Parallel downloading don't work in python threading

I'm building a parallel download library using threading module.

When I use my library, it downloads the file without error, but the video file doesn't have the same content as if I downloaded it through the browser.

I use threading for parallel downloading and I think I have a problem with threading.Lock and file.seek, but I can't figure out how to fix the problem.

This is my code:

import requests
import threading
from tqdm import tqdm
DOWNLOAD_CHUNK_SIZE = 1 << 20 # 1 MiB
class DownloadPart:
def __init__(self, url, byte_range) -> None:
self.url = url
self.byte_range = byte_range
self.lock = threading.Lock()
def download(self, file, pbar=None):
response = requests.get(
self.url,
headers={"Range": "bytes={}-{}".format(*self.byte_range)},
allow_redirects=True,
stream=True,
)
written = 0
for chunk in response.iter_content(chunk_size=DOWNLOAD_CHUNK_SIZE):
if chunk:
self.lock.acquire()
file.seek(self.byte_range[0] + written)
length = file.write(chunk)
file.flush()
written += length
pbar.update(length)
self.lock.release()
class Downloader:
def __init__(self, url, parts=10):
self.url = url
self.parts = parts
def _get_file_size(self) -> int:
info = requests.head(self.url, allow_redirects=True)
info.raise_for_status()
size = info.headers.get("content-length", None)
assert size
return int(size)
def download(self, filename):
file_size = self._get_file_size()
# file_size = 1024
size_per_part = file_size // self.parts
print(file_size, size_per_part)
file = open(filename, "wb")
pbar = tqdm(total=file_size)
threads = []
for index in range(self.parts):
# fix last part have more bytes
if index + 1 == self.parts:
byte_range = (size_per_part * index, file_size - 1)
else:
byte_range = (size_per_part * index, size_per_part * (index + 1) - 1)
thread = threading.Thread(
target=DownloadPart(self.url, byte_range).download, args=(file,), kwargs={"pbar": pbar}
)
thread.start()
threads.append(thread)
for thread in threads:
thread.join()
file.close()
URL = "https://s-delivery38.mxdcontent.net/v/8a5f59673042ed97c402be84ceeb20d9.mp4?s=TfiDzO2oBLrhub_GhToCiQ&e=1676489987&_t=1676476332"
d = Downloader(URL)
d.download("video.mp4")

How can I solve the problem with my library and get the same data in the file? Thank you for any help.

There were two problems with my code:

1. I found a solution to the first problem here. https://stackoverflow.com/a/25165183/14900791:
   > The Lock() function creates an entirely new lock - one that only the
   > thread calling the function can use. That's why it doesn't work,
   > because each thread is locking an entirely different lock.

2. Mixdrop (mxdcontent.net) only allows two videos in the same ip, so the code only works for two parts, the others got status code 509 (I didn't checked the status code so I didn't get an error).

import requests
import threading
from tqdm import tqdm

DOWNLOAD_CHUNK_SIZE = 1 << 20 # 1 MiB

# global lock instance
lock = threading.Lock()

class DownloadPart:
    def __init__(self, url, byte_range) -> None:
        self.url = url
        self.byte_range = byte_range

    def download(self, file, pbar=None):
        response = requests.get(
            self.url,
            headers={"Range": "bytes={}-{}".format(*self.byte_range)},
            allow_redirects=True,
            stream=True,
        )

        written = 0

        for chunk in response.iter_content(chunk_size=DOWNLOAD_CHUNK_SIZE):
            if chunk:
                lock.acquire()

                file.seek(self.byte_range[0] + written)
                length = file.write(chunk)
                file.flush()

                written += length
                pbar.update(length)

                lock.release()


class Downloader:
    def __init__(self, url, parts=10):
        self.url = url
        self.parts = parts

    def _get_file_size(self) -> int:
        info = requests.head(self.url, allow_redirects=True)

        info.raise_for_status()

        size = info.headers.get("content-length", None)
        assert size
        return int(size)

    def download(self, filename):
        file_size = self._get_file_size()
        # file_size = 1024

        size_per_part = file_size // self.parts

        print(file_size, size_per_part)

        file = open(filename, "wb")

        pbar = tqdm(total=file_size)

        threads = []
        for index in range(self.parts):
            # fix last part have more bytes
            if index + 1 == self.parts:
                byte_range = (size_per_part * index, file_size - 1)
            else:
                byte_range = (size_per_part * index, size_per_part * (index + 1) - 1)

            thread = threading.Thread(
                target=DownloadPart(self.url, byte_range).download, args=(file,), kwargs={"pbar": pbar}
            )
            thread.start()
            threads.append(thread)

        for thread in threads:
            thread.join()

        file.close()


URL = "https://s-delivery38.mxdcontent.net/v/8a5f59673042ed97c402be84ceeb20d9.mp4?s=TfiDzO2oBLrhub_GhToCiQ&e=1676489987&_t=1676476332"

d = Downloader(URL)

d.download("video.mp4")