+#lumbung.space rss & opds feed parser
+#© 2022 roel roscam abbing agplv3 etc
 
 import requests
 import jinja2
@@ -60,6 +60,21 @@ def create_frontmatter(entry):
         author = entry.author
     else:
         author = ''
+    
+    if 'authors' in entry:
+        authors = []
+        for a in entry.authors:
+            authors.append(a['name'])
+
+    if 'summary' in entry:
+        summary = entry.summary
+    else:
+        summary = ''
+
+    if 'publisher' in entry:
+        publisher = entry.publisher
+    else:
+        publisher = ''
 
     tags = []
     if 'tags' in entry:
@@ -67,15 +82,28 @@ def create_frontmatter(entry):
         for t in entry.tags:
             tags.append(t['term'])
 
-    frontmatter = {
-    'title':entry.title,
-    'date': published.format(),
-    'summary': '',
-    'author': author,
-    'original_link': entry.link,
-    'feed_name': entry['feed_name'],
-    'tags': str(tags)
-    }
+    if "opds" in entry:
+        frontmatter = {
+        'title':entry.title,
+        'date': published.format(),
+        'summary': summary,
+        'author': ",".join(authors),
+        'publisher': publisher,
+        'original_link': entry.links[0]['href'].replace('opds/cover/','books/'),
+        '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
 
@@ -100,32 +128,34 @@ def create_post(post_dir, entry):
         f.write(post)
         print('created post for', entry.title, '({})'.format(entry.link))
 
-def grab_media(post_directory, url):
+def grab_media(post_directory, url, prefered_name):
     """
     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
             response = requests.get(url, stream=True)
             if response.ok:
-                with open(os.path.join(post_directory, image), 'wb') as img_file:
-                    shutil.copyfileobj(response.raw, img_file)
-                    print('Downloaded cover image', image)
-                    return image
-            return image
-        elif os.path.exists(os.path.join(post_directory, image)):
-            return image
+                with open(os.path.join(post_directory, media_item), 'wb') as media_file:
+                    shutil.copyfileobj(response.raw, media_file)
+                    print('Downloaded media item', media_item)
+                    return media_item
+            return media_item
+        elif os.path.exists(os.path.join(post_directory, media_item)):
+            return media_item
 
     except Exception as e:
         print('Failed to download image', url)
         print(e)
     return url
 
-
 def parse_posts(post_dir, post_content):
     """
     parse the post content to for media items
@@ -147,6 +177,54 @@ def parse_posts(post_dir, post_content):
             iframe.decompose()
     return soup.decode()
 
+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)
+
+    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=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 grab_feed(feed_url):
     """
     check whether feed has been updated
@@ -219,6 +297,14 @@ for feed_url in feed_urls:
     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!")
+
         for entry in data.entries:
              # if 'tags' in entry:
              #     for tag in entry.tags:
@@ -228,15 +314,28 @@ for feed_url in feed_urls:
             entry['feed_name'] = feed_name
 
             post_name = slugify(entry.title)
+
+            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
-                create_post(post_dir, entry)
+                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
-                create_post(post_dir, entry)                
+                if opds_feed:
+                    create_opds_post(post_dir, entry)
+                else:
+                    create_post(post_dir, entry)                
                 existing_posts.remove(post_name) # create list of posts which have not been returned by the feed
 
         for post in existing_posts: