quill: add csv_summary.py artifact with sample and verifier

wizard-work/quill/csv_summary.py

@@ -0,0 +1,51 @@
+#!/usr/bin/env python3
+"""csv_summary.py — Read a CSV from stdin or a file argument and emit JSON summary statistics."""
+
+import sys
+import json
+import csv
+from collections import defaultdict
+from statistics import mean, median
+
+def summarize_csv(path=None):
+    fh = open(path, newline='') if path else sys.stdin
+    reader = csv.DictReader(fh)
+    rows = list(reader)
+    fh.close() if path else None
+
+    if not rows:
+        return {"rows": 0, "columns": [], "numeric": {}}
+
+    columns = reader.fieldnames or []
+    numeric = {}
+    for col in columns:
+        values = []
+        for row in rows:
+            try:
+                values.append(float(row[col]))
+            except (ValueError, TypeError):
+                pass
+        if values:
+            numeric[col] = {
+                "count": len(values),
+                "sum": round(sum(values), 4),
+                "mean": round(mean(values), 4),
+                "median": round(median(values), 4),
+                "min": round(min(values), 4),
+                "max": round(max(values), 4),
+            }
+
+    return {
+        "rows": len(rows),
+        "columns": columns,
+        "numeric": numeric,
+    }
+
+def main():
+    path = sys.argv[1] if len(sys.argv) > 1 else None
+    result = summarize_csv(path)
+    json.dump(result, sys.stdout, indent=2)
+    sys.stdout.write("\n")
+
+if __name__ == "__main__":
+    main()

wizard-work/quill/sample.csv

@@ -0,0 +1,6 @@
+name,age,salary,active
+Alice,30,50000,true
+Bob,25,45000,false
+Charlie,35,70000,true
+Diana,28,52000,true
+Eve,32,48000,false

wizard-work/quill/verify_summary.py

@@ -0,0 +1,34 @@
+#!/usr/bin/env python3
+"""verify_summary.py — Assert csv_summary.py output matches expected values."""
+import sys
+import json
+
+expected_rows = 5
+expected_columns = ["name", "age", "salary", "active"]
+expected_age = {
+    "count": 5,
+    "sum": 150.0,
+    "mean": 30.0,
+    "median": 30.0,
+    "min": 25.0,
+    "max": 35.0,
+}
+expected_salary = {
+    "count": 5,
+    "sum": 265000.0,
+    "mean": 53000.0,
+    "median": 50000.0,
+    "min": 45000.0,
+    "max": 70000.0,
+}
+
+data = json.load(sys.stdin)
+assert data["rows"] == expected_rows, f"rows: {data['rows']}"
+assert data["columns"] == expected_columns, f"columns: {data['columns']}"
+age = data["numeric"]["age"]
+for k, v in expected_age.items():
+    assert age[k] == v, f"age.{k}: {age[k]} != {v}"
+sal = data["numeric"]["salary"]
+for k, v in expected_salary.items():
+    assert sal[k] == v, f"salary.{k}: {sal[k]} != {v}"
+print("VERIFIED")