From 9b52c8cbac2ebece517b37fd38f1497cbea1951b Mon Sep 17 00:00:00 2001
From: kshitijrajsharma <skshitizraj@gmail.com>
Date: Tue, 17 Dec 2024 21:44:40 +0100
Subject: [PATCH] fix(folderstats): fixes folder stats not being populated

---
 API/download_metrics.py |  8 +++-
 src/app.py              | 81 ++++++++++++++++++++++++++++-------------
 2 files changed, 62 insertions(+), 27 deletions(-)

diff --git a/API/download_metrics.py b/API/download_metrics.py
index d33645d0..eba15bae 100644
--- a/API/download_metrics.py
+++ b/API/download_metrics.py
@@ -1,5 +1,6 @@
 # Standard library imports
 from datetime import datetime
+from typing import Optional
 
 # Third party imports
 from fastapi import APIRouter, Depends, HTTPException, Query
@@ -33,6 +34,11 @@ def get_stats(
         description="Group by: day, month, or quarter",
         regex=r"^(day|month|quarter|year)$",
     ),
+    folder: Optional[str] = Query(
+        None,
+        description="Folder to filter metrics by",
+        example="TM",
+    ),
     _: bool = Depends(staff_required),
 ):
     """
@@ -59,4 +65,4 @@ def get_stats(
         )
 
     metrics = DownloadMetrics()
-    return metrics.get_summary_stats(start_date, end_date, group_by)
+    return metrics.get_summary_stats(start_date, end_date, group_by, folder)
diff --git a/src/app.py b/src/app.py
index fb5e82d5..c9b2662e 100644
--- a/src/app.py
+++ b/src/app.py
@@ -29,7 +29,7 @@
 import sys
 import time
 import uuid
-from collections import namedtuple
+from collections import Counter, namedtuple
 from datetime import datetime, timedelta, timezone
 from json import dumps
 from json import loads as json_loads
@@ -2226,32 +2226,61 @@ def __init__(self) -> None:
         self.d_b = Database(dbdict)
         self.con, self.cur = self.d_b.connect()
 
-    def get_summary_stats(self, start_date, end_date, group_by):
-        """
-        Get summary metrics for raw-data-api downlaods
-        """
-
-        select_query = f"""
-            SELECT
-                date_trunc('{group_by}', date) as kwdate,
-                SUM((summary->>'downloads_count')::numeric) as total_downloads_count,
-                SUM((summary->>'uploads_count')::numeric) as total_uploads_count,
-                SUM((summary->>'unique_users')::numeric) as total_unique_users,
-                SUM((summary->>'unique_downloads')::numeric) as total_unique_downloads,
-                SUM((summary->>'interactions_count')::numeric) as total_interactions_count,
-                SUM((summary->>'upload_size')::numeric) as total_upload_size,
-                SUM((summary->>'download_size')::numeric) as total_download_size
-            FROM
-                metrics
-            WHERE
-                date BETWEEN '{start_date}' AND '{end_date}'
-            GROUP BY
-                kwdate
-            ORDER BY
-                kwdate
-        """
+    def get_summary_stats(self, start_date, end_date, group_by, folder=None):
+        """
+        Get summary metrics for raw-data-api downloads
+        """
+        if folder:
+            select_query = f"""
+                SELECT
+                    date_trunc('{group_by}', date) as kwdate,
+                    SUM((folders->'{folder}'->>'downloads_count')::numeric) as total_downloads_count,
+                    SUM((folders->'{folder}'->>'uploads_count')::numeric) as total_uploads_count,
+                    SUM((folders->'{folder}'->>'unique_users')::numeric) as total_unique_users,
+                    SUM((folders->'{folder}'->>'unique_downloads')::numeric) as total_unique_downloads,
+                    SUM((folders->'{folder}'->>'interactions_count')::numeric) as total_interactions_count,
+                    SUM((folders->'{folder}'->>'upload_size')::numeric) as total_upload_size,
+                    SUM((folders->'{folder}'->>'download_size')::numeric) as total_download_size,
+                    JSONB_AGG((folders->'{folder}'->>'locations')::json) as total_locations
+                FROM
+                    metrics
+                WHERE
+                    date BETWEEN '{start_date}' AND '{end_date}'
+                GROUP BY
+                    kwdate
+                ORDER BY
+                    kwdate
+            """
+        else:
+            select_query = f"""
+                SELECT
+                    date_trunc('{group_by}', date) as kwdate,
+                    SUM((summary->>'downloads_count')::numeric) as total_downloads_count,
+                    SUM((summary->>'uploads_count')::numeric) as total_uploads_count,
+                    SUM((summary->>'unique_users')::numeric) as total_unique_users,
+                    SUM((summary->>'unique_downloads')::numeric) as total_unique_downloads,
+                    SUM((summary->>'interactions_count')::numeric) as total_interactions_count,
+                    SUM((summary->>'upload_size')::numeric) as total_upload_size,
+                    SUM((summary->>'download_size')::numeric) as total_download_size,
+                    JSONB_AGG((summary->>'locations')::json) as total_locations
+                FROM
+                    metrics
+                WHERE
+                    date BETWEEN '{start_date}' AND '{end_date}'
+                GROUP BY
+                    kwdate
+                ORDER BY
+                    kwdate
+            """
 
         self.cur.execute(select_query)
         result = self.cur.fetchall()
         self.d_b.close_conn()
-        return [dict(item) for item in result]
+        result_lists = []
+        for item in result:
+            item["total_locations"] = dict(
+                sum((Counter(loc) for loc in item["total_locations"]), Counter())
+            )
+            result_lists.append(dict(item))
+
+        return result_lists