mirror of
https://github.com/TriM-Organization/Musicreater.git
synced 2025-06-05 03:45:23 +00:00
632 lines
22 KiB
Python
632 lines
22 KiB
Python
# -*- coding: utf-8 -*-
|
||
"""
|
||
新版本功能以及即将启用的函数
|
||
"""
|
||
|
||
|
||
"""
|
||
版权所有 © 2023 音·创 开发者
|
||
Copyright © 2023 all the developers of Musicreater
|
||
|
||
开源相关声明请见 仓库根目录下的 License.md
|
||
Terms & Conditions: License.md in the root directory
|
||
"""
|
||
|
||
# 睿乐组织 开发交流群 861684859
|
||
# Email TriM-Organization@hotmail.com
|
||
# 若需转载或借鉴 许可声明请查看仓库目录下的 License.md
|
||
|
||
import random
|
||
|
||
from .constants import INSTRUMENT_BLOCKS_TABLE
|
||
from .exceptions import *
|
||
from .main import MidiConvert
|
||
from .subclass import *
|
||
from .utils import *
|
||
from .types import Tuple, List, Dict
|
||
|
||
class FutureMidiConvertRSNB(MidiConvert):
|
||
"""
|
||
加入红石音乐适配
|
||
"""
|
||
|
||
music_command_list: Dict[int, SingleNoteBox]
|
||
"""音乐指令列表"""
|
||
|
||
@staticmethod
|
||
def soundID_to_block(sound_id: str, random_select: bool = False) -> str:
|
||
"""
|
||
将我的世界乐器名改作音符盒所需的对应方块名称
|
||
|
||
Parameters
|
||
----------
|
||
sound_id: str
|
||
将我的世界乐器名
|
||
random_select: bool
|
||
是否随机选取对应方块
|
||
|
||
Returns
|
||
-------
|
||
str方块名称
|
||
"""
|
||
try:
|
||
if random_select:
|
||
return random.choice(INSTRUMENT_BLOCKS_TABLE[sound_id])
|
||
else:
|
||
return INSTRUMENT_BLOCKS_TABLE[sound_id][0]
|
||
except KeyError:
|
||
return "air"
|
||
|
||
def to_note_list_in_delay(
|
||
self,
|
||
) -> Tuple[Dict[int, SingleNoteBox], int]:
|
||
"""
|
||
使用金羿的转换思路，将midi转换为我的世界音符盒音高、乐器及延迟列表，并输出每个音符之后的延迟
|
||
|
||
Returns
|
||
-------
|
||
tuple( Dict[int, SingleNoteBox], int音乐时长游戏刻 )
|
||
"""
|
||
|
||
self.to_music_channels()
|
||
|
||
tracks = {}
|
||
note_range = {}
|
||
InstID = -1
|
||
# cmd_amount = 0
|
||
|
||
# 此处 我们把通道视为音轨
|
||
for i in self.channels.keys():
|
||
# 如果当前通道为空 则跳过
|
||
if not self.channels[i]:
|
||
continue
|
||
|
||
# 第十通道是打击乐通道
|
||
SpecialBits = True if i == 9 else False
|
||
|
||
for track_no, track in self.channels[i].items():
|
||
for msg in track:
|
||
if msg[0] == "PgmC":
|
||
InstID = msg[1]
|
||
|
||
elif msg[0] == "NoteS":
|
||
# block_id = self.soundID_to_block((
|
||
# self.perc_inst_to_soundID_withX(msg[1])
|
||
# if SpecialBits
|
||
# else self.inst_to_souldID_withX(InstID)
|
||
# )[0])
|
||
|
||
# delaytime_now = round(msg[-1] / 50)
|
||
|
||
note_ = 2 ** ((msg[1] - 60) / 12)
|
||
|
||
try:
|
||
tracks[msg[-1]].append(
|
||
(
|
||
InstID,
|
||
note_,
|
||
)
|
||
)
|
||
except KeyError:
|
||
tracks[msg[-1]] = [(InstID, note_)]
|
||
|
||
try:
|
||
note_range[InstID]["max"] = max(
|
||
note_range[InstID]["max"], note_
|
||
)
|
||
note_range[InstID]["min"] = min(
|
||
note_range[InstID]["min"], note_
|
||
)
|
||
except KeyError:
|
||
note_range[InstID] = {"max": note_, "min": note_}
|
||
|
||
del InstID
|
||
all_ticks = list(tracks.keys())
|
||
all_ticks.sort()
|
||
results = []
|
||
print(note_range)
|
||
|
||
exit()
|
||
|
||
for i in range(len(all_ticks)):
|
||
for j in range(len(tracks[all_ticks[i]])):
|
||
results.append(
|
||
SingleCommand(
|
||
tracks[all_ticks[i]][j],
|
||
tick_delay=(
|
||
0
|
||
if j != 0
|
||
else (
|
||
all_ticks[i] - all_ticks[i - 1]
|
||
if i != 0
|
||
else all_ticks[i]
|
||
)
|
||
),
|
||
annotation="在{}播放{}%的{}音".format(
|
||
mctick2timestr(i), max_volume * 100, ""
|
||
),
|
||
)
|
||
)
|
||
|
||
self.music_command_list = results
|
||
self.music_tick_num = max(all_ticks)
|
||
return results, self.music_tick_num
|
||
|
||
|
||
class FutureMidiConvertM4(MidiConvert):
|
||
"""
|
||
加入插值算法优化音感
|
||
: 经测试，生成效果已经达到，感觉良好
|
||
"""
|
||
|
||
# 临时用的插值计算函数
|
||
@staticmethod
|
||
def _linear_note(
|
||
_note: SingleNote,
|
||
_apply_time_division: int = 100,
|
||
) -> List[Tuple[int, int, int, int, float],]:
|
||
"""传入音符数据，返回以半秒为分割的插值列表
|
||
:param _note: SingleNote 音符
|
||
:return list[tuple(int开始时间（毫秒）, int乐器, int音符, int力度（内置）, float音量（播放）),]"""
|
||
|
||
totalCount = int(_note.duration / _apply_time_division)
|
||
|
||
if totalCount == 0:
|
||
return [
|
||
(_note.start_time, _note.inst, _note.pitch, _note.velocity, 1),
|
||
]
|
||
# print(totalCount)
|
||
|
||
result: List[Tuple[int, int, int, int, float],] = []
|
||
|
||
for _i in range(totalCount):
|
||
result.append(
|
||
(
|
||
_note.start_time + _i * _apply_time_division,
|
||
_note.instrument,
|
||
_note.pitch,
|
||
_note.velocity,
|
||
((totalCount - _i) / totalCount),
|
||
)
|
||
)
|
||
|
||
return result
|
||
|
||
# 简单的单音填充
|
||
def to_command_list_in_score(
|
||
self,
|
||
scoreboard_name: str = "mscplay",
|
||
max_volume: float = 1.0,
|
||
speed: float = 1.0,
|
||
) -> Tuple[List[List[SingleCommand]], int, int]:
|
||
"""
|
||
使用金羿的转换思路，使用完全填充算法优化音感后，将midi转换为我的世界命令列表
|
||
|
||
Parameters
|
||
----------
|
||
scoreboard_name: str
|
||
我的世界的计分板名称
|
||
max_volume: float
|
||
最大播放音量，注意：这里的音量范围为(0,1]，如果超出将被处理为正确值，其原理为在距离玩家 (1 / volume -1) 的地方播放
|
||
speed: float
|
||
速度，注意：这里的速度指的是播放倍率，其原理为在播放音频的时候，每个音符的播放时间除以 speed
|
||
|
||
Returns
|
||
-------
|
||
tuple( list[list[SingleCommand指令,... ],... ], int指令数量, int音乐时长游戏刻 )
|
||
"""
|
||
|
||
if speed == 0:
|
||
raise ZeroSpeedError("播放速度仅可为正实数")
|
||
max_volume = 1 if max_volume > 1 else (0.001 if max_volume <= 0 else max_volume)
|
||
|
||
self.to_music_channels()
|
||
|
||
note_channels: Dict[int, List[SingleNote]] = empty_midi_channels(staff=[])
|
||
InstID = -1
|
||
|
||
# 此处 我们把通道视为音轨
|
||
for i in self.channels.keys():
|
||
# 如果当前通道为空 则跳过
|
||
if not self.channels[i]:
|
||
continue
|
||
|
||
# nowChannel = []
|
||
for track_no, track in self.channels[i].items():
|
||
noteMsgs = []
|
||
MsgIndex = []
|
||
|
||
for msg in track:
|
||
if msg[0] == "PgmC":
|
||
InstID = msg[1]
|
||
|
||
elif msg[0] == "NoteS":
|
||
noteMsgs.append(msg[1:])
|
||
MsgIndex.append(msg[1])
|
||
|
||
elif msg[0] == "NoteE":
|
||
if msg[1] in MsgIndex:
|
||
note_channels[i].append(
|
||
SingleNote(
|
||
InstID,
|
||
msg[1],
|
||
noteMsgs[MsgIndex.index(msg[1])][1],
|
||
noteMsgs[MsgIndex.index(msg[1])][2],
|
||
msg[-1] - noteMsgs[MsgIndex.index(msg[1])][2],
|
||
track_number=track_no,
|
||
)
|
||
)
|
||
noteMsgs.pop(MsgIndex.index(msg[1]))
|
||
MsgIndex.pop(MsgIndex.index(msg[1]))
|
||
|
||
del InstID
|
||
|
||
tracks = []
|
||
cmd_amount = 0
|
||
max_score = 0
|
||
|
||
# 此处 我们把通道视为音轨
|
||
for no, track in note_channels.items():
|
||
# 如果当前通道为空 则跳过
|
||
if not track:
|
||
continue
|
||
|
||
SpecialBits = True if no == 9 else False
|
||
|
||
track_now = []
|
||
|
||
for note in track:
|
||
for every_note in self._linear_note(
|
||
note, 100 if note.track_no == 0 else 500
|
||
):
|
||
soundID, _X = (
|
||
self.perc_inst_to_soundID_withX(note.pitch)
|
||
if SpecialBits
|
||
else self.inst_to_souldID_withX(note.inst)
|
||
)
|
||
|
||
score_now = round(every_note[0] / speed / 50)
|
||
|
||
max_score = max(max_score, score_now)
|
||
mc_pitch = 2 ** ((note.pitch - 60 - _X) / 12)
|
||
blockmeter = (
|
||
1
|
||
/ (1 if note.track_no == 0 else 0.9)
|
||
/ max_volume
|
||
/ every_note[4]
|
||
- 1
|
||
)
|
||
|
||
track_now.append(
|
||
SingleCommand(
|
||
self.execute_cmd_head.format(
|
||
"@a[scores=({}={})]".format(scoreboard_name, score_now)
|
||
.replace("(", r"{")
|
||
.replace(")", r"}")
|
||
)
|
||
+ "playsound {} @s ^ ^ ^{} {} {}".format(
|
||
soundID,
|
||
blockmeter,
|
||
note.velocity / 128,
|
||
"" if SpecialBits else mc_pitch,
|
||
),
|
||
annotation="在{}播放{}%({}BM)的{}音".format(
|
||
mctick2timestr(score_now),
|
||
max_volume * 100,
|
||
blockmeter,
|
||
"{}:{}".format(soundID, note.pitch),
|
||
),
|
||
),
|
||
)
|
||
|
||
cmd_amount += 1
|
||
|
||
if track_now:
|
||
self.music_command_list.extend(track_now)
|
||
tracks.append(track_now)
|
||
|
||
self.music_tick_num = max_score
|
||
return (tracks, cmd_amount, max_score)
|
||
|
||
# 简单的单音填充的延迟应用
|
||
def to_command_list_in_delay(
|
||
self,
|
||
max_volume: float = 1.0,
|
||
speed: float = 1.0,
|
||
player_selector: str = "@a",
|
||
) -> Tuple[List[SingleCommand], int, int]:
|
||
"""
|
||
使用金羿的转换思路，使用完全填充算法优化音感后，将midi转换为我的世界命令列表，并输出每个音符之后的延迟
|
||
|
||
Parameters
|
||
----------
|
||
max_volume: float
|
||
最大播放音量，注意：这里的音量范围为(0,1]，如果超出将被处理为正确值，其原理为在距离玩家 (1 / volume -1) 的地方播放音频
|
||
speed: float
|
||
速度，注意：这里的速度指的是播放倍率，其原理为在播放音频的时候，每个音符的播放时间除以 speed
|
||
player_selector: str
|
||
玩家选择器，默认为`@a`
|
||
|
||
Returns
|
||
-------
|
||
tuple( list[SingleCommand,...], int音乐时长游戏刻, int最大同时播放的指令数量 )
|
||
"""
|
||
|
||
if speed == 0:
|
||
raise ZeroSpeedError("播放速度仅可为正实数")
|
||
max_volume = 1 if max_volume > 1 else (0.001 if max_volume <= 0 else max_volume)
|
||
|
||
self.to_music_channels()
|
||
|
||
note_channels: Dict[int, List[SingleNote]] = empty_midi_channels(staff=[])
|
||
InstID = -1
|
||
|
||
# 此处 我们把通道视为音轨
|
||
for i in self.channels.keys():
|
||
# 如果当前通道为空 则跳过
|
||
if not self.channels[i]:
|
||
continue
|
||
|
||
# nowChannel = []
|
||
for track_no, track in self.channels[i].items():
|
||
noteMsgs = []
|
||
MsgIndex = []
|
||
|
||
for msg in track:
|
||
if msg[0] == "PgmC":
