Datasets:

PureOne
/

EIROQ-PHI

#!/usr/bin/env python3
"""Check this local release. Optional loader checks use the local HF dataset card.
The default checks use only the standard library and make no network requests.
This verifies packaging and data access, not scientific correctness.
"""
from __future__ import annotations

import argparse
import csv
import hashlib
import json
import math
from pathlib import Path, PurePosixPath
import re
import sys
import tempfile

ROOT = Path(__file__).resolve().parents[1]


def verify_manifest(base: Path) -> int:
    entries = set()
    for number, line in enumerate((base / "SHA256SUMS.txt").read_text(encoding="utf-8").splitlines(), 1):
        if not line.strip():
            continue
        match = re.fullmatch(r"([0-9a-f]{64})  (.+)", line)
        if not match:
            raise ValueError(f"Malformed manifest entry {number}.")
        expected, name = match.groups()
        relative = PurePosixPath(name)
        if (relative.is_absolute() or "\\" in name or ":" in name
                or any(part in {"", ".", ".."} for part in name.split("/"))
                or name in entries or name == "SHA256SUMS.txt"):
            raise ValueError(f"Unsafe or duplicate manifest path: {name}")
        path = base / name
        if not path.resolve().is_relative_to(base.resolve()):
            raise ValueError(f"Path escapes release: {name}")
        cursor = path
        while cursor != base:
            if cursor.is_symlink():
                raise ValueError(f"Unexpected symlink: {name}")
            cursor = cursor.parent
        if not path.is_file() or hashlib.sha256(path.read_bytes()).hexdigest() != expected:
            raise ValueError(f"Missing file or checksum mismatch: {name}")
        entries.add(name)
    if not entries:
        raise ValueError("Empty integrity manifest.")
    return len(entries)


def validate_tables() -> tuple[dict, list]:
    index = json.loads((ROOT / "dataset_index.json").read_text(encoding="utf-8"))
    configs = index["configurations"]
    names = [entry["config_name"] for entry in configs]
    paths = [entry["path"] for entry in configs]
    if len(configs) != 11 or len(set(names)) != 11 or len(set(paths)) != 11:
        raise ValueError("Expected 11 distinct CSV configurations.")
    found = {p.relative_to(ROOT).as_posix() for p in (ROOT / "research/data").glob("*.csv")}
    if set(paths) != found:
        raise ValueError("Dataset index does not cover exactly the research CSV tables.")
    total = 0
    for entry in configs:
        if entry["split"] != "train":
            raise ValueError("Unexpected split mapping.")
        path = ROOT / entry["path"]
        with path.open(newline="", encoding="utf-8") as handle:
            reader = csv.DictReader(handle)
            if reader.fieldnames != entry["columns"]:
                raise ValueError(f"Column mismatch: {entry['config_name']}")
            rows = list(reader)
        if len(rows) != entry["rows"]:
            raise ValueError(f"Row count mismatch: {entry['config_name']}")
        for row in rows:
            if set(row) != set(entry["columns"]) or any(v is None for v in row.values()):
                raise ValueError(f"Malformed CSV row: {entry['config_name']}")
            if not all(math.isfinite(float(value)) for value in row.values()):
                raise ValueError(f"Non-finite numeric cell: {entry['config_name']}")
        total += len(rows)
    if total != index["total_rows"] or total != 532:
        raise ValueError("Unexpected total row count.")

