Update Linkedin_Data_API_Calls.py

Linkedin_Data_API_Calls.py

@@ -69,7 +69,45 @@ def _get_sentiment_from_text(text_to_analyze):
 
     return {"label": dominant_sentiment_label, "counts": sentiment_counts}
 
-
+def get_post_media_category(post_content):
+    """
+    Determines the media category from the post's content object.
+    Args:
+        post_content (dict or None): The content dictionary of the post.
+    Returns:
+        str: The determined media category (e.g., "Video", "Article", "Document", "Image", "Multi-Image", "NONE").
+    """
+    if not post_content:
+        return "NONE"
+
+    # 1. Check for specific LinkedIn Video Component (from your original logic)
+    # You might want to refine this if 'mediaCategory' within the video component is more specific
+    if "com.linkedin.voyager.feed.render.LinkedInVideoComponent" in post_content:
+        # video_component_data = post_content.get("com.linkedin.voyager.feed.render.LinkedInVideoComponent", {})
+        # return video_component_data.get("mediaCategory", "Video") # Example if you want to use its specific category
+        return "Video"
+
+    # 2. Check for Article (based on your "old code" and examples)
+    if 'article' in post_content:
+        return "Article"
+
+    # 3. Check for Multi-Image (based on your "old code")
+    if 'multiImage' in post_content:
+        return "Multi-Image"
+
+    # 4. Check for Media (Document or Image - based on your "old code" and examples)
+    if 'media' in post_content:
+        media_item = post_content['media']
+        # Heuristic from your "old code": if 'title' is present, it's likely a Document.
+        if 'title' in media_item:
+            # Example: "content": {"media": {"title": "...", "id": "urn:li:document:..."}}
+            return "Document"
+        # Else, if 'id' is present (and no title was found for Document), assume Image.
+        elif 'id' in media_item:
+            # Example: "content": {"media": {"altText": "", "id": "urn:li:image:..."}}
+            return "Image"
+
+    return "NONE"
 # --- Post Retrieval Functions ---
 def fetch_linkedin_posts_core(comm_client_id, community_token, org_urn, count):
     """
@@ -266,9 +304,7 @@ def fetch_linkedin_posts_core(comm_client_id, community_token, org_urn, count):
             "published_at_iso": published_at_iso,
             "organization_urn": p.get("author", f"urn:li:organization:{org_urn.split(':')[-1]}"), 
             "is_ad": 'adContext' in p,
-            "media_category": p.get("content",{}).get("com.linkedin.voyager.feed.render.LinkedInVideoComponent",{}).get("mediaCategory") or \
-                              p.get("content",{}).get("com.linkedin.voyager.feed.render.ImageComponent",{}).get("mediaCategory") or \
-                              p.get("content",{}).get("com.linkedin.voyager.feed.render.ArticleComponent",{}).get("mediaCategory") or "NONE"
+            "media_category": get_post_media_category(p.get("content")),
         })
     logging.info(f"Processed {len(processed_raw_posts)} posts with core data.")
     return processed_raw_posts, stats_map, "DefaultOrgName"