|
||
InstID = msg[1]
|
||
|
||
elif msg[0] == "NoteS":
|
||
noteMsgs.append(msg[1:])
|
||
MsgIndex.append(msg[1])
|
||
|
||
elif msg[0] == "NoteE":
|
||
if msg[1] in MsgIndex:
|
||
note_channels[i].append(
|
||
SingleNote(
|
||
InstID,
|
||
msg[1],
|
||
noteMsgs[MsgIndex.index(msg[1])][1],
|
||
noteMsgs[MsgIndex.index(msg[1])][2],
|
||
msg[-1] - noteMsgs[MsgIndex.index(msg[1])][2],
|
||
track_number=track_no,
|
||
)
|
||
)
|
||
noteMsgs.pop(MsgIndex.index(msg[1]))
|
||
MsgIndex.pop(MsgIndex.index(msg[1]))
|
||
|
||
del InstID
|
||
|
||
tracks = {}
|
||
InstID = -1
|
||
# open("RES.TXT", "w", encoding="utf-8").write(str(note_channels))
|
||
|
||
# 此处 我们把通道视为音轨
|
||
for no, track in note_channels.items():
|
||
# 如果当前通道为空 则跳过
|
||
if not track:
|
||
continue
|
||
|
||
SpecialBits = True if no == 9 else False
|
||
|
||
for note in track:
|
||
liner_list = self._linear_note(note, 100 if note.track_no == 0 else 500)
|
||
for every_note in liner_list:
|
||
soundID, _X = (
|
||
self.perc_inst_to_soundID_withX(note.pitch)
|
||
if SpecialBits
|
||
else self.inst_to_souldID_withX(note.inst)
|
||
)
|
||
|
||
score_now = round(every_note[0] / speed / 50)
|
||
|
||
try:
|
||
tracks[score_now].append(
|
||
self.execute_cmd_head.format(player_selector)
|
||
+ f"playsound {soundID} @s ^ ^ ^{1 / (1 if note.track_no == 0 else 0.9) / max_volume / every_note[4] - 1} {note.velocity / 128} "
|
||
+ (
|
||
""
|
||
if SpecialBits
|
||
else f"{2 ** ((note.pitch - 60 - _X) / 12)}"
|
||
)
|
||
)
|
||
except KeyError:
|
||
tracks[score_now] = [
|
||
self.execute_cmd_head.format(player_selector)
|
||
+ f"playsound {soundID} @s ^ ^ ^{1 / (1 if note.track_no == 0 else 0.9) / max_volume / every_note[4] - 1} {note.velocity / 128} "
|
||
+ (
|
||
""
|
||
if SpecialBits
|
||
else f"{2 ** ((note.pitch - 60 - _X) / 12)}"
|
||
)
|
||
]
|
||
|
||
all_ticks = list(tracks.keys())
|
||
all_ticks.sort()
|
||
results: List[SingleCommand] = []
|
||
max_multi = 0
|
||
now_multi_delay = 0
|
||
now_multi = 0
|
||
|
||
for i in range(len(all_ticks)):
|
||
l = len(tracks[all_ticks[i]])
|
||
for j in range(l):
|
||
results.append(
|
||
SingleCommand(
|
||
tracks[all_ticks[i]][j],
|
||
tick_delay=(
|
||
0
|
||
if j != 0
|
||
else (
|
||
all_ticks[i] - all_ticks[i - 1]
|
||
if i != 0
|
||
else all_ticks[i]
|
||
)
|
||
),
|
||
annotation="由 音·创 生成",
|
||
)
|
||
)
|
||
if results[-1].delay + now_multi_delay <= 1:
|
||
now_multi += 1
|
||
now_multi_delay += results[-1].delay
|
||
else:
|
||
max_multi = max(max_multi, now_multi)
|
||
now_multi = 0
|
||
now_multi_delay = 0
|
||
|
||
self.music_command_list = results
|
||
self.music_tick_num = max(all_ticks)
|
||
return results, self.music_tick_num, max_multi
|
||
|
||
|
||
class FutureMidiConvertM5(MidiConvert):
|
||
"""
|
||
加入同刻偏移算法优化音感
|
||
"""
|
||
|
||
# 神奇的偏移音
|
||
def to_command_list_in_delay(
|
||
self,
|
||
max_volume: float = 1.0,
|
||
speed: float = 1.0,
|
||
player_selector: str = "@a",
|
||
) -> Tuple[List[SingleCommand], int]:
|
