Commit Graph

20 Commits

Author SHA1 Message Date
08c76c799c Rescue graph building test into subcommand 2024-11-23 21:02:52 +11:00
c69f2235a4 Generalise digest hex conversion 2024-11-23 18:30:04 +11:00
d9f5052d21 Housekeeping 2024-11-23 18:21:15 +11:00
14bd390114 Flatten TextInputStream wrapper and allow lines to be used directly on InputStream 2024-11-14 21:08:58 +11:00
c7185fa370 Faster UTF8 decode 2024-11-14 20:58:04 +11:00
ff38a831f2 Improve performance of TextInputStream 2024-11-14 20:40:55 +11:00
4c3fb214de idk anymore 2024-11-14 01:15:51 +11:00
775698620f apply MemInStream changes that required streams to be class-based 2024-11-14 01:04:37 +11:00
6a5da27440 Fix GZip.swift not being the actual struct name 2024-11-14 00:58:36 +11:00
da14fa8fd0 yep steam classes are faster... 2024-11-14 00:57:29 +11:00
451deb6aa3 Tried a few things to try making memory streams faster, the answer may be in just biting the bullet and making streams classes though 2024-11-14 00:36:15 +11:00
ea5dd3771b slightly more efficient line conversion 2024-11-12 21:34:06 +11:00
aff578b9a6 use zlib's full gzip support
simpler code and seems to be generally faster
2024-11-12 19:58:46 +11:00
3e47d6cd01 we are trapped in the tar (i made it 1ms faster yay for me the ratboy genius) 2024-11-12 18:41:20 +11:00
cee34c6f7a don't chunk tar reads for no reason (faster) 2024-11-12 18:37:46 +11:00
089a7dcfe1 zlib-based homecooked gzip reader for significantly faster decompression times 2024-11-12 09:37:46 +11:00
cf5e1a3f35 basic package graph implemented 2024-11-11 21:06:37 +11:00
078397f451 Replace hex decoding closure with something safer and hopefully more performant 2024-11-10 22:16:19 +11:00
d6b0acf807 Update copyright headers 2024-11-10 17:51:53 +11:00
941dfae317 Initial implementation of APKINDEX, fetching, reading, parsing, & merging 2024-11-08 21:22:33 +11:00