Datasets:
File size: 5,055 Bytes
7bcd860 6d853a7 7bcd860 6d853a7 7bcd860 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 | """Validate the produced Parquet tables against the canonical schemas."""
from __future__ import annotations
import pyarrow as pa
import pyarrow.compute as pc
import pyarrow.parquet as pq
from common import ensure_dirs, get_logger, load_config, p, utc_now, write_json
from schema import (
AUTHORS_DIM, CITATION_EDGES, COUNTRY_COLLABORATION, COUNTRY_YEAR_COUNTS,
INSTITUTIONS_DIM, INSTITUTION_ACTIVITY, SOURCES_DIM, TOPIC_GROWTH,
TOPIC_YEAR_COUNTS, TOPICS_DIM, WORK_AUTHORS, WORK_INSTITUTIONS, WORK_TOPICS,
WORKS_FULL,
)
# rel -> (schema, required non-null columns, primary-key column or None)
EXPECTED = {
"data/works.parquet": (WORKS_FULL, ["work_id", "title", "publication_year"], "work_id"),
"data/work_authors.parquet": (WORK_AUTHORS, ["work_id"], None),
"data/work_institutions.parquet": (WORK_INSTITUTIONS, ["work_id", "institution_id"], None),
"data/work_topics.parquet": (WORK_TOPICS, ["work_id", "topic_id"], None),
"data/authors.parquet": (AUTHORS_DIM, ["author_id"], "author_id"),
"data/institutions.parquet": (INSTITUTIONS_DIM, ["institution_id"], "institution_id"),
"data/topics.parquet": (TOPICS_DIM, ["topic_id"], "topic_id"),
"data/sources.parquet": (SOURCES_DIM, ["source_id"], "source_id"),
"data/citation_edges.parquet": (CITATION_EDGES, ["citing_work_id", "cited_work_id"], None),
"analytics/topic_year_counts.parquet": (TOPIC_YEAR_COUNTS, ["topic_id", "publication_year"], None),
"analytics/topic_growth.parquet": (TOPIC_GROWTH, ["topic_id"], "topic_id"),
"analytics/country_year_counts.parquet": (COUNTRY_YEAR_COUNTS, ["country_code", "publication_year"], None),
"analytics/country_collaboration.parquet": (COUNTRY_COLLABORATION, ["country_a", "country_b"], None),
"analytics/institution_activity.parquet": (INSTITUTION_ACTIVITY, ["institution_id", "publication_year"], None),
}
def run(cfg: dict | None = None) -> dict:
cfg = cfg or load_config()
ensure_dirs(cfg)
logger = get_logger("diversa.validate", p(cfg, "logs") / "pipeline.log")
root = p(cfg, "data").parent
results = {"checked_at": utc_now(), "tables": {}, "errors": [], "warnings": []}
for rel, (schema, required, pk) in EXPECTED.items():
path = root / rel
entry = {"path": rel}
if not path.exists():
entry["status"] = "missing"
results["errors"].append(f"{rel} missing")
results["tables"][rel] = entry
continue
pf = pq.ParquetFile(path)
arrow = pf.schema_arrow
names = arrow.names
missing_cols = [c for c in schema.names if c not in names]
extra_cols = [c for c in names if c not in schema.names]
type_mismatch = []
for f in schema:
if f.name in names:
actual = arrow.field(f.name).type
if actual != f.type and str(actual).replace("element", "item") != str(f.type):
type_mismatch.append(f"{f.name}: {actual} != {f.type}")
entry.update({"rows": pf.metadata.num_rows, "columns": len(names),
"missing_columns": missing_cols, "extra_columns": extra_cols,
"type_mismatch": type_mismatch})
if missing_cols:
results["errors"].append(f"{rel} missing columns {missing_cols}")
if type_mismatch:
results["errors"].append(f"{rel} type mismatch {type_mismatch}")
if extra_cols:
results["warnings"].append(f"{rel} extra columns {extra_cols}")
cols = [c for c in required if c in names]
if cols:
tbl = pq.read_table(path, columns=cols)
for c in cols:
nn = tbl.column(c).null_count
if nn:
results["errors"].append(f"{rel}:{c} has {nn} nulls")
list_cols = [f.name for f in schema if pa.types.is_list(f.type) and f.name in names]
for c in list_cols:
flat = pc.list_flatten(pq.read_table(path, columns=[c]).column(c))
if flat.null_count:
results["errors"].append(
f"{rel}:{c} has {flat.null_count} null elements inside lists")
if pk and pk in names:
vals = pq.read_table(path, columns=[pk]).column(pk).to_pylist()
dup = len(vals) - len(set(vals))
entry["duplicate_pk"] = dup
if dup:
results["errors"].append(f"{rel} has {dup} duplicate {pk}")
entry["status"] = "ok" if not missing_cols and not type_mismatch else "schema_error"
results["tables"][rel] = entry
logger.info("validated %-45s rows=%-9s status=%s",
rel, f"{entry.get('rows', 0):,}", entry["status"])
results["status"] = "pass" if not results["errors"] else "fail"
write_json(p(cfg, "metadata") / "validation_report.json", results)
logger.info("validation %s (%d errors, %d warnings)",
results["status"], len(results["errors"]), len(results["warnings"]))
return results
if __name__ == "__main__":
run()
|