• 新增了 dumper 的并发读取功能。
• ips dump 和 ips pack 命令现在支持 reader_jobs 参数，可用于调整读取操作的并发数。
• 实现了基于数据密度的 IP 数据库分段功能，有效解决了不同 IP 段数据密度差异的问题。
• 更新了 IPv6 正则匹配规则，移除 IPv6 Link-local 规格的匹配。

示例

读取 GeoLite2-City.mmdb 数据，重新打包生成 ipdb 数据库，1 个并发与 8 个并发的耗时对比，平台是 MBP M1 Max (8 cores)

$ time ./ips pack -i ~/.ips/GeoLite2-City.mmdb -o 1.ipdb --reader-jobs 1
./ips pack -i ~/.ips/GeoLite2-City.mmdb -o 1.ipdb --reader-jobs 1  596.91s user 54.99s system 132% cpu 8:13.25 total

$ time ./ips pack -i ~/.ips/GeoLite2-City.mmdb -o 1.ipdb --reader-jobs 8
./ips pack -i ~/.ips/GeoLite2-City.mmdb -o 1.ipdb --reader-jobs 8  698.84s user 25.46s system 777% cpu 1:33.19 total

• Added concurrent reading functionality to the dumper.
• The ips dump and ips pack commands have been enhanced to support the reader_jobs parameter, allowing adjustment of the concurrency level for reading operations.
• Implemented segmentation of the IP database based on data density, effectively addressing the issue of data density variations across different IP segments.
• Updated IPv6 regular matching rules to remove IPv6 Link-local specification matching.

Example

Reading GeoLite2-City.mmdb data and repackaging it into an ipdb database, comparing the time taken with 1 and 8 concurrent jobs, the platform used is MBP M1 Max (8 cores)

$ time ./ips pack -i ~/.ips/GeoLite2-City.mmdb -o 1.ipdb --reader-jobs 1
./ips pack -i ~/.ips/GeoLite2-City.mmdb -o 1.ipdb --reader-jobs 1  596.91s user 54.99s system 132% cpu 8:13.25 total

$ time ./ips pack -i ~/.ips/GeoLite2-City.mmdb -o 1.ipdb --reader-jobs 8
./ips pack -i ~/.ips/GeoLite2-City.mmdb -o 1.ipdb --reader-jobs 8  698.84s user 25.46s system 777% cpu 1:33.19 total