232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331 | def run_quality_rules(df: Any, rules: list[dict], *, dataset_name: str = "unknown", table_name: str = "unknown", engine: str = "auto") -> dict:
"""Execute quality rules against a dataframe and return structured results.
Parameters
----------
df : Any
Value used by this callable.
rules : Any
Value used by this callable.
Returns
-------
dict
Structured output produced by this callable.
"""
resolved_engine = _resolve_engine(df, engine)
row_count = len(df) if resolved_engine == "pandas" else df.count()
results = []
for idx, rule in enumerate(rules):
severity, sev_msg = _normalize_severity(rule)
rule = dict(rule)
rule_id = rule.get("rule_id", f"DQ{idx + 1:03d}")
rule_type = rule.get("rule_type")
base = {
"rule_id": rule_id,
"rule_type": rule_type or "unknown",
"column_name": rule.get("column"),
"columns": rule.get("columns"),
"severity": severity,
"status": "failed",
"action": SEVERITY_TO_ACTION[severity],
"failed_count": 1,
"total_count": int(row_count),
"failed_pct": 100.0 if row_count else 0.0,
"threshold": None,
"message": "Invalid rule",
"reason": rule.get("reason"),
}
if not rule_type:
base["message"] = "Missing required key: rule_type"
results.append(base)
continue
if rule_type not in SUPPORTED_RULE_TYPES:
base.update({"status": "skipped", "action": "allow", "failed_count": 0, "failed_pct": 0.0, "message": "Unsupported rule type"})
results.append(base)
continue
required = {
"not_null": ["column"],
"unique": ["column"],
"unique_combination": ["columns"],
"accepted_values": ["column", "accepted_values"],
"range_check": ["column"],
"regex_check": ["column", "pattern"],
"row_count_min": ["min_count"],
"row_count_between": ["min_count", "max_count"],
"freshness_check": ["column", "max_age_days"],
}[rule_type]
missing_keys = [k for k in required if k not in rule]
if missing_keys:
base["message"] = f"Missing required keys: {', '.join(missing_keys)}"
results.append(base)
continue
cols_to_check = rule.get("columns") or ([rule["column"]] if "column" in rule else [])
missing_cols = [c for c in cols_to_check if c not in getattr(df, "columns", [])]
if missing_cols:
base["message"] = f"Missing required column(s): {', '.join(missing_cols)}"
results.append(base)
continue
try:
failed_count, total_count, threshold, msg = (_pandas_rule(df, rule, row_count) if resolved_engine == "pandas" else _spark_rule(df, rule, row_count))
except ValueError as exc:
base["message"] = str(exc)
results.append(base)
continue
if sev_msg:
msg = f"{msg}. {sev_msg}"
results.append(_result_from_counts(rule, severity, failed_count, total_count, threshold, msg))
blocking = any(r["status"] == "failed" and r["action"] == "block" for r in results)
warning = any(r["status"] == "failed" and r["action"] == "warn" for r in results)
status = "failed" if blocking else "warning" if warning else "passed"
summary = {
"total_rules": len(results),
"passed_rules": sum(1 for r in results if r["status"] == "passed"),
"failed_rules": sum(1 for r in results if r["status"] == "failed"),
"skipped_rules": sum(1 for r in results if r["status"] == "skipped"),
}
return {
"dataset_name": dataset_name,
"table_name": table_name,
"engine": resolved_engine,
"status": status,
"can_continue": not blocking,
"results": _to_jsonable(results),
"summary": _to_jsonable(summary),
"generated_at": _now_iso(),
}
|