This fixes both, replay-exclusive keybindings blocking vanilla ones as well as
vanilla keybindings blocking replay-exclusive ones.
The magic happens in Mixin_ContextualKeyBindings.
Aside from being best practice, this is necessary for the recording CRC to be
written to the replay file (cause that only happens on close).
Also potentially fixes https://github.com/ReplayMod/ReplayMod/issues/144 though
that should already be covered by ZipReplayFile tracking and closing all of its
`OutputStream`s before saving.
We didn't used to do that because it would have required us to serialize and
then de-serialize as MCProtocolLib only to re-serialize each packet.
With ReplayStudio v2, we only have to convert to from MC's ByteBuf types to
ReplayStudio's ByteBuf (fairly trivial).
This also allows us to automatically make use of ReplayStudio's backwards
compatibility features: We didn't used to record the login phase pre-1.14 and we
still don't, so we now would have to inject a fake login packet (because we're
no longer generating legacy-version replay files) but ReplayOutputStream can
do that for us.
Without this commit, any pre-1.14 replays generated are unplayable due to the
missing LoginSuccess packet.
In the case of #118, this was resulting in the EntityTrackerUpdate packet (which
contains info on which skin layers should be visible) sometimes ending up before
the SpawnPlayer packet, hence the players outer skin layers being invisible at
the start of the recording.
The new preprocessor version brings first-party support for fabric, in
particular for automatically fetching mappings loom and for
automatically remapping between MC versions via intermediary mappings.
This will come in handy when updating to 1.15.
It also supports automatic remapping between fabric and forge on the
same MC version (by going mcp<->srg<->mojang<->intermediary<->yarn).
Therefore this commit replaces the previous 1.13.2 (forge) subproject
with a 1.14.4-forge one. Instead of manually providing full mappings for
1.14.4-fabric to 1.13.2-forge and partial (classes) for 1.13.2-forge to
1.12.2-forge, we now only need partial (classes) for 1.12.2-forge to
1.14.4-forge and preprocessor will take care of the forge to fabric
step (in fact, our mapping file for that is currently completely empty).
In an attempt to write an IDE (IntelliJ) plugin for quicker and easier
working with the preprocessor (e.g. immediately mapping single files
from within the IDE, jumping between different versions for the same
file), the preprocessor gradle plugin declaration now requires you to
explicitly specify variables in the common.gradle and the overall
relationship between projects in the build.gradle (latter is required
anyway, so the plugin can know where it should map between forge and
fabric and which subprojects are which versions).
Since that necessitated some changes to the build.gradle file, I took
the opportunity to convert it to Kotlin. As such we now also use
Gradle's newer plugin-block with the pluginManagement-block instead of
manually declaring buildScript (not for the common.gradle though), which
imo is nicer anyway and comes with various advantages (see gradle docs).
There were also some remapping bugs fixed and some new ones
introduced (e.g. manually declared inner class mappings seem to no
longer function properly under certain conditions).
There's also support for remapping Kotlin now. Just saying.
For some reason hypixel is sending duplicate entity meta data entries
in their spawn packets.
We'll now be ignoring all but the last one for each id (should match
how vanilla handles these).
- Switch from ECJ to IDEA (with potential for kotlin)
- Can pull SRG mappings from FG, reducing amount of manual mappings (basically just class mappings and fixing errors in SRG mappings)
The preprocessor should have complained about these but it didn't for
some reason. This commit isn't fixing any imminent bug, it's just
precautionary (and required to compile once the preprocessor is fixed).
