mirror of
https://github.com/Hommy-master/capcut-mate.git
synced 2026-08-28 23:27:50 +08:00
优化资源文件下载，增加断点续传的能力。
This commit is contained in:
Hommy
+201
-45
@@ -310,6 +310,113 @@ def _http_get(url: str, **kwargs) -> requests.Response:
|
||||
return requests.get(url, headers=headers, **kwargs)
|
||||
|
||||
|
||||
# 草稿文件列表里既有 json 元数据，也有音视频/图片。仅后者需要 HTTP Range 断点续传。
|
||||
_MEDIA_RESOURCE_EXTENSIONS = frozenset(
|
||||
{
|
||||
".mp4",
|
||||
".mov",
|
||||
".avi",
|
||||
".mkv",
|
||||
".webm",
|
||||
".m4v",
|
||||
".flv",
|
||||
".wmv",
|
||||
".ts",
|
||||
".mpeg",
|
||||
".mpg",
|
||||
".3gp",
|
||||
".mp3",
|
||||
".wav",
|
||||
".aac",
|
||||
".m4a",
|
||||
".flac",
|
||||
".ogg",
|
||||
".wma",
|
||||
".aiff",
|
||||
".aif",
|
||||
".opus",
|
||||
".amr",
|
||||
".jpg",
|
||||
".jpeg",
|
||||
".png",
|
||||
".gif",
|
||||
".webp",
|
||||
".bmp",
|
||||
".tiff",
|
||||
".tif",
|
||||
".heic",
|
||||
".heif",
|
||||
".ico",
|
||||
".svg",
|
||||
}
|
||||
)
|
||||
|
||||
|
||||
def _extract_extension(url_or_path: str) -> str:
|
||||
"""从 URL 或本地路径取出小写扩展名（忽略 query）。"""
|
||||
if not url_or_path:
|
||||
return ""
|
||||
path = urlparse(url_or_path).path if "://" in url_or_path else url_or_path
|
||||
return os.path.splitext(path)[1].lower()
|
||||
|
||||
|
||||
def _is_media_resource(url_or_path: str) -> bool:
|
||||
"""
|
||||
判断是否为需要断点续传的资源文件（视频/图片/音频）。
|
||||
|
||||
json、bin、草稿元数据等返回 False，保持「失败即整文件重下」的原有行为。
|
||||
"""
|
||||
return _extract_extension(url_or_path) in _MEDIA_RESOURCE_EXTENSIONS
|
||||
|
||||
|
||||
def _local_file_size(path: str) -> int:
|
||||
"""返回本地文件字节数；不存在或无法读取时视为 0。"""
|
||||
try:
|
||||
if os.path.isfile(path):
|
||||
return os.path.getsize(path)
|
||||
except OSError:
|
||||
return 0
|
||||
return 0
|
||||
|
||||
|
||||
def _resume_request_headers(local_path: str) -> Tuple[Optional[Dict[str, str]], int]:
|
||||
"""
|
||||
按本地半成品大小构造 Range 请求头。
|
||||
|
||||
无半成品时返回 (None, 0)，调用方应发普通 GET，请求形态与改造前一致。
|
||||
"""
|
||||
size = _local_file_size(local_path)
|
||||
if size <= 0:
|
||||
return None, 0
|
||||
return {"Range": f"bytes={size}-"}, size
|
||||
|
||||
|
||||
def _is_download_success_status(status_code: int, resume_from: int) -> bool:
|
||||
"""200 始终成功；206 仅在确实携带断点（resume_from>0）时视为续传成功。"""
|
||||
if status_code == 200:
|
||||
return True
|
||||
return status_code == 206 and resume_from > 0
|
||||
|
||||
|
||||
def _write_http_body_to_file(
|
||||
response: requests.Response, file_path: str, *, append: bool
|
||||
) -> None:
|
||||
"""
|
||||
将 HTTP 响应体写入本地文件。
|
||||
|
||||
append=True：断点续传，在已有内容后追加（配合 206）。
|
||||
append=False：覆盖写入（JSON 等非资源，或服务端忽略 Range 返回 200 时的整文件重下）。
|
||||
"""
|
||||
parent_dir = os.path.dirname(file_path)
|
||||
if parent_dir:
|
||||
os.makedirs(parent_dir, exist_ok=True)
|
||||
mode = "ab" if append else "wb"
|
||||
with open(file_path, mode) as out:
|
||||
for chunk in response.iter_content(chunk_size=8192):
|
||||
if chunk:
|
||||
out.write(chunk)
|
||||
|
||||
|
||||
def _resolve_download_target_path(
|
||||
file_url: str, target_dir: str
