2025/5/9
🚀 新功能
- 新增支持获取接口
归属地
与运营商
(利用纯真IP数据库
实现),支持关键字过滤,可通过配置location
与isp
生成想要的结果,建议优先使用靠近使用环境的归属地与本机网络运营商,以提升播放效果(#1058) - 新增支持无需开启测速的情况下,可对接口进行
排序
,输出结果日志
🌟 优化
- 优化
IPv6
结果进入缓存 - 调整
冻结结果的阈值
,加入最大延迟
与最小速率
限制 - 调整默认配置
ipv_type_prefer = auto
,即根据网络环境自动选择排序IPv4与IPv6结果的优先级 - 频道结果日志文件更名为
result.log
- 更新部分配置参数描述
🐛 修复
🗑️ 移除
- 移除无效的
IPv6订阅源
Note
有小伙伴对部署后首次更新时间变长有疑问,其实这是正常的。
因为从v1.7.0
开始,为了提升频道测速准确性,默认对接口进行全量测速。
目前首次运行一般30分钟
左右,如果是新增的频道比较多首次运行时间会比较长。
但这并不会影响使用,由于默认模板已经内置了部分更新结果(output/cache.pkl.gz
),部署后可立即访问使用。
同时测速阶段可根据历史数据跳过无效接口,无需担心,后续更新所需时间会明显减少。
如果你介意,可开启Host共享模式(speed_test_filter_host = True
),相同Host的接口会共享测速结果,可以大幅降低测速所需时间,但结果准确性也会下降。
English
🚀 New Features
- Added support for obtaining interface
location
andISP
(implemented using theIPIP database
), supports keyword
filtering. You can configurelocation
andisp
to generate desired results. It is recommended to prioritize the
location and ISP close to the usage environment to improve playback performance (#1058). - Added support for sorting interfaces and outputting result logs without enabling speed testing.
🌟 Optimizations
- Optimized caching of
IPv6
results. - Adjusted the
frozen result threshold
by addingmaximum latency
andminimum speed
limits. - Adjusted the default configuration
ipv_type_prefer = auto
, which automatically prioritizes sorting of IPv4 and IPv6
results based on the network environment. - Renamed the channel result log file to
result.log
. - Updated descriptions of some configuration parameters.
🐛 Bug Fixes
- Fixed the issue with matching
IPv6 results with parameters
(#1048). - Fixed the failure to generate whitelist results (#1055).
🗑️ Removals
- Removed invalid
IPv6 subscription sources
.
[!NOTE]
Some users have raised concerns about the longer initial update time after deployment. This is actually normal.
Starting fromv1.7.0
, to improve the accuracy of channel speed tests,
full speed testing of interfaces is enabled by default.
The first run usually takes about30 minutes
. If there are many new channels, the initial run time may be longer.
However, this does not affect usage, as the default template already includes some pre-updated results
(output/cache.pkl.gz
), allowing immediate access after deployment.
During the speed test phase, invalid interfaces can be skipped based on historical data, so there is no need to worry.
Subsequent updates will take significantly less time.
If you are concerned, you can enable Host sharing mode (speed_test_filter_host = True
), where interfaces with the
same Host share speed test results. This can greatly reduce the time required for speed testing,
but the accuracy of the results may decrease.