File size: 8,407 Bytes
fab5954 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 | #!/usr/bin/env python3
"""Multi-connection HTTP Range download with resume (sidecar progress file)."""
from __future__ import annotations
import json
import os
import sys
import threading
import time
import urllib.error
import urllib.request
from concurrent.futures import ThreadPoolExecutor, as_completed
URL = os.environ.get(
"DL_URL",
"https://data.bris.ac.uk/datasets/tar/2v6cgv1x04ol22qp9rm9x2j6a7.zip",
)
DEST = os.environ.get(
"DL_DEST",
os.path.join(os.path.dirname(os.path.abspath(__file__)), "2v6cgv1x04ol22qp9rm9x2j6a7.zip"),
)
WORKERS = int(os.environ.get("DL_WORKERS", "16"))
CHUNK = int(os.environ.get("DL_CHUNK", str(32 * 1024 * 1024)))
PROGRESS_PATH = DEST + ".progress.json"
LOCK_PATH = DEST + ".lock"
print_lock = threading.Lock()
progress_lock = threading.Lock()
bytes_lock = threading.Lock()
session_bytes = 0
def log(msg: str) -> None:
with print_lock:
print(msg, flush=True)
def head_length(url: str) -> int:
req = urllib.request.Request(url, method="HEAD")
with urllib.request.urlopen(req, timeout=60) as resp:
length = resp.headers.get("Content-Length")
accept = resp.headers.get("Accept-Ranges", "")
if "bytes" not in accept.lower():
log(f"WARN Accept-Ranges={accept!r}")
if not length:
raise SystemExit("no Content-Length")
return int(length)
def load_progress(total: int, dest_size: int) -> list[list[int]]:
if os.path.exists(PROGRESS_PATH):
with open(PROGRESS_PATH) as f:
data = json.load(f)
if data.get("url") == URL and data.get("total") == total:
return merge_intervals(data.get("completed") or [])
if 0 < dest_size < total:
return [[0, dest_size]]
return []
def save_progress(total: int, completed: list[list[int]]) -> None:
tmp = PROGRESS_PATH + ".tmp"
payload = {"url": URL, "total": total, "completed": merge_intervals(completed)}
with open(tmp, "w") as f:
json.dump(payload, f)
os.replace(tmp, PROGRESS_PATH)
def merge_intervals(intervals: list[list[int]]) -> list[list[int]]:
if not intervals:
return []
items = sorted((int(a), int(b)) for a, b in intervals if b > a)
out = [list(items[0])]
for a, b in items[1:]:
if a <= out[-1][1]:
out[-1][1] = max(out[-1][1], b)
else:
out.append([a, b])
return out
def gaps(total: int, completed: list[list[int]]) -> list[tuple[int, int]]:
done = merge_intervals(completed)
pos = 0
missing: list[tuple[int, int]] = []
for a, b in done:
if pos < a:
missing.append((pos, a))
pos = max(pos, b)
if pos < total:
missing.append((pos, total))
return missing
def split_gaps(missing: list[tuple[int, int]], chunk: int) -> list[tuple[int, int]]:
tasks = []
for a, b in missing:
s = a
while s < b:
e = min(s + chunk, b)
tasks.append((s, e))
s = e
return tasks
def fetch_range(url: str, start: int, end: int, fd: int, retries: int = 8) -> int:
"""Download [start, end) into fd. Returns bytes written."""
last_err: Exception | None = None
for attempt in range(retries):
written = 0
offset = start
try:
req = urllib.request.Request(
url,
headers={
"Range": f"bytes={start}-{end - 1}",
"User-Agent": "parallel-resume-get/1.0",
},
)
with urllib.request.urlopen(req, timeout=120) as resp:
if resp.status not in (200, 206):
raise RuntimeError(f"HTTP {resp.status}")
while offset < end:
want = min(1024 * 1024, end - offset)
buf = resp.read(want)
if not buf:
raise RuntimeError(f"short read at {offset}")
os.pwrite(fd, buf, offset)
n = len(buf)
offset += n
written += n
with bytes_lock:
global session_bytes
session_bytes += n
if offset != end:
raise RuntimeError(f"incomplete {offset} != {end}")
return written
except (urllib.error.URLError, TimeoutError, OSError, RuntimeError) as exc:
last_err = exc
time.sleep(min(2 ** attempt, 30))
raise RuntimeError(f"range {start}-{end} failed: {last_err}")
def main() -> int:
if os.path.exists(LOCK_PATH):
try:
with open(LOCK_PATH) as f:
old = int(f.read().strip() or "0")
os.kill(old, 0)
log(f"another downloader is running (pid {old})")
return 1
except (OSError, ValueError):
pass
with open(LOCK_PATH, "w") as f:
f.write(str(os.getpid()))
try:
total = head_length(URL)
dest_size = os.path.getsize(DEST) if os.path.exists(DEST) else 0
completed = load_progress(total, dest_size)
log(f"url={URL}")
log(f"dest={DEST}")
log(f"total={total} ({total / 1024**3:.2f} GiB) workers={WORKERS} chunk={CHUNK}")
log(f"already={sum(b - a for a, b in completed)} bytes in {len(completed)} interval(s)")
flags = os.O_RDWR | os.O_CREAT
fd = os.open(DEST, flags, 0o644)
try:
if dest_size != total:
os.ftruncate(fd, total)
tasks = split_gaps(gaps(total, completed), CHUNK)
if not tasks:
log("nothing to download")
return 0
log(f"tasks={len(tasks)} remaining={sum(e - s for s, e in tasks)}")
t0 = time.time()
done_tasks = 0
def worker(se: tuple[int, int]) -> tuple[int, int]:
s, e = se
fetch_range(URL, s, e, fd)
with progress_lock:
completed.append([s, e])
save_progress(total, completed)
return s, e
with ThreadPoolExecutor(max_workers=WORKERS) as pool:
futs = [pool.submit(worker, se) for se in tasks]
last_report = t0
last_bytes = 0
pending = set(futs)
while pending:
finished = []
for fut in list(pending):
if fut.done():
finished.append(fut)
pending.remove(fut)
for fut in finished:
fut.result()
done_tasks += 1
now = time.time()
if now - last_report >= 5 or not pending:
with bytes_lock:
cur = session_bytes
dt = now - last_report
inst = (cur - last_bytes) / dt if dt > 0 else 0
avg = cur / (now - t0) if now > t0 else 0
remain = sum(e - s for s, e in tasks) - cur
eta = remain / inst if inst > 0 else float("inf")
log(
f"chunks {done_tasks}/{len(tasks)} "
f"+{cur / 1024**3:.2f} GiB this run "
f"{inst / 1024**2:.2f} MiB/s inst "
f"{avg / 1024**2:.2f} MiB/s avg "
f"eta {eta / 60:.1f} min"
)
last_report = now
last_bytes = cur
if pending:
time.sleep(0.4)
finally:
os.close(fd)
final = merge_intervals(completed)
if gaps(total, final):
log(f"INCOMPLETE gaps={gaps(total, final)[:5]}")
return 1
size = os.path.getsize(DEST)
if size != total:
log(f"size mismatch {size} != {total}")
return 1
log(f"DONE {DEST} size={size}")
try:
os.remove(PROGRESS_PATH)
except OSError:
pass
return 0
finally:
try:
os.remove(LOCK_PATH)
except OSError:
pass
if __name__ == "__main__":
raise SystemExit(main())
|