Stardust
I just want to write codes in the best way.
LaTexDSL
-
Requirements:
- MiKTeX 2.9.6300
- PyInstaller for cpy3
- python
Just see how the source codes transformed to the target...
- test.tex
\documentclass[UTF8]{ctexart}
\usepackage{amsmath,amssymb,latexsym}
\begin{document}
@:func a b[ r"{a}\left( {b} \right)".format(a=a,b=b)]
@:div a b [r"{a}/{b}".format(a=a,b=b)]
\noindent proof 1.1 Exercise (1) turn
$\forall y \in func f A \cap func f B $, turn
$ y \in func f A \wedge y \in func f B \rightarrow$
$$ \exists x_1 \in A , func f x_1 = y \wedge \exists x_2 \in B , func f x_2 =y , \rightarrow $$
$$ x=x_1=x_2 \rightarrow x \in A \cup B \rightarrow y = func f x \in func f <<A \cup B>> $$
[]
turn
turn
\noindent proof 1.1 Exercise (2) turn
$\forall y \in div <<func f X>> <<func f A>> $, turn
$y \in func f X \wedge y \notin func f A $, turn
$\exists x_1 \in X , func f x_1 = y \wedge \forall x \in func f A, func f x \not= y \rightarrow$,
$$y = func f <<\overline x>>, x \in X \wedge x \notin A \rightarrow $$
$$y \in func f <<div X A>> $$
[]
- target.tex
\documentclass[UTF8]{ctexart}
\usepackage{amsmath,amssymb,latexsym}
\begin{document}
\noindent proof 1.1 Exercise (1) \\
$\forall y \in f\left( A \right) \cap f\left( B \right) $, \\
$ y \in f\left( A \right) \wedge y \in f\left( B \right) \rightarrow$
$$ \exists x_1 \in A , f\left( x_1 \right) = y \wedge \exists x_2 \in B , f\left( x_2 \right) =y , \rightarrow $$
$$ x=x_1=x_2 \rightarrow x \in A \cup B \rightarrow y = f\left( x \right) \in f\left( A \cup B \right) $$
[]
\\
\\
\noindent proof 1.1 Exercise (2) \\
$\forall y \in f\left( X \right)/f\left( A \right) $, \\
$y \in f\left( X \right) \wedge y \notin f\left( A \right) $, \\
$\exists x_1 \in X , f\left( x_1 \right) = y \wedge \forall x \in f\left( A \right), f\left( x \right) \not= y \rightarrow$,
$$y = f\left( \overline x \right), x \in X \wedge x \notin A \rightarrow $$
$$y \in f\left( X/A \right) $$
[]
\end{document}
Uh... I think it help a lot with my homework...
Just pack startex.py to .exe in Windows, and then
startex <tagfile> <outputfile>
Pattern Matching
It must be boring to be involved with the arguments like "which is the best programming language?".
For sure it seems most likely that the answer is Haskell.
- Guard:
sampleFunc:: Int->Int->String
sampleFunc x y
| x<10 = "digit"
| (x>10)&&(y<20) = "teens"
| (x%10==0) = "tens"
| otherwise = "tens and digit"
But in some degree, it's not practicable for me to use Haskell as my main weapon, as the result of there're quite many comprehensive libraries writing in Python available.
Also, using Python can be somewhat a tradition in Scientific Researching.
Uh, maybe the most important point is that I'm not clever at all. And I does have a coding style with both OO and FP, which is not pure at all.
But I cannot live without pattern matching yet...
So..
>>patMatch({1,2,3},{1,2,3},partial=False)
>>True
>>patMatch({1,2},{1,2,3})
>>True
>>patMatch({1,2},{1,2,3},partial=False)
>>False
>>patMatch([1,2,Any(int)],[1,2,3])
>>True
# and more examples can be given.
patMatch([1,2,Seq(int,atleast=2),0.5],[1,2,3,10,0.5]) #->True
patMatch([1,2,Seq(float,atleast=2),0.5],[1,2,3,10,0.5]) #->True
patMatch((1,2,Seq(float)),[1,2]) # -> True
patMatch((1,2,Seq(float,atleast=2)),[1,2]) # -> False
patMacth([[1,2,3],[Seq(int),[Seq(int)]]],[[1,2,3],[1,[1]]] ) #->True
patMatch([Any(dict)],[dict(a=[1,2,3],b=[2,3,4])]) # ->True
dictionary= {'a':1,'b':20}
patMatch(dict(a=Any(int),b=20),dictionary) #-> True
class sampleClass:
def __init__(self,a,b,c):
self.a=a
self.b=b
self.c=c
def func(self):
dosomething
instance=sampleClass(1,15,20)
patMatch(sampleClass(1,Any(int),Any(int)),instance) # -> TrueUse this module conveniently:
from Stardust import *
matcher0=PM([1,"2",2])
print (matcher.match([1,Any(str),2])) #->True
matcher1=PM([1,2,3,4,""])
print (matcher2.match([1,Seq(int),""])) #-> True
class sample:
def __init__(self,a,b,c):
self.a=a
self.b=b
self.c=c
def dosome(self):pass
matcher2=PM([[sample(1,2,3),dict(a=2,c=5,d=7)],[],[]])
matcher2.match([[sample(Any(),*[Any(int)]*2,dict(a=2)],Seq(list)],partial=True) #->TrueAnd more examples can be found in runsamples.py
Welcome to find out the bugs in this tiny library.
If you use Python and you want to code in a more rational style, hope you can benefit from my works.
Strict Type in Python
If you prefer languages which is more likely to do type checking and you have to write python sometimes, try this.
@strict.args(int,int)
@strict.ret(float)
def f(a,b):
return a*b*1.0
>> f(1.0, 2.0)
TypeError: Type of argnument 0 should <class 'int'>
>> f(1, 2)
>> 2.0
@strict.args(str, int , name = str)
@strict.ret(str)
def repeat(a, b, name = '<unknown>' ):
print(f'in closure : {name}')
return a*b
>> repeat("2", 3)
in closure : <unknown>
>> '222'
>> repeat("2","2")
TypeError: Type of argnument 1 should <class 'int'>And there're many nice properties to use the strict you can find it for yourself.
The source codes is tiny, you can see it here
Making some extensions based on it might be something very very interesting!
TCO in CPython
See tco.ipynb.
import functools
from collections import namedtuple
param_struct = namedtuple('param_struct', ['args', 'kwargs'])
def tco(func):
def call(*args, **kwargs):
try:
old = func.__globals__[func.__name__]
func.__globals__[func.__name__] = lambda *args, **kwargs : param_struct(args, kwargs)
res = param_struct(args, kwargs)
while True:
res = func(*res.args, **res.kwargs)
if isinstance(res, param_struct):
continue
break
func.__globals__[func.__name__] = old
return res
except Exception as e:
func.__globals__[func.__name__] = old
raise e
return callI'm so glad to see that tail call optimization could be put into actual use in Python.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent