Upload folder using huggingface_hub

utils/nyt_archive_downloader.py

@@ -0,0 +1,107 @@
+import json
+import logging
+import os
+import random
+from concurrent.futures import Future, ThreadPoolExecutor
+from datetime import datetime
+from typing import Dict, List, Tuple
+
+import requests
+from tenacity import retry, wait_exponential, stop_after_delay
+
+logging.basicConfig(level=logging.INFO)
+LOGGER = logging.getLogger(__name__)
+
+API_KEYS = [
+    "6ZQc4XoaQPnHYOzKPLZR6kCR3ibSiafMxiWGCPeTx2VYI8cv",  # Bruno
+    "MTXkBDV8cEEs9GQk6MXrhg7DfXeqAA8Fyx28sCcn7axcXeTc",  # User 1
+    "ATqaJHb8eWZraTHOn2IO9TRAlkSYPkbaTDtVNaUzPVtbUVXu",  # User 2
+    "EEnruAuDHEhLyArTgPA3hbwthxl02jWxOBRxSMlIJG5bpgWz",  # User 3
+    "LdgKBnXAUSwWlDk0E2sf2SItbBEUvB31r1LHTf2z1fJweg5O",  # User 4
+    "rip8LQojkkYnvKeaFB5J3InNBgOuOE0QTVbmOBR8cwLLKbAM",  # User 5
+    "lsUcGIR8B3LmFKqlcRB3rqEWOh7Hv7swm9162dAHiDvSqB0G",  # User 6
+    "ngxrdtaLKqhaAL45dSMokE07XC76MmTLWcSc8P6Gq3QRlJLk",  # User 7
+    "Vnd8jKCkt8YqjTJNtCdYqLXHHhf27QgRV7r52dE1CQObUpMi",  # User 8
+    "MDSNT0GTV4yTjxLz6nUqjjRz0kyU1k679gfzUf8VlZvHK5eu",  # User 9
+]
+
+YEAR_START = 1927
+YEAR_END = 2026
+
+JSON_RAW_DIR_DEST = "./data/nyt_news_dataset/bulk"
+REQ_TIMEOUT = 60 * 10  # 10 minutes
+THREAD_MULTIPLIER = 1
+N_PROC = 8
+
+
+def _initialize_folder(folder_path: str) -> None:
+    if not os.path.exists(folder_path):
+        os.makedirs(folder_path)
+
+
+# 24 hours = 24 * 60 * 60 = 86,400 seconds
+# Max wait capped at 5 minutes = 300 seconds
+@retry(
+    stop=stop_after_delay(86400),
+    wait=wait_exponential(multiplier=2, min=1, max=300) # Backoff up to 5 minutes, than every 5 minutes
+)
+def _fetch_nyt_month(
+    year: int,
+    month: int,
+    api_key_list: List[str] = API_KEYS,
+    timeout: int = REQ_TIMEOUT,
+    json_out_dir: str = JSON_RAW_DIR_DEST,
+) -> None:
+    
+    _initialize_folder(json_out_dir)
+    file_name = f"nyt_{year}_{month}.json"
+    f_path = os.path.join(json_out_dir, file_name)
+    if os.path.exists(f_path):
+        LOGGER.warning(f"File {f_path} already exists!")
+        return
+    
+    url = f"https://api.nytimes.com/svc/archive/v1/{year}/{month}.json"
+    params = {"api-key": random.choice(api_key_list)}
+    LOGGER.info(f"Fetching articles for {year}-{month} from {url}")
+    response = requests.get(url, params=params, timeout=timeout)
+
+    if response.status_code != 200:
+        LOGGER.error(f"HTTP response {response.status_code} for {url}!")
+        raise ValueError(f"HTTP response {response.status_code}!")
+    
+    LOGGER.info(f"Fetched articles for {year}-{month} from {response.url}")
+
+    json_obj = response.json()
+    if not json_obj:
+        LOGGER.warning(f"No data found for {year}-{month} from {response.url}")
+
+    with open(f_path, "w", encoding="utf-8") as f:
+        json.dump(json_obj, f, ensure_ascii=False, indent=4)
+    LOGGER.info(f"Saved JSON file to {f_path}")
+
+
+def fetch_nyt_archive(
+    start_year: int = YEAR_START,
+    end_year: int = YEAR_END,
+    thread_mul: int = THREAD_MULTIPLIER,
+    n_proc: int = N_PROC,
+) -> None:
+    max_workers = n_proc * thread_mul
+
+    futures: List[Tuple[str, Future]] = []
+    with ThreadPoolExecutor(max_workers=max_workers) as executor:
+        for year in range(start_year, end_year + 1):
+            for month in range(1, 13):
+                future_exec = executor.submit(_fetch_nyt_month, year, month)
+                futures.append((f"{year}-{month}", future_exec))
+
+    for str_key, future in futures:
+        try:
+            future.result()
+        except Exception:
+            LOGGER.exception(f"Error processing {str_key}. Skipped!")
+    LOGGER.info("Done!")
+
+
+if __name__ == "__main__":
+    fetch_nyt_archive()