|
||||
) -> Tuple[str, Optional[str]]:
|
||||
@@ -715,16 +822,33 @@ def _download_single_file(file_url: str, target_dir: str) -> None:
|
||||
retry_count = 0
|
||||
|
||||
full_file_path, url_draft_id = _resolve_download_target_path(file_url, target_dir)
|
||||
# 仅视频/图片/音频走 Range 续传；json 等仍每次整文件覆盖下载。
|
||||
enable_resume = _is_media_resource(file_url) or _is_media_resource(full_file_path)
|
||||
|
||||
while retry_count <= max_retries:
|
||||
try:
|
||||
response = _http_get(
|
||||
file_url,
|
||||
timeout=(_REQUEST_CONNECT_TIMEOUT, _REQUEST_READ_TIMEOUT),
|
||||
stream=True,
|
||||
)
|
||||
extra_headers = None
|
||||
resume_from = 0
|
||||
if enable_resume:
|
||||
extra_headers, resume_from = _resume_request_headers(full_file_path)
|
||||
if resume_from > 0:
|
||||
logger.info(
|
||||
"Resume media download from byte %s: %s",
|
||||
resume_from,
|
||||
file_url,
|
||||
)
|
||||
|
||||
get_kwargs = {
|
||||
"timeout": (_REQUEST_CONNECT_TIMEOUT, _REQUEST_READ_TIMEOUT),
|
||||
"stream": True,
|
||||
}
|
||||
# 无半成品时不传 headers，保持与改造前完全相同的 GET 调用形态。
|
||||
if extra_headers:
|
||||
get_kwargs["headers"] = extra_headers
|
||||
|
||||
response = _http_get(file_url, **get_kwargs)
|
||||
try:
|
||||
if response.status_code != 200:
|
||||
if not _is_download_success_status(response.status_code, resume_from):
|
||||
if not _is_retryable_http_status(response.status_code):
|
||||
status = response.status_code
|
||||
logger.error(
|
||||
@@ -763,14 +887,14 @@ def _download_single_file(file_url: str, target_dir: str) -> None:
|
||||
_sleep_transient_http_backoff(retry_count, response)
|
||||
continue
|
||||
|
||||
parent_dir = os.path.dirname(full_file_path)
|
||||
if parent_dir:
|
||||
os.makedirs(parent_dir, exist_ok=True)
|
||||
|
||||
with open(full_file_path, "wb") as out:
|
||||
for chunk in response.iter_content(chunk_size=8192):
|
||||
if chunk:
|
||||
out.write(chunk)
|
||||
# 206：从断点追加；200：整文件覆盖（含服务端忽略 Range 的情况）。
|
||||
append = response.status_code == 206 and resume_from > 0
|
||||
if resume_from > 0 and response.status_code == 200:
|
||||
logger.info(
|
||||
"Server ignored Range, re-downloading from start: %s",
|
||||
file_url,
|
||||
)
|
||||
_write_http_body_to_file(response, full_file_path, append=append)
|
||||
finally:
|
||||
response.close()
|
||||
|
||||
@@ -1040,15 +1164,31 @@ def _download_remote_material_raising(
|
||||
fallback_ext: str,
|
||||
) -> str:
|
||||
"""下载 URL 素材；失败抛出 DraftDownloadAbort。"""
|
||||
# 本函数只拉取音视频/图片（含无扩展名的 CDN URL），始终允许断点续传。
|
||||
local_path: Optional[str] = None
|
||||
for attempt in range(_MAX_RETRIES + 1):
|
||||
response = None
|
||||
try:
|
||||
response = _http_get(
|
||||
file_url,
|
||||
timeout=(_REQUEST_CONNECT_TIMEOUT, _REQUEST_READ_TIMEOUT),
|
||||
stream=True,
|
||||
)
|
||||
if response.status_code != 200:
|
||||
extra_headers = None
|
||||
resume_from = 0
|
||||
if local_path:
|
||||
extra_headers, resume_from = _resume_request_headers(local_path)
|
||||
if resume_from > 0:
|
||||
logger.info(
|
||||
"Resume remote material from byte %s: %s",
|
||||
resume_from,
|
||||
file_url,
|
||||
)
|
||||
|
||||
get_kwargs = {
|
||||
"timeout": (_REQUEST_CONNECT_TIMEOUT, _REQUEST_READ_TIMEOUT),
|
||||
"stream": True,
|
||||
}
|
||||
if extra_headers:
|
||||
get_kwargs["headers"] = extra_headers
|
||||
|
||||
response = _http_get(file_url, **get_kwargs)
|
||||
if not _is_download_success_status(response.status_code, resume_from):
|
||||
if not _is_retryable_http_status(response.status_code):
|
||||
status = response.status_code
|
||||
logger.error(
|
||||
@@ -1082,19 +1222,20 @@ def _download_remote_material_raising(
|
||||
response.close()
|
||||
continue
|
||||
|
||||
ext = _infer_ext_from_content_type(
|
||||
response.headers.get("Content-Type"), fallback_ext
|
||||
)
|
||||
filename = _build_material_filename(base_name, ext)
|
||||
local_path = os.path.join(target_dir, "assets", sub_dir, filename)
|
||||
if local_path is None:
|
||||
ext = _infer_ext_from_content_type(
|
||||
response.headers.get("Content-Type"), fallback_ext
|
||||
)
|
||||
filename = _build_material_filename(base_name, ext)
|
||||
local_path = os.path.join(target_dir, "assets", sub_dir, filename)
|
||||
|
||||
parent_dir = os.path.dirname(local_path)
|
||||
if parent_dir:
|
||||
os.makedirs(parent_dir, exist_ok=True)
|
||||
with open(local_path, "wb") as f:
|
||||
for chunk in response.iter_content(chunk_size=8192):
|
||||
if chunk:
|
||||
f.write(chunk)
|
||||
append = response.status_code == 206 and resume_from > 0
|
||||
if resume_from > 0 and response.status_code == 200:
|
||||
logger.info(
|
||||
"Server ignored Range, re-downloading from start: %s",
|
||||
file_url,
|
||||
)
|
||||
_write_http_body_to_file(response, local_path, append=append)
|
||||
return local_path
|
||||
except DraftDownloadAbort:
|
||||
raise
|
||||
@@ -1152,14 +1293,29 @@ def _download_remote_file(file_url: str, local_path: str) -> bool:
|
||||
|
||||
def _download_remote_file_raising(file_url: str, local_path: str) -> None:
|
||||
"""下载单个 URL 素材；失败抛出 DraftDownloadAbort。"""
|
||||
enable_resume = _is_media_resource(file_url) or _is_media_resource(local_path)
|
||||
for attempt in range(_MAX_RETRIES + 1):
|
||||
try:
|
||||
response = _http_get(
|
||||
file_url,
|
||||
timeout=(_REQUEST_CONNECT_TIMEOUT, _REQUEST_READ_TIMEOUT),
|
||||
stream=True,
|
||||
)
|
||||
if response.status_code != 200:
|
||||
extra_headers = None
|
||||
resume_from = 0
|
||||
if enable_resume:
|
||||
extra_headers, resume_from = _resume_request_headers(local_path)
|
||||
if resume_from > 0:
|
||||
logger.info(
|
||||
"Resume remote file from byte %s: %s",
|
||||
resume_from,
|
||||
file_url,
|
||||
)
|
||||
|
||||
get_kwargs = {
|
||||
"timeout": (_REQUEST_CONNECT_TIMEOUT, _REQUEST_READ_TIMEOUT),
|
||||
"stream": True,
|
||||
}
|
||||
if extra_headers:
|
||||
get_kwargs["headers"] = extra_headers
|
||||
|
||||
response = _http_get(file_url, **get_kwargs)
|
||||
if not _is_download_success_status(response.status_code, resume_from):
|
||||
if not _is_retryable_http_status(response.status_code):
|
||||
status = response.status_code
|
||||
logger.error(
|
||||
@@ -1193,13 +1349,13 @@ def _download_remote_file_raising(file_url: str, local_path: str) -> None:
|
||||
response.close()
|
||||
continue
|
||||
|
||||
parent_dir = os.path.dirname(local_path)
|
||||
if parent_dir:
|
||||
os.makedirs(parent_dir, exist_ok=True)
|
||||
with open(local_path, "wb") as f:
|
||||
for chunk in response.iter_content(chunk_size=8192):
|
||||
if chunk:
|
||||
f.write(chunk)
|
||||
append = response.status_code == 206 and resume_from > 0
|
||||
if resume_from > 0 and response.status_code == 200:
|
||||
logger.info(
|
||||
"Server ignored Range, re-downloading from start: %s",
|
||||
file_url,
|
||||
)
|
||||
_write_http_body_to_file(response, local_path, append=append)
|
||||
return
|
||||
except DraftDownloadAbort:
|
||||
raise
|
||||
|
||||
@@ -0,0 +1,292 @@
|
||||
"""草稿下载：资源文件断点续传；JSON 等非资源仍整文件重下。"""
|
||||
import os
|
||||
import tempfile
|
||||
from unittest.mock import MagicMock, patch
|
||||
|
||||
import pytest
|
||||
import requests
|
||||
|
||||
import src.utils.draft_downloader as dd
|
||||
|
||||
|
||||
@pytest.fixture
|
||||
def no_sleep():
|
||||
with patch.object(dd, "time") as m_time:
|
||||
m_time.sleep = MagicMock()
|
||||
yield m_time
|
||||
|
||||
|
||||
def _stream_response(
|
||||
chunks,
|
||||
status: int = 200,
|
||||
headers=None,
|
||||
raise_after=None,
|
||||
) -> MagicMock:
|
||||
r = MagicMock()
|
||||
r.status_code = status
|
||||
r.headers = headers or {}
|
||||
r.close = MagicMock()
|
||||
|
||||
def iter_content(chunk_size=8192):
|
||||
for chunk in chunks:
|
||||
yield chunk
|
||||
if raise_after is not None:
|
||||
raise raise_after
|
||||
|
||||
r.iter_content = iter_content
|
||||
return r
|
||||
|
||||
|
||||
def _range_from_call(call) -> str:
|
||||
headers = call.kwargs.get("headers") or {}
|
||||
return headers.get("Range", "")
|
||||
|
||||
|
||||
class TestIsMediaResource:
|
||||
@pytest.mark.parametrize(
|
||||
"value",
|
||||
[
|
||||
"clip.mp4",
|
||||
"https://cdn.example.com/a.MP4?token=1",
|
||||
r"C:\draft\assets\videos\x.mov",
|
||||
"photo.PNG",
|
||||
"https://x.test/img.jpg",
|
||||
"audio.mp3",
|
||||
"track.wav",
|
||||
"pic.webp",
|
||||
],
|
||||
)
|
||||
def test_media_extensions_are_detected(self, value: str) -> None:
|
||||
assert dd._is_media_resource(value) is True
|
||||
|
||||
@pytest.mark.parametrize(
|
||||
"value",
|
||||
[
|
||||
"draft_content.json",
|
||||
"draft_meta_info.json",
|
||||
"https://cdn.example.com/app/output/draft/20251204214904ccb1af38/a.bin",
|
||||
"notes.txt",
|
||||
"https://x.test/foo.image?sig=1",
|
||||
"",
|
||||
],
|
||||
)
|
||||
def test_non_media_extensions_are_excluded(self, value: str) -> None:
|
||||
assert dd._is_media_resource(value) is False
|
||||
|
||||
|
||||
class TestResumeHelpers:
|
||||
def test_resume_headers_none_when_file_missing(self, tmp_path) -> None:
|
||||
headers, resume_from = dd._resume_request_headers(str(tmp_path / "miss.mp4"))
|
||||
assert headers is None
|
||||
assert resume_from == 0
|
||||
|
||||
