File size: 2,655 Bytes
e00f999 | 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 | #!/usr/bin/env python3
"""Check the dataset's files against their documented shape. Exit 1 on the first kind of problem found.
python3 scripts/validate.py
"""
import csv
import sys
from common import ROLES, data, read_jsonl
problems = []
def check(cond, msg):
if not cond:
problems.append(msg)
labels = read_jsonl(data("roles", "labels.jsonl"))
ids = set()
projects = {r["project"]: r for r in csv.DictReader(open(data("projects.csv"), encoding="utf-8"))}
for l in labels:
for k in ("id", "project", "commit", "module", "path", "url", "qualified_name", "name", "kind", "role", "structure"):
check(k in l, f"{l.get('id')}: missing {k}")
check(l["id"] not in ids, f"{l['id']}: duplicate id")
ids.add(l["id"])
check(l["id"] == f"{l['project']}/{l['qualified_name']}", f"{l['id']}: id is not project/qualified_name")
check(l["role"] in ROLES, f"{l['id']}: unknown role {l['role']}")
check(l["project"] in projects, f"{l['id']}: project not in projects.csv")
check(l["commit"] == projects.get(l["project"], {}).get("commit"), f"{l['id']}: commit differs from projects.csv")
check(l["path"].endswith(".java") and "src/main/java/" in l["path"], f"{l['id']}: path is not main Java")
check(l["kind"] in ("class", "abstract class", "interface", "enum", "record", "annotation"), f"{l['id']}: kind {l['kind']}")
check(set(l["structure"]["derived_marks"]) <= {"api", "ext"}, f"{l['id']}: unknown mark")
for p, r in projects.items():
n = sum(1 for l in labels if l["project"] == p)
check(n == int(r["labelled"]), f"{p}: {n} labels, projects.csv says {r['labelled']}")
csv_rows = list(csv.DictReader(open(data("roles", "labels.csv"), encoding="utf-8")))
check(len(csv_rows) == len(labels), "labels.csv and labels.jsonl differ in length")
features = read_jsonl(data("features", "features.jsonl"))
check({f["id"] for f in features} == ids, "features.jsonl does not cover exactly the labelled types")
try:
commits = read_jsonl(data("history", "commits.jsonl"))
except FileNotFoundError:
commits = []
problems.append("data/history/commits.jsonl missing")
for c in commits:
modified = {m["qualified_name"] for m in c["modified"]}
for k, shown in c["shown_before"].items():
check(set(shown) <= modified, f"{c['project']} {c['commit'][:10]}: {k} shows a type not modified")
check(c["project"] in projects, f"{c['commit']}: unknown project")
if problems:
print("\n".join(problems[:50]))
print(f"{len(problems)} problem(s)")
sys.exit(1)
print(f"ok: {len(labels)} labels, {len(features)} feature rows, {len(commits)} commits, {len(projects)} projects")
|