canho's picture
Add files using upload-large-folder tool
667c0a2 verified
Raw History Blame Contribute Delete
1.8 kB
"""Download an open URL to disk with provenance (sha256, type, status).
Source-agnostic: an open-access publisher PDF or any directly fetchable URL.
arXiv links are handled upstream by arxiv_lib (higher fidelity), so they do not
normally reach here.
"""
from __future__ import annotations
import hashlib
from pathlib import Path
import requests
BROWSER_UA = {
"User-Agent": "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 "
"(KHTML, like Gecko) Chrome/124.0 Safari/537.36",
"Accept": "application/pdf,text/html,*/*",
}
def _ext_for(content_type: str, head: bytes) -> str:
ct = (content_type or "").lower()
low = head[:512].lower()
if head[:4] == b"%PDF" or "pdf" in ct:
return "pdf"
if "html" in ct or b"<!doctype html" in low or b"<html" in low:
return "html"
if "xml" in ct:
return "xml"
return "bin"
def download(url: str, dest_dir: str | Path, stem: str, timeout: int = 60) -> dict:
dest_dir = Path(dest_dir)
dest_dir.mkdir(parents=True, exist_ok=True)
r = requests.get(url, headers=BROWSER_UA, timeout=timeout, allow_redirects=True)
r.raise_for_status()
content = r.content
ctype = r.headers.get("Content-Type", "").split(";")[0].strip()
ext = _ext_for(ctype, content[:512])
path = dest_dir / f"{stem}.{ext}"
path.write_bytes(content)
return {
"path": str(path),
"ext": ext,
"bytes": len(content),
"sha256": hashlib.sha256(content).hexdigest(),
"content_type": ctype,
"http_status": r.status_code,
"final_url": r.url,
}
if __name__ == "__main__":
import sys, json
print(json.dumps(download(sys.argv[1], "data/raw",
sys.argv[2] if len(sys.argv) > 2 else "dl"), indent=2))