def test_resume_headers_use_existing_size(self, tmp_path) -> None:
|
||||
path = tmp_path / "part.mp4"
|
||||
path.write_bytes(b"hello")
|
||||
headers, resume_from = dd._resume_request_headers(str(path))
|
||||
assert resume_from == 5
|
||||
assert headers == {"Range": "bytes=5-"}
|
||||
|
||||
def test_success_status_206_only_when_resuming(self) -> None:
|
||||
assert dd._is_download_success_status(200, 0) is True
|
||||
assert dd._is_download_success_status(200, 10) is True
|
||||
assert dd._is_download_success_status(206, 10) is True
|
||||
assert dd._is_download_success_status(206, 0) is False
|
||||
assert dd._is_download_success_status(404, 10) is False
|
||||
|
||||
|
||||
class TestDownloadSingleFileResume:
|
||||
_BASE = "https://capcut.example.com"
|
||||
_DRAFT = "20251204214904ccb1af38"
|
||||
_TIMEOUT = (dd._REQUEST_CONNECT_TIMEOUT, dd._REQUEST_READ_TIMEOUT)
|
||||
_HEADERS = dd._REQUEST_HEADERS
|
||||
|
||||
def _url(self, name: str) -> str:
|
||||
return f"{self._BASE}/app/output/draft/{self._DRAFT}/{name}"
|
||||
|
||||
def test_first_media_request_has_no_range_header(self, no_sleep) -> None:
|
||||
"""首次下载资源文件不带 Range，GET 形态与改造前一致。"""
|
||||
file_url = self._url("Resources/clip.mp4")
|
||||
with tempfile.TemporaryDirectory() as td:
|
||||
resp = _stream_response([b"ab", b"cd"])
|
||||
with patch.object(dd, "requests") as m_req:
|
||||
m_req.get.return_value = resp
|
||||
m_req.exceptions = requests.exceptions
|
||||
assert dd.download_single_file(file_url, td) is True
|
||||
m_req.get.assert_called_once_with(
|
||||
file_url,
|
||||
timeout=self._TIMEOUT,
|
||||
stream=True,
|
||||
headers=self._HEADERS,
|
||||
)
|
||||
out = os.path.join(td, "Resources", "clip.mp4")
|
||||
with open(out, "rb") as f:
|
||||
assert f.read() == b"abcd"
|
||||
|
||||
def test_media_retry_sends_range_and_appends(self, no_sleep) -> None:
|
||||
"""中途断开后，重试从已写入字节续传，206 响应追加到原文件。"""
|
||||
file_url = self._url("assets/clip.mp4")
|
||||
first = _stream_response(
|
||||
[b"hello"],
|
||||
raise_after=requests.exceptions.ChunkedEncodingError("truncated"),
|
||||
)
|
||||
second = _stream_response([b"world"], status=206)
|
||||
|
||||
with tempfile.TemporaryDirectory() as td:
|
||||
with patch.object(dd, "requests") as m_req:
|
||||
m_req.get.side_effect = [first, second]
|
||||
m_req.exceptions = requests.exceptions
|
||||
assert dd.download_single_file(file_url, td) is True
|
||||
|
||||
assert m_req.get.call_count == 2
|
||||
assert "Range" not in (m_req.get.call_args_list[0].kwargs.get("headers") or {})
|
||||
assert _range_from_call(m_req.get.call_args_list[1]) == "bytes=5-"
|
||||
out = os.path.join(td, "assets", "clip.mp4")
|
||||
with open(out, "rb") as f:
|
||||
assert f.read() == b"helloworld"
|
||||
|
||||
def test_media_retry_overwrites_when_server_ignores_range(self, no_sleep) -> None:
|
||||
"""服务端忽略 Range 返回 200 时，整文件覆盖，避免拼出损坏文件。"""
|
||||
file_url = self._url("assets/clip.mp4")
|
||||
first = _stream_response(
|
||||
[b"hello"],
|
||||
raise_after=requests.exceptions.ChunkedEncodingError("truncated"),
|
||||
)
