pput.py 文件源码

def main_loop(self, concurrency=4, worker_class=UploadWorker):
        chunk_index = 0
        self._begin_upload()
        self._workers = self._start_workers(concurrency, worker_class=worker_class)
        while self._pending_chunks or not self.stream_handler.finished:
            self._check_workers()  # raise exception and stop everything if any worker has crashed
            # print "main_loop p:{} o:{} i:{}".format(
            #     self._pending_chunks, self.outbox.qsize(), self.inbox.qsize())
            # consume results first as this is a quick operation
            self._handle_results()
            chunk = self.stream_handler.get_chunk()
            if chunk:
                # s3 multipart index is 1 based, increment before sending
                chunk_index += 1
                self._send_chunk(chunk_index, chunk)
        self._finish_upload()
        self.results.sort()
        return multipart_etag(r[1] for r in self.results)