Add deduplication of source page views

models/base/docs.md

@@ -3,5 +3,6 @@
 {% docs segment_web_page_views %}
 
 This is a base model for Segment's web page views table. It does some straightforward renaming and parsing of Segment raw data in this table.
+If a page view id has multiple entries in the source table then deduplication is done to keep the row with the earliest `received_at` timestamp.
 
 {% enddocs %}

models/base/segment_web_page_views.sql

@@ -4,6 +4,24 @@ with source as (
 
 ),
 
+row_numbering as (
+
+    select
+        *,
+        row_number() over (partition by id order by received_at asc) as row_num
+    from source
+
+),
+
+deduped as (
+
+    select
+        *
+    from row_numbering
+    where row_num = 1
+
+),
+
 renamed as (
 
     select
@@ -50,7 +68,7 @@ renamed as (
 
         {% endif %}
 
-    from source
+    from deduped
 
 ),