草，怎么突然感觉最近闹得有点欢…… 很容易被拉清单的感觉

清单风险++

不过对于最后一条我补充一句，这张是为了讽刺效果这么填的（而且我也不觉得3,5,7张的大字全对）
我觉得美国在少数极端个例的平等以及和谐上还要下工夫，比如说经常有枪击案，以及小部分美国人对他国留学生不友好甚至不公正等问题

应该说美国自己要「优先」是可以理解，那也比背地里使阴招追求同样的目的但拿一堆正人君子的理由粉饰要好啊……

你说这个清单为什么不能把武汉失职的那些人都拉出来批批啊…… 为什么清单只拉无辜的异见民众啊
这都秋后了，该算账了

那个用来切分音频的考虑到要接触更多新东西有点费心，暂时不做了（而且不做也不是不能用 audacity 弄，理论上那样也可以用大致打了吐字轴的字幕辅助但 C / Scheme(Nyquist) 插件更困难……），说到底批量重命名程序很简单：

from shutil import move
for fp in argv[1:]:
move(fp, mapFilename(transform, fp))
这样而已…… 接下来考虑下星之所在，还是扒谱 Dream It Possible 好了，毕竟是华为的

https://www.youtube.com/watch?v=AzhePXOeWMg
不愧是华为，办事精甚细腻，让人拍案叫绝

天啊…… 现在我看『精甚细腻』『颐使气指』『金科律玉』『风狂雨骤』『发酵(xiao)』这些都找不出问题了，甚至有点习惯这些拼写方法了
我小学大概也是读错过赡养的吧？ 没有没有没有…… 其实这些也不算易错的感觉

$ youtube-dl --proxy http://localhost:12333 -f 251 'https://www.youtube.com/watch?v=AzhePXOeWMg'
[youtube] AzhePXOeWMg: Downloading webpage
[download] Destination: #DreamItPossible Karaoke-AzhePXOeWMg.webm
[download] 40.4% of 3.32MiB at 74.07KiB/s ETA 00:27
🌚 -f 251 ... 没办法啊

251 webm audio only tiny  151k , opus @160k (48000Hz), 3.92MiB 

这是音质最高的

[DuangSUSE@duangsuse]~/文档% ~/Projects/audio_to_midi_melodia/audio_to_midi_melodia.py DreamItPossible_Vocal.ogg DreamIt.mid
Loading audio...
Extracting melody f0 with MELODIA...
Converting Hz to MIDI notes...
Saving MIDI to disk...
Conversion complete.
用上高科技自动扒谱了…… 不过感觉也没那么好，还是只能辅助拿音高

I will run I will climb I will so -o -ar
I'm un de -feat -ed o o oh
Jum -ping out of my skin, pull the ch -o -rd
Yeah I be -lieve it o oh
The past, is e -very thing we were Don't make us who we are

[60, 58, 60, 60, 58, 60, 60, 58, 60, 68, 67, 58, 58, 55, 55, 55, 55, 56, 55, 60, 58, 60, 60, 58, 60, 60, 58, 60, 68, 67, 58, 58, 55, 55, 55, 58, 60, 63, 65, 63, 65, 67, 65, 63, 65, 63, 65, 67, 65, 63, 63, 62, 63]

So I'll dream, until I make it real
And all I see is stars
It's not until you fall that you fly

When your dreams comes alive you're unstoppable
Take a shot, chase the
Sun, find the beautiful
We will glow in the dark
Turning dust to gold
And we'll dream
It possible
Possible
And we'll dream
It possible
I will chase, I will reach, I will fly
Until I'm breaking, until I'm breaking
Out of my cage, like a bird in the night
I know I'm changing, I know I'm changing
In, into something big,
Better than before
And if it takes, takes a thousand lives
Then it's worth fighting for
It's not until you fall that you fly
When your dreams come
Alive you're unstoppable
Take a shot, chase the
Sun, find the beautiful
We will glow in the dark
Turning dust to gold
And we'll dream
It possible
Possible
From the bottom to the top
We're sparking wild fire's
Never quit and never stop
The rest of our lives
From the bottom to the top
We're sparking wild fire's
Never quit and never stop
It's not until you fall that you fly
When your dreams come
Alive you're unstoppable
Take a shot, chase the
Sun, find the beautiful
We will glow in the dark
Turning dust to gold
And we'll dream
It possible
Possible
And we'll dream
It possible

……Oh 我最讨厌扒这种很好听的歌了，显然要想扒谱熟悉歌词是基本要求，但对一个 native speaker 而言显然那样更容易处理这个断词的问题……

显然为了避免再次返工，我们这次必须得先处理好断词的问题…… 毕竟这个和普通的歌词还是有很大区别，我是按音高分音符(所以一个词也可以占多音符)的

undefated 这个词 synthv 直接读起来有点怪，还是按中文的思路拆成 un -defated 两个音好了

第一次很失败…… 这支曲子很多变，突破了我对英文的猜测，总是不能按自己之前策划的操作，因为我只认音高…… 显然这样的话音高表也得重弄

……终于知道 MuseScore 里怎么进行基本的顺序编辑了，只能剪切粘贴啊

#linux #science #tools 🤔 可编程、录制的星图软件， 里面 Sun 叫 Sol, 估计是巴西的后端

from re import split
def template(name, verb, text):
def subst(word):
if word == name: return "%1"
elif word == verb: return "%2"
else: return word #< can be a dict getOr
return "".join(map(subst, split(f"({name}|{verb})", text)))

from random import shuffle
def refill(templ, name, verb, sep = "。"):
corpus = shuffle(sep.split(templ))
sep.join(lambda part: part %(name, verb) for sentence in corpus) #Python #code

&eval=renderTextElement=(p,f,t)=>element("a",configured(withPoint(...p),withFont(...f),withText(t)))
现在可以直接魔改渲染脚本