In the structured data model:
Posts are objects on a topic-page and are right now identified by their id-tag. Then the pages canonical-url is somehow used to create an identifier – also see:
- “identifier” property - Background notes - Data Model - schema.org.
- identifier - Schema.org Property
identifier:
https://meta.discourse.org/t/feedback-on-new-hamburger-and-user-menus/post_123
The identifier might be of type URL (IRI for JSON-LD), like it is in this case. This (automatically created) url does not need to match the canonical url of the post itself, it only needs to be unique.
Setting the identifier to the specific value of the canonical url ensures it is unique and should not do any harm.
And one step further: add the canonical url as property url to the post..
Crawler-view: topic page with posts
Current state:
<div id="post_123"
itemscope="" itemtype="http://schema.org/DiscussionForumPosting"
class="topic-body crawler-post">
<div class="crawler-post-meta">
[…]
Using canonical-url as identifier via attribute itemid and setting property url…
<div id="post_123"
itemid="https://meta.discourse.org/t/feedback-on-new-hamburger-and-user-menus/32519?page=6#post_123"
itemscope="" itemtype="http://schema.org/DiscussionForumPosting"
class="topic-body crawler-post">
<div class="crawler-post-meta">
<link itemprop="url" href="https://meta.discourse.org/t/feedback-on-new-hamburger-and-user-menus/32519?page=6#post_121">
[…]
… results in: https://search.google.com/test/rich-results/result?id=_k2xr2Uf2tCLTReR69-SLg&hl=en
