#!/usr/bin/env python3 """Download selected GitHub Actions artifact files through ranged ZIP reads.""" import argparse import io import json from pathlib import Path import re import subprocess import time import urllib.error import urllib.request import zipfile import zlib class NoRedirect(urllib.request.HTTPRedirectHandler): def redirect_request(self, request, response, code, message, headers, url): return None def download_url(repository, artifact): token = subprocess.check_output(['gh', 'auth', 'token'], text=True).strip() request = urllib.request.Request( f'https://api.github.com/repos/{repository}/actions/artifacts/{artifact}/zip?request_time={time.time_ns()}', headers={'Authorization': 'Bearer ' + token, 'Accept': 'application/vnd.github+json', 'Cache-Control': 'no-cache'}) try: urllib.request.build_opener(NoRedirect).open(request, timeout=60) except urllib.error.HTTPError as error: if error.code in (301, 302, 303, 307, 308): url = error.headers['Location'] if url.startswith('https://'): return url raise raise ValueError('GitHub did not return an artifact download location') class RangedZip(io.RawIOBase): def __init__(self, url, refresh_url=None): self.url, self.position, self.small = url, 0, None self.refresh_url = refresh_url with urllib.request.urlopen(urllib.request.Request(url, headers={'Range': 'bytes=0-0'}), timeout=60) as response: match = re.fullmatch(r'bytes 0-0/(\d+)', response.headers.get('Content-Range', '')) if response.status != 206 or not match: raise ValueError('Artifact storage must support HTTP byte ranges') self.size = int(match[1]) if self.size <= 1024 * 1024: self.small = self.read(self.size) self.position = 0 def seekable(self): return True def readable(self): return True def tell(self): return self.position def seek(self, offset, whence=io.SEEK_SET): position = offset + (self.position if whence == io.SEEK_CUR else self.size if whence == io.SEEK_END else 0) if position < 0: raise ValueError('Negative ZIP position') self.position = position return position def read(self, size=-1): size = min(self.size - self.position, size if size >= 0 else self.size) if size <= 0: return b'' if self.small is not None: data = self.small[self.position:self.position + size] self.position += size return data start, end = self.position, self.position + size - 1 for attempt in range(3): try: request = urllib.request.Request(self.url, headers={'Range': f'bytes={start}-{end}'}) with urllib.request.urlopen(request, timeout=120) as response: if response.status != 206 or response.headers.get('Content-Range') != f'bytes {start}-{end}/{self.size}': raise ValueError('Artifact range response differs from the requested bytes') data = response.read(size) if len(data) != size: raise OSError('Artifact range was truncated') self.position += size return data except (OSError, urllib.error.URLError): if attempt == 2: raise if self.refresh_url: self.url = self.refresh_url() time.sleep(attempt + 1) def main(): parser = argparse.ArgumentParser(description=__doc__) parser.add_argument('repository', help='OWNER/REPOSITORY; uses the active gh account') parser.add_argument('run', type=int) parser.add_argument('artifact') parser.add_argument('output', type=Path) parser.add_argument('--file', action='append', help='exact archive path; repeat to select files') args = parser.parse_args() if not re.fullmatch(r'[\w.-]+/[\w.-]+', args.repository): parser.error('repository must be OWNER/REPOSITORY') result = json.loads(subprocess.check_output([ 'gh', 'api', f'repos/{args.repository}/actions/runs/{args.run}/artifacts?per_page=100'])) matches = [item for item in result['artifacts'] if item['name'] == args.artifact and not item['expired']] if len(matches) != 1: parser.error('require one unexpired artifact with the selected name') refresh = lambda: download_url(args.repository, matches[0]['id']) with RangedZip(refresh(), refresh) as source, zipfile.ZipFile(source) as archive: selected = args.file or [item.filename for item in archive.infolist() if not item.is_dir()] for name in selected: relative = Path(name) if relative.is_absolute() or '..' in relative.parts: raise ValueError('Artifact path escapes the output directory') target = args.output / relative if target.exists(): raise FileExistsError('Choose a new output directory: ' + str(target)) target.parent.mkdir(parents=True, exist_ok=True) temporary = target.with_name(target.name + '.download') info, position, crc = archive.getinfo(name), 0, 0 if temporary.exists(): with temporary.open('rb') as previous: for data in iter(lambda: previous.read(64 * 1024 * 1024), b''): position += len(data) crc = zlib.crc32(data, crc) if position > info.file_size: raise ValueError('Partial download exceeds the selected artifact file') print(f'{target}: resuming at {position} bytes', flush=True) with archive.open(name) as incoming, temporary.open('ab') as outgoing: incoming.seek(position) for data in iter(lambda: incoming.read(64 * 1024 * 1024), b''): outgoing.write(data) position += len(data) crc = zlib.crc32(data, crc) if position != info.file_size or crc != info.CRC: raise zipfile.BadZipFile('Downloaded file differs from the ZIP entry: ' + name) temporary.replace(target) print(f'{target}: {target.stat().st_size} bytes', flush=True) if __name__ == '__main__': try: main() except urllib.error.HTTPError as error: raise SystemExit(f'Artifact download HTTP error {error.code}') from None except urllib.error.URLError: raise SystemExit('Artifact download connection failed') from None