|
||||
second = _stream_response([b"FULLFILE"], status=200)
|
||||
|
||||
with tempfile.TemporaryDirectory() as td:
|
||||
with patch.object(dd, "requests") as m_req:
|
||||
m_req.get.side_effect = [first, second]
|
||||
m_req.exceptions = requests.exceptions
|
||||
assert dd.download_single_file(file_url, td) is True
|
||||
out = os.path.join(td, "assets", "clip.mp4")
|
||||
with open(out, "rb") as f:
|
||||
assert f.read() == b"FULLFILE"
|
||||
|
||||
def test_json_retry_overwrites_without_range(self, no_sleep) -> None:
|
||||
"""JSON 失败重试必须整文件重下，即使本地已有半成品也不发 Range。"""
|
||||
file_url = self._url("draft_meta_info.json")
|
||||
first = _stream_response(
|
||||
[b'{"x":'],
|
||||
raise_after=requests.exceptions.ChunkedEncodingError("truncated"),
|
||||
)
|
||||
second = _stream_response([b'{"ok": true}'])
|
||||
|
||||
with tempfile.TemporaryDirectory() as td:
|
||||
with patch.object(dd, "requests") as m_req:
|
||||
m_req.get.side_effect = [first, second]
|
||||
m_req.exceptions = requests.exceptions
|
||||
assert dd.download_single_file(file_url, td) is True
|
||||
|
||||
second_headers = m_req.get.call_args_list[1].kwargs.get("headers") or {}
|
||||
assert "Range" not in second_headers
|
||||
out = os.path.join(td, "draft_meta_info.json")
|
||||
with open(out, "rb") as f:
|
||||
assert f.read() == b'{"ok": true}'
|
||||
|
||||
def test_bin_retry_overwrites_without_range(self, no_sleep) -> None:
|
||||
file_url = self._url("assets/x.bin")
|
||||
first = _stream_response(
|
||||
[b"123"],
|
||||
raise_after=requests.exceptions.ReadTimeout("stalled"),
|
||||
)
|
||||
second = _stream_response([b"abc"])
|
||||
with tempfile.TemporaryDirectory() as td:
|
||||
with patch.object(dd, "requests") as m_req:
|
||||
m_req.get.side_effect = [first, second]
|
||||
m_req.exceptions = requests.exceptions
|
||||
assert dd.download_single_file(file_url, td) is True
|
||||
assert "Range" not in (m_req.get.call_args_list[1].kwargs.get("headers") or {})
|
||||
out = os.path.join(td, "assets", "x.bin")
|
||||
with open(out, "rb") as f:
|
||||
assert f.read() == b"abc"
|
||||
|
||||
def test_image_resume_same_as_video(self, no_sleep) -> None:
|
||||
file_url = self._url("Resources/pic.png")
|
||||
first = _stream_response(
|
||||
[b"PNG"],
|
||||
raise_after=requests.exceptions.ChunkedEncodingError("truncated"),
|
||||
)
|
||||
second = _stream_response([b"DATA"], status=206)
|
||||
with tempfile.TemporaryDirectory() as td:
|
||||
with patch.object(dd, "requests") as m_req:
|
||||
m_req.get.side_effect = [first, second]
|
||||
m_req.exceptions = requests.exceptions
|
||||
assert dd.download_single_file(file_url, td) is True
|
||||
assert _range_from_call(m_req.get.call_args_list[1]) == "bytes=3-"
|
||||
out = os.path.join(td, "Resources", "pic.png")
|
||||
with open(out, "rb") as f:
|
||||
assert f.read() == b"PNGDATA"
|
||||
|
||||
|
||||
class TestDownloadRemoteMaterialResume:
|
||||
def test_extensionless_cdn_url_still_resumes(self, no_sleep) -> None:
|
||||
"""无扩展名的图片 CDN URL 也走续传（本函数只下载素材）。"""
|
||||
