fixed reg_data using reg_history

pipelines.py

@@ -1222,7 +1222,7 @@ def process_reg_history():
             output1.write(f"  [{dt}] {msg}\n")
 
 def recreate_reg_data():
-    import csv
+
     from collections import defaultdict
     from datetime import datetime
 
@@ -1230,43 +1230,55 @@ def recreate_reg_data():
         dt = datetime.strptime(row['datetime'], "%Y-%m-%dT%H-%M")
         crn = row['crn']
         enrolled = int(row['enrolled'])
-        return dt, crn, enrolled
+        return dt, row['datetime'], crn, enrolled
 
     def reduce_latest_per_day(rows):
-        latest = defaultdict(dict)  # latest[crn][date_str] = (dt, enrolled)
+        latest = defaultdict(dict)  # latest[crn][date] = (dt, ts, enrolled)
+        latest_ts_by_date = {}      # date → (dt, ts) for header naming
 
         for row in rows:
-            dt, crn, enrolled = parse_row(row)
-            day = dt.date().isoformat()
-            if day not in latest[crn] or dt > latest[crn][day][0]:
-                latest[crn][day] = (dt, enrolled)
-        return latest
+            dt, full_ts, crn, enrolled = parse_row(row)
+            date_str = dt.date().isoformat()
+            ts_header = dt.strftime("%Y-%m-%dT%H")  # <-- this is what we want
 
-    def pivot_to_table(latest_data):
-        all_dates = sorted({day for crn in latest_data for day in latest_data[crn]})
-        crns = sorted(latest_data)
+            # for each crn, per day, keep latest reading
+            if date_str not in latest[crn] or dt > latest[crn][date_str][0]:
+                latest[crn][date_str] = (dt, ts_header, enrolled)
+
+            # also record latest timestamp per day for consistent column headers
+            if date_str not in latest_ts_by_date or dt > latest_ts_by_date[date_str][0]:
+                latest_ts_by_date[date_str] = (dt, ts_header)
+
+        return latest, [ts for _, ts in sorted(latest_ts_by_date.values())]
+
+    def pivot_table(latest, headers):
+        crns = sorted(latest)
         table = []
 
         for crn in crns:
             row = [crn]
-            for day in all_dates:
-                val = latest_data[crn].get(day, (None, None))[1]
-                row.append(str(val) if val is not None else "")
+            for ts in headers:
+                date_str = ts[:10]  # match on YYYY-MM-DD
+                val = latest[crn].get(date_str)
+                if val and val[1] == ts:
+                    row.append(str(val[2]))
+                else:
+                    row.append("")
             table.append(row)
 
-        return ["crn"] + all_dates, table
+        return ['crn'] + headers, table
 
     with open("cache/reg_history_fa25.csv", newline='') as f:
         fieldnames = ['datetime', 'crn', 'course', 'teacher', 'max', 'enrolled', 'waitlistmax', 'waitlisted']
         reader = csv.DictReader(f, fieldnames=fieldnames)
         rows = list(reader)
 
-    latest = reduce_latest_per_day(rows)
-    header, table = pivot_to_table(latest)
+    latest, headers = reduce_latest_per_day(rows)
+    header_row, table = pivot_table(latest, headers)
 
     with open("cache/reg_data_fa25.csv", "w", newline='') as f:
         writer = csv.writer(f)
-        writer.writerow(header)
+        writer.writerow(header_row)
         writer.writerows(table)