From 1836081ef9855d632768987e1d11a05b9dcf9e9f Mon Sep 17 00:00:00 2001
From: Jolan Thomassin <98430140+JolanThomassin@users.noreply.github.com>
Date: Thu, 14 Mar 2024 18:42:07 +0000
Subject: [PATCH 1/2] Fixes #79, SQL query getting random crawl and score

---
 ailab/db/finesse/test_queries/__init__.py | 36 +++++++++++++++++++++++
 1 file changed, 36 insertions(+)

diff --git a/ailab/db/finesse/test_queries/__init__.py b/ailab/db/finesse/test_queries/__init__.py
index 6883e9a..cb5d916 100644
--- a/ailab/db/finesse/test_queries/__init__.py
+++ b/ailab/db/finesse/test_queries/__init__.py
@@ -44,3 +44,39 @@ def get_random_chunk(cursor, schema_version, seed=None):
 
     cursor.execute(query)
     return cursor.fetchall()
+
+
+def get_random_document_score(cursor, schema_version, seed=None):
+    if seed is None:
+        seed = math.sin(time.time())
+
+    # Execute the SET commands separately
+    cursor.execute(f'SET SEARCH_PATH TO "{schema_version}", public;')
+    cursor.execute(f"SET SEED TO {seed};")
+
+    query = """
+        WITH random_crawl AS (  
+            SELECT id  
+            FROM crawl  
+            ORDER BY
+                floor(random() * (
+                    SELECT
+                        COUNT(*)
+                    FROM
+                        Chunk
+                ))
+            LIMIT
+                1  
+        )  
+        SELECT  
+            cr.id AS crawl_id, cr.url AS crawl_url, sc.score, sc.score_type   
+        FROM  
+            crawl cr  
+        INNER JOIN  
+            score sc ON cr.id = sc.entity_id  
+        WHERE  
+            cr.id = (SELECT id FROM random_crawl)  
+    """
+
+    cursor.execute(query)
+    return cursor.fetchall()
\ No newline at end of file

From 1510b8d06a44389316b7f84932d550cd743a26f1 Mon Sep 17 00:00:00 2001
From: Jolan Thomassin <98430140+JolanThomassin@users.noreply.github.com>
Date: Thu, 14 Mar 2024 18:42:16 +0000
Subject: [PATCH 2/2] Fixes #79, New script printing random crawl

---
 bin/testing-current-score.py | 45 ++++++++++++++++++++++++++++++++++++
 bin/testing-current-score.sh |  6 +++++
 2 files changed, 51 insertions(+)
 create mode 100644 bin/testing-current-score.py
 create mode 100755 bin/testing-current-score.sh

diff --git a/bin/testing-current-score.py b/bin/testing-current-score.py
new file mode 100644
index 0000000..20f9cc5
--- /dev/null
+++ b/bin/testing-current-score.py
@@ -0,0 +1,45 @@
+import ailab.db as db
+
+from ailab.db.finesse.test_queries import get_random_document_score
+
+
+class NoChunkFoundError(Exception):
+    pass
+
+
+## This is a comment.
+def evaluate_random_document(project_db):  
+  
+    if project_db is None:  
+        print("Database connection failed.")  
+        return None  
+  
+    with project_db.cursor() as cursor:  
+  
+        random_chunk = get_random_document_score(cursor, "louis_v005")  
+  
+        if not random_chunk:  
+            raise NoChunkFoundError("No chunk found in the database.")  
+  
+        print("\n-------------")
+        print("crawl_id:", random_chunk[0]["crawl_id"])  
+        print("crawl_url:", random_chunk[0]["crawl_url"])
+        print("\n") 
+        
+        print("-------------")
+        for chunk in random_chunk:  
+            print("score_type:", chunk["score_type"])  
+            print("score:", chunk["score"])  
+            print("\n")  
+        
+    return  
+
+
+
+def main():
+    project_db = db.connect_db()
+    evaluate_random_document(project_db)
+
+
+if __name__ == "__main__":
+    main()
diff --git a/bin/testing-current-score.sh b/bin/testing-current-score.sh
new file mode 100755
index 0000000..47d1dfd
--- /dev/null
+++ b/bin/testing-current-score.sh
@@ -0,0 +1,6 @@
+#!/bin/bash
+DIRNAME=$(dirname "$0")
+. "$DIRNAME"/lib.sh
+
+
+PYTHONPATH=$PROJECT_DIR python "$DIRNAME"/testing-current-score.py