    text = (ROOT / "README.md").read_text(encoding="utf-8")
    if not text.startswith("---\n") or "\n---\n" not in text[4:]:
        raise ValueError("Missing dataset-card front matter.")
    front = text.split("---", 2)[1]
    # This accepts the simple generated list shape; the optional HF check below
    # also parses the full YAML through DatasetCard and exercises its mappings.
    blocks = re.findall(r"^  - config_name: ([a-z_]+)\n(.*?)(?=^  - config_name:|\Z)", front, re.M | re.S)
    mappings = {}
    for name, block in blocks:
        files = re.findall(r"^        path: (\S+)\s*$", block, re.M)
        splits = re.findall(r"^      - split: (\S+)\s*$", block, re.M)
        if len(files) != 1 or splits != ["train"] or name in mappings:
            raise ValueError("Unexpected card configuration structure.")
        mappings[name] = files[0]
    if mappings != {entry["config_name"]: entry["path"] for entry in configs}:
        raise ValueError("Card mappings and dataset index differ.")
    if front.count("default: true") != 1 or not re.search(r"config_name: rank_law\n    default: true", front):
        raise ValueError("Expected rank_law as the unique default configuration.")
    if re.search(r"^license:", front, re.M):
        raise ValueError("Unexpected license grant in dataset card.")
    return {"configurations": len(configs), "total_rows": total, "numeric_csv_cells": "finite"}, configs


def validate_links_and_syntax() -> dict:
    checked = 0
    documents = [ROOT / "README.md", *(ROOT / "docs").glob("*.md")]
    for document in documents:
        for target in re.findall(r"!?\[[^\]]*\]\(([^)]+)\)", document.read_text(encoding="utf-8")):
            target = target.split("#", 1)[0]
            if not target or re.match(r"^[a-zA-Z][a-zA-Z0-9+.-]*:", target):
                continue
            path = document.parent / target
            if not path.resolve().is_relative_to(ROOT) or not path.exists():
                raise ValueError(f"Broken local link in {document.name}: {target}")
            checked += 1
    python_files = [ROOT / "upload_to_hf.py", ROOT / "tools/validate_release.py"]
    for path in python_files:
        compile(path.read_text(encoding="utf-8"), path.name, "exec")
    return {"local_links_checked": checked, "python_helpers_compiled": len(python_files)}


def check_datasets(configs: list) -> dict:
    import datasets
    import huggingface_hub
    from datasets import load_dataset
    from huggingface_hub import DatasetCard

    card = DatasetCard((ROOT / "README.md").read_text(encoding="utf-8"))
    declared = card.data.to_dict()["configs"]
    expected = {entry["config_name"]: entry["path"] for entry in configs}
    actual = {entry["config_name"]: entry["data_files"][0]["path"] for entry in declared}
    if expected != actual:
        raise ValueError("HF card parser found unexpected mappings.")
    checks = []
    with tempfile.TemporaryDirectory(prefix="eiroq-local-datasets-") as cache:
        for entry in configs:
            data = load_dataset(str(ROOT), name=entry["config_name"], split="train", cache_dir=cache)
            if len(data) != entry["rows"] or data.column_names != entry["columns"]:
                raise ValueError(f"Local HF loader mismatch: {entry['config_name']}")
            checks.append({"configuration": entry["config_name"], "rows": len(data)})
    return {"status": "passed", "datasets_version": datasets.__version__,
            "huggingface_hub_version": huggingface_hub.__version__, "checks": checks,
            "scope": "local card parsing and CSV loading; live Hub viewer not tested"}


def main() -> int:
    parser = argparse.ArgumentParser(description=__doc__)
    parser.add_argument("--check-datasets", action="store_true", help="Load all 11 local configs using datasets.")
    args = parser.parse_args()
    try:
        result = {"release": "EIROQ-PHI Hugging Face preparation 1.0.0-hf.1",
                  "outer_manifest_entries_verified": verify_manifest(ROOT),
                  "original_manifest_entries_verified": verify_manifest(ROOT / "research")}
        result["tables"], configs = validate_tables()
        result["links_and_syntax"] = validate_links_and_syntax()
        result["hf_loader"] = check_datasets(configs) if args.check_datasets else {"status": "not requested"}
        result["passed"] = True
        result["limitations"] = ["No upload performed", "No new laboratory measurements",
                                  "Original scientific test totals are preserved recorded evidence, not newly rerun here"]
        print(json.dumps(result, indent=2))
        return 0
    except Exception as exc:
        print(f"Validation failed ({type(exc).__name__}): {exc}", file=sys.stderr)
        return 1


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

	#!/usr/bin/env python3
	"""Check this local release. Optional loader checks use the local HF dataset card.