url = "https://p3-bot-workflow-sign.byteimg.com/tos-cn-i-mdko3gqilj/foo.png~tplv.image?x=1"
|
||||
first = _stream_response(
|
||||
[b"img-"],
|
||||
headers={"Content-Type": "image/png"},
|
||||
raise_after=requests.exceptions.ChunkedEncodingError("truncated"),
|
||||
)
|
||||
second = _stream_response(
|
||||
[b"rest"],
|
||||
status=206,
|
||||
headers={"Content-Type": "image/png"},
|
||||
)
|
||||
with tempfile.TemporaryDirectory() as td:
|
||||
with patch.object(dd, "requests") as m_req:
|
||||
m_req.get.side_effect = [first, second]
|
||||
m_req.exceptions = requests.exceptions
|
||||
path = dd._download_remote_material(url, td, "images", "双行", ".mp4")
|
||||
assert path is not None
|
||||
assert path.endswith(".png")
|
||||
with open(path, "rb") as f:
|
||||
assert f.read() == b"img-rest"
|
||||
assert _range_from_call(m_req.get.call_args_list[1]) == "bytes=4-"
|
||||
|
||||
def test_first_material_request_has_no_range(self, no_sleep) -> None:
|
||||
url = "https://cdn.example.com/v?id=1"
|
||||
resp = _stream_response([b"mp4"], headers={"Content-Type": "video/mp4"})
|
||||
with tempfile.TemporaryDirectory() as td:
|
||||
with patch.object(dd, "requests") as m_req:
|
||||
m_req.get.return_value = resp
|
||||
m_req.exceptions = requests.exceptions
|
||||
path = dd._download_remote_material(url, td, "videos", "clip1", ".bin")
|
||||
assert path is not None
|
||||
first_headers = m_req.get.call_args.kwargs.get("headers") or {}
|
||||
assert "Range" not in first_headers
|
||||
m_req.get.assert_called_once()
|
||||
|
||||
|
||||
class TestDownloadRemoteFileResume:
|
||||
def test_mp4_retry_appends_on_206(self, no_sleep) -> None:
|
||||
first = _stream_response(
|
||||
[b"AAA"],
|
||||
raise_after=requests.exceptions.ChunkedEncodingError("truncated"),
|
||||
)
|
||||
second = _stream_response([b"BBB"], status=206)
|
||||
with tempfile.TemporaryDirectory() as td:
|
||||
out = os.path.join(td, "a.mp4")
|
||||
with patch.object(dd, "requests") as m_req:
|
||||
m_req.get.side_effect = [first, second]
|
||||
m_req.exceptions = requests.exceptions
|
||||
assert dd._download_remote_file("https://x.test/a.mp4", out) is True
|
||||
assert _range_from_call(m_req.get.call_args_list[1]) == "bytes=3-"
|
||||
with open(out, "rb") as f:
|
||||
assert f.read() == b"AAABBB"
|
||||
|
||||
def test_non_media_remote_file_does_not_resume(self, no_sleep) -> None:
|
||||
first = _stream_response(
|
||||
[b"AAA"],
|
||||
raise_after=requests.exceptions.ChunkedEncodingError("truncated"),
|
||||
)
|
||||
second = _stream_response([b"ZZZ"])
|
||||
with tempfile.TemporaryDirectory() as td:
|
||||
out = os.path.join(td, "a.bin")
|
||||
with patch.object(dd, "requests") as m_req:
|
||||
m_req.get.side_effect = [first, second]
|
||||
m_req.exceptions = requests.exceptions
|
||||
assert dd._download_remote_file("https://x.test/a.bin", out) is True
|
||||
assert "Range" not in (m_req.get.call_args_list[1].kwargs.get("headers") or {})
|
||||
with open(out, "rb") as f:
|
||||
assert f.read() == b"ZZZ"
|
||||
Reference in New Issue
Block a user