Spaces:
Running on Zero
Running on Zero
File size: 8,957 Bytes
9936912 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 | #!/usr/bin/env python3
"""Download ranked open-book candidates with Scrapy and verify each PDF."""
from __future__ import annotations
import argparse
import hashlib
import json
import re
import unicodedata
from datetime import datetime, timezone
from pathlib import Path
import pymupdf
import scrapy
from scrapy.crawler import CrawlerProcess
PROJECT_ROOT = Path(__file__).resolve().parents[1]
DEFAULT_INPUT = PROJECT_ROOT / "data" / "sources" / "discovered_open_books.jsonl"
DEFAULT_OUTPUT = PROJECT_ROOT / "data" / "raw" / "open_books"
DEFAULT_SUMMARY = PROJECT_ROOT / "data" / "sources" / "downloaded_open_books_summary.json"
def sha256_bytes(data: bytes) -> str:
return hashlib.sha256(data).hexdigest()
def slug(text: str, limit: int = 80) -> str:
text = unicodedata.normalize("NFKD", text).encode("ascii", "ignore").decode()
text = re.sub(r"[^a-zA-Z0-9]+", "_", text).strip("_").lower()
return (text[:limit].rstrip("_") or "untitled")
def source_id(candidate: dict) -> str:
identity = candidate.get("oai_identifier") or candidate["pdf_url"]
return f"{candidate['source_catalog']}_{hashlib.sha256(identity.encode()).hexdigest()[:16]}"
def filename(candidate: dict) -> str:
url_hash = hashlib.sha256(candidate["pdf_url"].encode()).hexdigest()[:10]
return f"{slug(candidate['title'])}_{url_hash}.pdf"
def load_candidates(path: Path, min_score: int, max_books: int) -> list[dict]:
rows = [json.loads(line) for line in path.open(encoding="utf-8")]
rows = [row for row in rows if row["relevance_score"] >= min_score]
rows.sort(key=lambda row: (-row["relevance_score"], row["title"].lower(), row["pdf_url"]))
return rows[:max_books] if max_books else rows
class OpenBookDownloadSpider(scrapy.Spider):
name = "open_book_downloads"
custom_settings = {
"ROBOTSTXT_OBEY": True,
"CONCURRENT_REQUESTS": 2,
"CONCURRENT_REQUESTS_PER_DOMAIN": 2,
"DOWNLOAD_DELAY": 0.5,
"AUTOTHROTTLE_ENABLED": True,
"AUTOTHROTTLE_START_DELAY": 0.5,
"AUTOTHROTTLE_MAX_DELAY": 10.0,
"RETRY_TIMES": 4,
"DOWNLOAD_TIMEOUT": 180,
"USER_AGENT": "controlai-open-corpus-research/0.1",
"LOG_LEVEL": "INFO",
}
def __init__(
self,
candidates: list[dict],
output_dir: str,
summary_path: str,
max_mib: int,
*args,
**kwargs,
) -> None:
super().__init__(*args, **kwargs)
self.candidates = candidates
self.output_dir = Path(output_dir)
self.summary_path = Path(summary_path)
self.max_bytes = int(max_mib) * 1024 * 1024
self.lock_path = self.output_dir / "source_lock.json"
self.output_dir.mkdir(parents=True, exist_ok=True)
self.records = []
if self.lock_path.exists():
self.records = json.loads(self.lock_path.read_text(encoding="utf-8"))
self.known_urls = {record["url"] for record in self.records}
self.known_hashes = {record["sha256"] for record in self.records}
self.downloaded = 0
self.skipped = 0
self.duplicates = 0
self.failures = []
async def start(self):
for candidate in self.candidates:
if candidate["pdf_url"] in self.known_urls:
self.skipped += 1
continue
yield scrapy.Request(
candidate["pdf_url"],
callback=self.save_pdf,
errback=self.download_failed,
cb_kwargs={"candidate": candidate},
meta={"download_maxsize": self.max_bytes, "download_warnsize": self.max_bytes},
dont_filter=True,
)
def save_pdf(self, response: scrapy.http.Response, candidate: dict):
data = response.body
if not data.startswith(b"%PDF"):
self.failures.append(
{
"title": candidate["title"],
"url": candidate["pdf_url"],
"error": f"not a PDF (content-type={response.headers.get('Content-Type')!r})",
}
)
return
digest = sha256_bytes(data)
if digest in self.known_hashes:
self.duplicates += 1
return
try:
with pymupdf.open(stream=data, filetype="pdf") as document:
page_count = document.page_count
if page_count < 1:
raise ValueError("PDF has no pages")
except Exception as error:
self.failures.append(
{
"title": candidate["title"],
"url": candidate["pdf_url"],
"error": f"invalid PDF: {type(error).__name__}: {error}",
}
)
return
name = filename(candidate)
destination = self.output_dir / name
partial = destination.with_suffix(".pdf.part")
partial.write_bytes(data)
partial.replace(destination)
record = {
"source_id": source_id(candidate),
"title": candidate["title"],
"authors": candidate.get("creators", []) or candidate.get("contributors", []),
"filename": name,
"url": candidate["pdf_url"],
"catalog": candidate["source_catalog"],
"oai_identifier": candidate.get("oai_identifier"),
"publishers": candidate.get("publishers", []),
"license": candidate.get("licenses", []),
"corpus_tier": "supplemental_open_book",
"coverage": candidate.get("matched_phrases", []),
"relevance_score": candidate["relevance_score"],
"bytes": len(data),
"page_count": page_count,
"sha256": digest,
"downloaded_at": datetime.now(timezone.utc).isoformat(),
}
self.records.append(record)
self.known_urls.add(record["url"])
self.known_hashes.add(digest)
self.downloaded += 1
yield record
def download_failed(self, failure):
candidate = failure.request.cb_kwargs["candidate"]
self.failures.append(
{
"title": candidate["title"],
"url": candidate["pdf_url"],
"error": failure.getErrorMessage(),
}
)
def closed(self, reason: str) -> None:
self.records.sort(key=lambda record: (record["title"].lower(), record["url"]))
self.lock_path.write_text(
json.dumps(self.records, indent=2, ensure_ascii=False) + "\n", encoding="utf-8"
)
summary = {
"selected_candidates": len(self.candidates),
"downloaded_this_run": self.downloaded,
"already_downloaded": self.skipped,
"exact_content_duplicates": self.duplicates,
"failures": self.failures,
"total_downloaded_books": len(self.records),
"total_pages": sum(record["page_count"] for record in self.records),
"total_bytes": sum(record["bytes"] for record in self.records),
"close_reason": reason,
"generated_at": datetime.now(timezone.utc).isoformat(),
}
self.summary_path.parent.mkdir(parents=True, exist_ok=True)
self.summary_path.write_text(
json.dumps(summary, indent=2, ensure_ascii=False) + "\n", encoding="utf-8"
)
def main() -> None:
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument("--input", type=Path, default=DEFAULT_INPUT)
parser.add_argument("--output", type=Path, default=DEFAULT_OUTPUT)
parser.add_argument("--summary", type=Path, default=DEFAULT_SUMMARY)
parser.add_argument("--min-score", type=int, default=20)
parser.add_argument("--max-books", type=int, default=25, help="0 downloads every selected candidate")
parser.add_argument("--max-mib", type=int, default=150, help="Reject a single PDF larger than this")
parser.add_argument("--dry-run", action="store_true")
args = parser.parse_args()
candidates = load_candidates(args.input, args.min_score, args.max_books)
print(f"Selected {len(candidates)} candidates with score >= {args.min_score}")
for candidate in candidates[:20]:
print(f"- {candidate['relevance_score']:3d} {candidate['title']}")
if len(candidates) > 20:
print(f"... and {len(candidates) - 20} more")
if args.dry_run:
print("Dry run only. Remove --dry-run to download and verify PDFs.")
return
process = CrawlerProcess()
process.crawl(
OpenBookDownloadSpider,
candidates=candidates,
output_dir=str(args.output),
summary_path=str(args.summary),
max_mib=args.max_mib,
)
process.start()
if args.summary.exists():
print(args.summary.read_text(encoding="utf-8"))
print(f"PDFs: {args.output}")
if __name__ == "__main__":
main()
|