perfect-postcode/download_arcgis_data.py

122 lines
4.1 KiB
Python

#!/usr/bin/env python3
"""Download ArcGIS data and convert to Parquet."""
# Run it with:
# uv run download_arcgis_data.py
import time
import zipfile
import httpx
import polars as pl
from pathlib import Path
from tqdm import tqdm
URL = "https://www.arcgis.com/sharing/rest/content/items/077631e063eb4e1ab43575d01381ec33/data"
BASE_DATA_PATH = Path("./data_sources")
BASE_DATA_PATH.mkdir(exist_ok=True)
DOWNLOAD_PATH = BASE_DATA_PATH / "arcgis_data.zip"
EXTRACT_PATH = BASE_DATA_PATH / "arcgis_extracted"
PARQUET_PATH = BASE_DATA_PATH / "arcgis_data.parquet"
MAX_RETRIES = 3
def download_with_progress(url: str, output_path: Path) -> None:
"""Download a file with progress bar and retry logic."""
for attempt in range(1, MAX_RETRIES + 1):
try:
with httpx.stream(
"GET",
url,
follow_redirects=True,
timeout=httpx.Timeout(30.0, read=None),
) as response:
response.raise_for_status() # pyright: ignore[reportUnusedCallResult]
total = int(response.headers.get("content-length", 0))
with open(output_path, "wb") as f, tqdm(
total=total,
unit="B",
unit_scale=True,
unit_divisor=1024,
desc="Downloading",
) as pbar:
for chunk in response.iter_bytes(chunk_size=8192):
f.write(chunk)
pbar.update(len(chunk))
return # Success
except (httpx.ConnectError, httpx.ReadTimeout) as e:
if attempt < MAX_RETRIES:
wait = 2**attempt
print(f"Attempt {attempt} failed: {e}. Retrying in {wait}s...")
time.sleep(wait)
else:
raise
def extract_zip(zip_path: Path, extract_path: Path) -> list[Path]:
"""Extract ZIP file and return list of extracted files."""
print("Extracting ZIP file...")
extract_path.mkdir(exist_ok=True)
with zipfile.ZipFile(zip_path, "r") as zf:
zf.extractall(extract_path)
return [extract_path / name for name in zf.namelist()]
def find_data_file(extract_path: Path) -> Path:
"""Find the main data file (CSV, XLSX, or similar) in extracted files."""
# Look for common data file extensions
for ext in ["*.csv", "*.xlsx", "*.xls", "*.json", "*.geojson"]:
files = list(extract_path.rglob(ext))
if files:
# Return the largest file if multiple found
return max(files, key=lambda f: f.stat().st_size)
raise FileNotFoundError(f"No data file found in {extract_path}")
def convert_to_parquet(data_path: Path, parquet_path: Path) -> None:
"""Convert data file to Parquet using Polars."""
print(f"Converting {data_path.name} to Parquet...")
suffix = data_path.suffix.lower()
if suffix == ".csv":
df = pl.read_csv(data_path, try_parse_dates=True)
elif suffix in [".xlsx", ".xls"]:
df = pl.read_excel(data_path)
elif suffix in [".json", ".geojson"]:
df = pl.read_json(data_path)
else:
raise ValueError(f"Unsupported file format: {suffix}")
df.write_parquet(parquet_path, compression="zstd")
print(f"Saved to {parquet_path}")
print(f"Rows: {df.height:,}")
print(f"Columns: {df.columns}")
print(f"Original size: {data_path.stat().st_size / 1024**2:.1f} MB")
print(f"Parquet size: {parquet_path.stat().st_size / 1024**2:.1f} MB")
def main() -> None:
if not DOWNLOAD_PATH.exists():
download_with_progress(URL, DOWNLOAD_PATH)
else:
print(f"File already exists at {DOWNLOAD_PATH}, skipping download")
# Check if it's a ZIP file
if zipfile.is_zipfile(DOWNLOAD_PATH):
extracted_files = extract_zip(DOWNLOAD_PATH, EXTRACT_PATH)
print(f"Extracted {len(extracted_files)} files")
data_file = find_data_file(EXTRACT_PATH)
else:
# Not a ZIP, treat as direct data file
data_file = DOWNLOAD_PATH
convert_to_parquet(data_file, PARQUET_PATH)
if __name__ == "__main__":
main()