	The default checks use only the standard library and make no network requests.
	This verifies packaging and data access, not scientific correctness.
	"""
	from __future__ import annotations

	import argparse
	import csv
	import hashlib
	import json
	import math
	from pathlib import Path, PurePosixPath
	import re
	import sys
	import tempfile

	ROOT = Path(__file__).resolve().parents[1]


	def verify_manifest(base: Path) -> int:
	entries = set()
	for number, line in enumerate((base / "SHA256SUMS.txt").read_text(encoding="utf-8").splitlines(), 1):
	if not line.strip():
	continue
	match = re.fullmatch(r"([0-9a-f]{64}) (.+)", line)
	if not match:
	raise ValueError(f"Malformed manifest entry {number}.")
	expected, name = match.groups()
	relative = PurePosixPath(name)
	if (relative.is_absolute() or "\\" in name or ":" in name
	or any(part in {"", ".", ".."} for part in name.split("/"))
	or name in entries or name == "SHA256SUMS.txt"):
	raise ValueError(f"Unsafe or duplicate manifest path: {name}")
	path = base / name
	if not path.resolve().is_relative_to(base.resolve()):
	raise ValueError(f"Path escapes release: {name}")
	cursor = path
	while cursor != base:
	if cursor.is_symlink():
	raise ValueError(f"Unexpected symlink: {name}")
	cursor = cursor.parent
	if not path.is_file() or hashlib.sha256(path.read_bytes()).hexdigest() != expected:
	raise ValueError(f"Missing file or checksum mismatch: {name}")
	entries.add(name)
	if not entries:
	raise ValueError("Empty integrity manifest.")
	return len(entries)


	def validate_tables() -> tuple[dict, list]:
	index = json.loads((ROOT / "dataset_index.json").read_text(encoding="utf-8"))
	configs = index["configurations"]
	names = [entry["config_name"] for entry in configs]
	paths = [entry["path"] for entry in configs]
	if len(configs) != 11 or len(set(names)) != 11 or len(set(paths)) != 11:
	raise ValueError("Expected 11 distinct CSV configurations.")
	found = {p.relative_to(ROOT).as_posix() for p in (ROOT / "research/data").glob("*.csv")}
	if set(paths) != found:
	raise ValueError("Dataset index does not cover exactly the research CSV tables.")
	total = 0
	for entry in configs:
	if entry["split"] != "train":
	raise ValueError("Unexpected split mapping.")
	path = ROOT / entry["path"]
	with path.open(newline="", encoding="utf-8") as handle:
	reader = csv.DictReader(handle)
	if reader.fieldnames != entry["columns"]:
	raise ValueError(f"Column mismatch: {entry['config_name']}")
	rows = list(reader)
	if len(rows) != entry["rows"]:
	raise ValueError(f"Row count mismatch: {entry['config_name']}")
	for row in rows:
	if set(row) != set(entry["columns"]) or any(v is None for v in row.values()):
	raise ValueError(f"Malformed CSV row: {entry['config_name']}")
	if not all(math.isfinite(float(value)) for value in row.values()):
	raise ValueError(f"Non-finite numeric cell: {entry['config_name']}")
	total += len(rows)
	if total != index["total_rows"] or total != 532:
	raise ValueError("Unexpected total row count.")

	text = (ROOT / "README.md").read_text(encoding="utf-8")
	if not text.startswith("---\n") or "\n---\n" not in text[4:]:
	raise ValueError("Missing dataset-card front matter.")
	front = text.split("---", 2)[1]
	# This accepts the simple generated list shape; the optional HF check below
	# also parses the full YAML through DatasetCard and exercises its mappings.
	blocks = re.findall(r"^ - config_name: ([a-z_]+)\n(.*?)(?=^ - config_name:\|\Z)", front, re.M \| re.S)
	mappings = {}
	for name, block in blocks:
	files = re.findall(r"^ path: (\S+)\s*$", block, re.M)
	splits = re.findall(r"^ - split: (\S+)\s*$", block, re.M)
	if len(files) != 1 or splits != ["train"] or name in mappings:
	raise ValueError("Unexpected card configuration structure.")
	mappings[name] = files[0]
	if mappings != {entry["config_name"]: entry["path"] for entry in configs}:
	raise ValueError("Card mappings and dataset index differ.")
	if front.count("default: true") != 1 or not re.search(r"config_name: rank_law\n default: true", front):
	raise ValueError("Expected rank_law as the unique default configuration.")
	if re.search(r"^license:", front, re.M):
	raise ValueError("Unexpected license grant in dataset card.")
	return {"configurations": len(configs), "total_rows": total, "numeric_csv_cells": "finite"}, configs


	def validate_links_and_syntax() -> dict:
	checked = 0
	documents = [ROOT / "README.md", (ROOT / "docs").glob(".md")]
	for document in documents:
	for target in re.findall(r"!?\[[^\]]*\]\(([^)]+)\)", document.read_text(encoding="utf-8")):
	target = target.split("#", 1)[0]
	if not target or re.match(r"^[a-zA-Z][a-zA-Z0-9+.-]*:", target):
	continue
	path = document.parent / target
	if not path.resolve().is_relative_to(ROOT) or not path.exists():
	raise ValueError(f"Broken local link in {document.name}: {target}")
	checked += 1
	python_files = [ROOT / "upload_to_hf.py", ROOT / "tools/validate_release.py"]
	for path in python_files:
	compile(path.read_text(encoding="utf-8"), path.name, "exec")
	return {"local_links_checked": checked, "python_helpers_compiled": len(python_files)}


	def check_datasets(configs: list) -> dict:
	import datasets
	import huggingface_hub
	from datasets import load_dataset
	from huggingface_hub import DatasetCard

	card = DatasetCard((ROOT / "README.md").read_text(encoding="utf-8"))
	declared = card.data.to_dict()["configs"]
	expected = {entry["config_name"]: entry["path"] for entry in configs}
	actual = {entry["config_name"]: entry["data_files"][0]["path"] for entry in declared}
	if expected != actual:
	raise ValueError("HF card parser found unexpected mappings.")
	checks = []
	with tempfile.TemporaryDirectory(prefix="eiroq-local-datasets-") as cache:
	for entry in configs:
	data = load_dataset(str(ROOT), name=entry["config_name"], split="train", cache_dir=cache)
	if len(data) != entry["rows"] or data.column_names != entry["columns"]:
	raise ValueError(f"Local HF loader mismatch: {entry['config_name']}")
	checks.append({"configuration": entry["config_name"], "rows": len(data)})
	return {"status": "passed", "datasets_version": datasets.__version__,
	"huggingface_hub_version": huggingface_hub.__version__, "checks": checks,
	"scope": "local card parsing and CSV loading; live Hub viewer not tested"}


	def main() -> int:
	parser = argparse.ArgumentParser(description=__doc__)
	parser.add_argument("--check-datasets", action="store_true", help="Load all 11 local configs using datasets.")
	args = parser.parse_args()
	try:
	result = {"release": "EIROQ-PHI Hugging Face preparation 1.0.0-hf.1",
	"outer_manifest_entries_verified": verify_manifest(ROOT),
	"original_manifest_entries_verified": verify_manifest(ROOT / "research")}
	result["tables"], configs = validate_tables()
	result["links_and_syntax"] = validate_links_and_syntax()
	result["hf_loader"] = check_datasets(configs) if args.check_datasets else {"status": "not requested"}
	result["passed"] = True
	result["limitations"] = ["No upload performed", "No new laboratory measurements",
	"Original scientific test totals are preserved recorded evidence, not newly rerun here"]
	print(json.dumps(result, indent=2))
	return 0
	except Exception as exc:
	print(f"Validation failed ({type(exc).__name__}): {exc}", file=sys.stderr)
	return 1


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