From 6db1239e49c51ca7869540b4d398e4a8abfb05d0 Mon Sep 17 00:00:00 2001
From: Rory Sawyer <rory@sawyer.dev>
Date: Thu, 19 Oct 2023 12:52:16 -0400
Subject: [PATCH] feat: support multi-part problem responses

---
 macros/get_problem_id.sql                  |  2 +-
 models/problems/fact_problem_responses.sql | 15 +++++++++++++++
 2 files changed, 16 insertions(+), 1 deletion(-)

diff --git a/macros/get_problem_id.sql b/macros/get_problem_id.sql
index bebd0158..362a2253 100644
--- a/macros/get_problem_id.sql
+++ b/macros/get_problem_id.sql
@@ -3,5 +3,5 @@
 -- either the problem id is at the end of the object ID
 -- or is followed by '/answer' or '/hint'
 {% macro get_problem_id(object_id) %}
-   regexpExtract(object_id, 'xblock/([\w\d-@\+:]*)', 1)
+   regexpExtract(object_id, 'xblock/([\w\d-\+:]*@problem\+block@[\w\d][^_]*)(_\d_\d)?', 1)
 {% endmacro %}
diff --git a/models/problems/fact_problem_responses.sql b/models/problems/fact_problem_responses.sql
index 89d347a0..832d71dc 100644
--- a/models/problems/fact_problem_responses.sql
+++ b/models/problems/fact_problem_responses.sql
@@ -32,3 +32,18 @@ from
     join {{ ref('dim_course_blocks')}} blocks
          on (responses.course_key = blocks.course_key
              and responses.problem_id = blocks.block_id)
+group by
+    -- multi-part questions include an extra record for the response to the first
+    -- part of the question. this group by clause eliminates the duplicate record
+    emission_time,
+    org,
+    course_key,
+    course_name,
+    course_run,
+    problem_id,
+    problem_name,
+    problem_name_with_location,
+    actor_id,
+    responses,
+    success,
+    attempts