th.py

# -*- coding: utf-8 -*-
from threading import Thread,Lock,Event,Condition
import multiprocessing as mlti

class orth:#(th.Thread):
    lock=Lock()
orEv=Event()
def thpr(*a):
    with orth.lock:
        print(a,end="-")
def run(cb,args):
    t=Thread(target=cb,args=args)
    t.start()
    return t
def runi(sl,cb,data,mod,vl):
    liImz=[Thread(target=cb, args=(vl,x,data,[])) for x in range(mod)]
    for j in liImz: j.start()
    return liImz
def thrun(thFind,iList,nm,ars):
    "thFind(ars,sepnm,iList,letlist)"
    #l=list.split(nm)
    #(len(letlist)*i)/nm
    #separater_span_block_colon_range_窓関数_
    sep=len(iList)/nm
    sepnm=int(sep+1)#+0.5)
    #sepnm=len(iList)//nm+1#floordivision
    letlist=[]#[[] for i in range(nm)]
    liImz=[]
    for i in range(nm):
        #th
        letlist.append([])
        if not(sepnm*(i+1)>=len(iList)):
            rang=(sepnm*i,sepnm*(i+1))
            imz=Thread(target=thFind,args=(ars,rang,iList,letlist[i]))
            liImz.append(imz)
        else:
            rang=(sepnm*i,len(iList))
            imz=Thread(target=thFind,args=(ars,rang,iList,letlist[i]))
            liImz.append(imz)
            break
    for f in liImz: f.start()
    for f in liImz:
        res=getattr(f,b'\x6A\x6F\x69\x6E'.decode())()
    return letlist
def threadAny(thFind,iList,ars):
    "thFind(ars,sepnm,iList,letlist)"
    letlist=[[] for i in range(len(iList))]
    liImz=[]
    for i,j in enumerate(iList):
        def crProc(proc):
            def f1(*a):
                proc(*a)
                orEv.set()
            return f1
        imz=Thread(target=crProc(thFind),args=(ars,j,letlist[i]))
        liImz.append(imz)
    for j in liImz: j.start()
    for i,j in enumerate(liImz):
        orEv.clear()#block<>set() green light
        n=orEv.wait(timeout=5)
        #call set() elsewhere
        if not n:return 0
        rn1=[j for j in letlist if j][0]
    return rn1





import concurrent
from concurrent.futures import ProcessPoolExecutor, ThreadPoolExecutor
def pRun(Cb,list,th=20,fil=()):
    executor = concurrent.futures.ProcessPoolExecutor(max_workers=nm)
    futures = [executor.submit(Cb,t) for t in list]#start


    # as_completed()は、与えられたfuturesの要素を完了順にたどるイテレータを返す。
    for future in concurrent.futures.as_completed(futures):# block
        print(future.result()) # digest()の戻り値が表示される。
    # (as_completedをすべてイテレートしているので、実際にはこの時点で完了していないスレッドは無いはず。)
    executor.shutdown()
#thpool
def pool(cb,urls,th=20,ve=(),pl="t"):
    li2=[None]*len(urls)
    pl={"t":ThreadPoolExecutor, "p":ProcessPoolExecutor}[pl]
    with pl(max_workers=th) as executor:
        future_to_url = {executor.submit(cb, url,*ve): x for x,url in enumerate(urls)}
        for future in concurrent.futures.as_completed(future_to_url):
            x = future_to_url[future]
            li2[x]=future.result()
    return li2
##print(thpe(lambda ab,cd:ab+"23",["A","B"],23,["CD"]))







def main():#incremental search
    import time
    def timeo(duration=0.2):
        def fi(te="something"):
            for j in te:
                time.sleep(duration)
                yield j
                newflag=0
                if(newflag):#cansel this thread
                    return 2
            return 1
        return fi
    global pro,pr1;
    def pro(conn,cb,ars,li):
        for j in cb(ars):
            conn.send(j)
            ri=conn.recv()
            print(ri)
    pr1=timeo()
    #
    ps=[]
    for j in "inc":
        pare, child = mlti.Pipe()
        ps.append(( mlti.Process(target=pro, args=(child,1234,(j*9),None)),pare,child))
        #cant pickle local callback
    for p in ps:
        p[0].start()
        time.sleep(0.5)
##    print pare.recv()   # prints "[23, None, 'hello']"
##    for p in ps:        jw(p,)

if __name__ == '__main__':
##    main()
    from multiprocessing import shared_memory
##    m=mlti.shared_memory.SharedMemory("fil",True,1)
##    m=mlti.shared_memory.SharedMemory("fil",False,1)
##    print(m)
    import mmap
    m=mmap.mmap(-1,3,"fil")
    print(m.tell(),m.read())