File size: 6,945 Bytes
37ae1e2
 
 
 
 
 
 
 
 
 
 
 
 
 
ae03c09
37ae1e2
affaa54
 
37ae1e2
 
 
 
 
 
 
 
 
 
affaa54
 
ae03c09
 
37ae1e2
 
 
 
 
 
 
 
 
 
 
 
ae03c09
 
 
 
 
 
 
37ae1e2
 
 
 
 
 
 
 
 
affaa54
 
4df4d00
affaa54
 
 
 
 
 
 
 
 
37ae1e2
 
 
 
 
 
 
 
 
 
 
 
 
ae03c09
 
 
 
37ae1e2
 
 
 
ae03c09
 
 
409cbb4
 
ae03c09
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
37ae1e2
 
 
 
 
 
 
 
 
 
 
 
ae03c09
 
 
 
 
 
 
 
37ae1e2
 
 
 
 
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
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
#!/usr/bin/env python3
"""Validate bibliographic metadata in the generated resource dataset."""

from __future__ import annotations

import csv
import re
import sys
from datetime import datetime, timezone
from pathlib import Path


ROOT = Path(__file__).resolve().parents[1]
RESOURCES = ROOT / "data" / "resources.csv"
ARXIV_PUBLICATIONS = ROOT / "data" / "arxiv_publication_audit.csv"
YEAR_RE = re.compile(r"^(19|20)\d{2}$")
PROJECT_GITHUB_REPO = "chaoyue0307/awesome-loop-engineering"
PROJECT_PUBLICATION_SURFACES = {"GitHub", "GitHub Releases", "GitHub Discussions"}
REQUIRED_COLUMNS = {
    "authors",
    "publication_date",
    "publication_year",
    "publication_venue",
    "publisher",
    "doi",
    "publication_note",
    "primary_category",
    "metadata_source",
    "github_repo",
    "url_kind",
    "canonical_url",
    "evidence_class",
}


def main() -> int:
    with RESOURCES.open(encoding="utf-8", newline="") as handle:
        reader = csv.DictReader(handle)
        missing = REQUIRED_COLUMNS - set(reader.fieldnames or [])
        if missing:
            print(f"Missing publication columns: {', '.join(sorted(missing))}", file=sys.stderr)
            return 1
        rows = list(reader)

    if not ARXIV_PUBLICATIONS.exists():
        print("Missing data/arxiv_publication_audit.csv", file=sys.stderr)
        return 1
    with ARXIV_PUBLICATIONS.open(encoding="utf-8", newline="") as handle:
        publication_rows = list(csv.DictReader(handle))
    publications = {row["arxiv_id"]: row for row in publication_rows}

    failures: list[str] = []
    maximum_year = datetime.now(timezone.utc).year + 1
    for row in rows:
        row_id = row["row_id"]
        year = row["publication_year"]
        date = row["publication_date"]

        if not row["publisher"]:
            failures.append(f"{row_id}: publisher/source platform is missing")
        if row["publisher"].strip().lower() == "awesome loop engineering":
            failures.append(f"{row_id}: project name is exposed as a publisher")
        if row["url_kind"] != "external" and row["evidence_class"] == "repository-native" and (
            row["publication_venue"] != "GitHub" or row["publisher"] != "GitHub"
        ):
            failures.append(f"{row_id}: repository-native artifact must use GitHub as its publishing platform")
        if row["github_repo"].lower() == PROJECT_GITHUB_REPO and (
            row["publication_venue"] not in PROJECT_PUBLICATION_SURFACES or row["publisher"] != "GitHub"
        ):
            failures.append(f"{row_id}: project surface must identify GitHub, not the repository name, as its venue")
        if PROJECT_GITHUB_REPO in row["publication_venue"].lower():
            failures.append(f"{row_id}: project repository slug is exposed as a publication venue")
        if not row["metadata_source"]:
            failures.append(f"{row_id}: metadata provenance is missing")
        if year:
            if not YEAR_RE.fullmatch(year) or not 1900 <= int(year) <= maximum_year:
                failures.append(f"{row_id}: invalid publication year {year!r}")
            if date and not date.startswith(year):
                failures.append(f"{row_id}: publication date {date!r} disagrees with year {year!r}")
        if row["resource_type"] == "Paper":
            if not row["authors"]:
                failures.append(f"{row_id}: paper authors are missing")
            if not year:
                failures.append(f"{row_id}: paper publication year is missing")
        if row["arxiv_id"]:
            decision = publications.get(row["arxiv_id"])
            if not decision:
                failures.append(f"{row_id}: arXiv publication decision is missing")
                continue
            if not row["authors"]:
                failures.append(f"{row_id}: arXiv authors are missing")
            if not date:
                failures.append(f"{row_id}: arXiv publication date is missing")
            if decision["status"] == "preprint-only":
                if row["publisher"] != "arXiv":
                    failures.append(f"{row_id}: preprint-only row has publisher {row['publisher']!r}")
                if row["metadata_source"] not in {"arxiv-api", "arxiv-html-meta"}:
                    failures.append(f"{row_id}: preprint metadata is not backed by a primary arXiv source")
                expected_evidence = "benchmark" if row["resource_type"] == "Benchmark" else "research-preprint"
                if row["evidence_class"] != expected_evidence:
                    failures.append(f"{row_id}: preprint-only row has evidence class {row['evidence_class']!r}")
            else:
                expected = {
                    "publication_date": decision["publication_date"],
                    "publication_year": decision["publication_year"],
                    "publication_venue": decision["publication_venue"],
                    "publisher": decision["publisher"],
                    "doi": decision["doi"],
                    "canonical_url": decision["published_url"],
                    "metadata_source": decision["metadata_source"],
                }
                for field, expected_value in expected.items():
                    if row[field] != expected_value:
                        failures.append(
                            f"{row_id}: {field} differs from the verified publication record "
                            f"({row[field]!r} != {expected_value!r})"
                        )
                if row["publisher"] == "arXiv":
                    failures.append(f"{row_id}: resolved publication still uses arXiv as publisher")
                expected_evidence = "benchmark" if row["resource_type"] == "Benchmark" else "research-paper"
                if row["evidence_class"] != expected_evidence:
                    failures.append(f"{row_id}: resolved publication has evidence class {row['evidence_class']!r}")
        doi = row["doi"].lower()
        if doi.startswith(("http://", "https://", "doi:")):
            failures.append(f"{row_id}: DOI is not normalized")

    if failures:
        print("Publication metadata check failed:", file=sys.stderr)
        for failure in failures[:50]:
            print(f"- {failure}", file=sys.stderr)
        if len(failures) > 50:
            print(f"- ... and {len(failures) - 50} more", file=sys.stderr)
        return 1

    resource_arxiv_ids = {row["arxiv_id"] for row in rows if row["arxiv_id"]}
    if resource_arxiv_ids != set(publications):
        print("Publication metadata check failed:", file=sys.stderr)
        print("- arXiv decision IDs do not match exported resource IDs", file=sys.stderr)
        return 1

    resolved = sum(row["status"] in {"accepted", "published"} for row in publication_rows)
    print(f"Validated publication metadata for {len(rows)} resources ({resolved} arXiv venue upgrades).")
    return 0


if __name__ == "__main__":
    raise SystemExit(main())