||
"""
|
||
使用金羿的转换思路，使用同刻偏移算法优化音感后，将midi转换为我的世界命令列表，并输出每个音符之后的延迟
|
||
|
||
Parameters
|
||
----------
|
||
max_volume: float
|
||
最大播放音量，注意：这里的音量范围为(0,1]，如果超出将被处理为正确值，其原理为在距离玩家 (1 / volume -1) 的地方播放音频
|
||
speed: float
|
||
速度，注意：这里的速度指的是播放倍率，其原理为在播放音频的时候，每个音符的播放时间除以 speed
|
||
player_selector: str
|
||
玩家选择器，默认为`@a`
|
||
|
||
Returns
|
||
-------
|
||
tuple( list[SingleCommand,...], int音乐时长游戏刻 )
|
||
"""
|
||
|
||
if speed == 0:
|
||
raise ZeroSpeedError("播放速度仅可为正实数")
|
||
max_volume = 1 if max_volume > 1 else (0.001 if max_volume <= 0 else max_volume)
|
||
|
||
self.to_music_channels()
|
||
|
||
tracks = {}
|
||
InstID = -1
|
||
|
||
# 此处 我们把通道视为音轨
|
||
for i in self.channels.keys():
|
||
# 如果当前通道为空 则跳过
|
||
if not self.channels[i]:
|
||
continue
|
||
|
||
# 第十通道是打击乐通道
|
||
SpecialBits = True if i == 9 else False
|
||
|
||
# nowChannel = []
|
||
|
||
for track_no, track in self.channels[i].items():
|
||
for msg in track:
|
||
if msg[0] == "PgmC":
|
||
InstID = msg[1]
|
||
|
||
elif msg[0] == "NoteS":
|
||
soundID, _X = (
|
||
self.perc_inst_to_soundID_withX(msg[1])
|
||
if SpecialBits
|
||
else self.inst_to_souldID_withX(InstID)
|
||
)
|
||
|
||
score_now = round(msg[-1] / float(speed) / 50)
|
||
# print(score_now)
|
||
|
||
try:
|
||
tracks[score_now].append(
|
||
self.execute_cmd_head.format(player_selector)
|
||
+ f"playsound {soundID} @s ^ ^ ^{128 / max_volume / msg[2] - 1} {msg[2] / 128} "
|
||
+ (
|
||
""
|
||
if SpecialBits
|
||
else f"{2 ** ((msg[1] - 60 - _X) / 12)}"
|
||
)
|
||
)
|
||
except KeyError:
|
||
tracks[score_now] = [
|
||
self.execute_cmd_head.format(player_selector)
|
||
+ f"playsound {soundID} @s ^ ^ ^{128 / max_volume / msg[2] - 1} {msg[2] / 128} "
|
||
+ (
|
||
""
|
||
if SpecialBits
|
||
else f"{2 ** ((msg[1] - 60 - _X) / 12)}"
|
||
)
|
||
]
|
||
|
||
all_ticks = list(tracks.keys())
|
||
all_ticks.sort()
|
||
results = []
|
||
|
||
for i in range(len(all_ticks)):
|
||
for j in range(len(tracks[all_ticks[i]])):
|
||
results.append(
|
||
SingleCommand(
|
||
tracks[all_ticks[i]][j],
|
||
tick_delay=(
|
||
(
|
||
0
|
||
if (
|
||
(all_ticks[i + 1] - all_ticks[i])
|
||
/ len(tracks[all_ticks[i]])
|
||
< 1
|
||
)
|
||
else 1
|
||
)
|
||
if j != 0
|
||
else (
|
||
(
|
||
all_ticks[i]
|
||
- all_ticks[i - 1]
|
||
- (
|
||
0
|
||
if (
|
||
(all_ticks[i] - all_ticks[i - 1])
|
||
/ len(tracks[all_ticks[i - 1]])
|
||
< 1
|
||
)
|
||
else (len(tracks[all_ticks[i - 1]]) - 1)
|
||
)
|
||
)
|
||
if i != 0
|
||
else all_ticks[i]
|
||
)
|
||
),
|
||
annotation="在{}播放{}%的{}音".format(
|
||
mctick2timestr(
|
||
i + 0
|
||
if (
|
||
(all_ticks[i + 1] - all_ticks[i])
|
||
/ len(tracks[all_ticks[i]])
|
||
< 1
|
||
)
|
||
else j
|
||
),
|
||
max_volume * 100,
|
||
"",
|
||
),
|
||
)
|
||
)
|
||
|
||
self.music_command_list = results
|
||
self.music_tick_num = max(all_ticks)
|
||
return results, self.music_tick_num
|
||
|
||
|
||
class FutureMidiConvertM6(MidiConvert):
|
||
"""
|
||
加入插值算法优化音感，但仅用于第一音轨
|
||
"""
|
||
|
||
# TODO 没写完的！！！！
|