Twoim zadaniem, jeśli nie chcesz odmówić, jest napisanie programu, który „golfa” fragment danego kodu w wybranym języku.

Podstawowa funkcjonalność:

• Usuń zewnętrzne białe znaki
• Usuń komentarze

„Zaawansowana” funkcjonalność:

• Połącz deklaracje ( int x; int y; int z;do int x, y, z;)
• Używaj krótszych odmian podstawowych struktur kontrolnych ( while(1)zastępuje się for(;;))
• Przenieś kod ( i = 0; while(i < 10) { /* ... */ ++i; }do for(i = 0; i < 10; ++i) { /* ... */ })
• Itp

Python z Python

Robi mnóstwo rzeczy, w tym zmianę nazw zmiennych, usuwanie niepotrzebnych białych znaków i komentarzy oraz umieszczanie jak najwięcej w jednym wierszu. Nie zawsze działa całkowicie z bardziej zaawansowaną składnią Pythona i będę kontynuować aktualizację z wszelkimi poprawkami.

Kod:

import string
import keyword
import pkgutil

builtins = __builtins__.__dict__.keys()

vars = {}
#imported = builtins+string.__dict__.keys()+['append','extend','count','index','insert','pop','remove','reverse','sort']
multiline = ''
ml_last = ''
strings = []
defined = []
undefined = []

def get_name(name):
    if name.startswith('__'):
        vars[name] = name
        return name
    if name in vars:
        return vars[name]

    for c in string.letters+'_':
        if c not in vars.values():
            vars[name] = c
            return c

    for c0 in string.letters+'_':
        for c1 in string.letters+string.digits+'_':
            if c0+c1 not in vars.values():
                if c0+c1 in keyword.kwlist:
                    continue
                vars[name] = c0+c1
                return c0+c1

def replace_names(expr,defining=False,prefix = '',assign=True):
    if ';' in expr:
        ns = ''
        for e in expr.split(';'):
            ns += replace_names(e,assign=assign)+';'
        return ns[:-1]

    global multiline
    expr = expr.strip()
    if expr in ['']+keyword.kwlist:
        return expr
    if expr == '""':
        return '"'+strings.pop(0)+'"'
    if expr == "''":
        return "'"+strings.pop(0)+"'"

    if '=' in expr and assign:
        e = expr[:]
        vals = ['']
        while '=' in e:
            i = e.index('=')
            if e != '' and e[0] == '=':
                vals[-1] += '='
                e = e[1:]
                continue
            if e[i-1] not in '!<>*/+-%' and e[i+1] != '=' and (vals[-1]+e[:i]).count('(') == (vals[-1]+e[:i]).count(')'):
                vals[-1] += e[:i]
                e = e[i+1:]
                vals.append('')
            else:
                vals[-1] += e[:i+1]
                e = e[i+1:]

        if len(vals) > 1:
            vals[-1] += e
            ns = ''
            left,val = vals[:-1],vals[-1]
            for l in left:
                rs = replace_names(l,True,assign=assign)
                ns += rs+'='
            ns += replace_names(val,assign=assign)
            return ns

    if expr[0] in ['(','[','{']:
        try:
            delimit = expr[0]
            i = 0; level = 1
            while level > 0:
                i += 1
                char = expr[i]
                if char in '([{':
                    level += 1
                if char in ')]}':
                    level -= 1
            inner = expr[1:i]
            rest = expr[i+1:]
            return expr[0]+replace_names(inner,defining,assign=False)+expr[i]+replace_names(rest,defining,assign=assign)

        except IndexError:
            multiline = expr
            return ''

    if expr.startswith('for') and not expr.endswith('in'):
        varname = ''
        curword = ''
        for i,char in enumerate(expr):
            if char in string.letters+string.digits+'_':
                curword += char
            else:
                if curword == 'in':
                    break
                curword = ''
            varname += char
        rest = expr[i:]

        dpart = replace_names(varname[3:-2],True,assign=assign)
        rpart = replace_names(rest,assign=assign)
        return 'for' + ' '*(dpart[0] in string.letters+string.digits+'_') + dpart + ' '*(dpart[-1] in string.letters+string.digits+'_') + 'in' + ' '*(rpart[0] in string.letters+string.digits+'_') + rpart

    if expr.startswith('lambda'):
        args = expr.split('lambda',1)[1].split(':')[0]
        replace_names(args,True,assign=assign)

    poses = ['' if e == -1 else e for e in (expr.find(char) for char in ['(','[','{'])]
    pos = min(poses)
    if pos != '':
        delimit = '([{'[poses.index(pos)]
        first,rest = expr.split(delimit,1)
        return replace_names(first,defining,assign=assign)+replace_names(delimit+rest,defining,assign=assign)

    multiline = ''
    if ' ' in expr:
        ns = ''
        for sub in expr.split(' '):
            rs = replace_names(sub,defining,assign=assign)
            if rs == '':
                continue
            if ns != '' and (ns[-1] in string.letters+string.digits+'_' and rs[0] in string.letters+string.digits+'_'):
                ns += ' '+rs
            else:
                ns += rs
        return ns

    for cmp in ['**=','//=','==','!=','<>','<=','>=','+=','-=','*=','/=','%=','//','**','<<','>>','<','>','+','-','*','/','%','&','|','^','~',':',',','.']:
        if cmp in expr:
            ns = ''
            for sub in expr.split(cmp):
                rs = replace_names(sub,defining,prefix,assign=assign)+cmp
                ns += rs 
                if cmp == '.':
                    prefix += rs 
            return ns[:-len(cmp)]

    if expr[0] in string.digits:
        return expr
    if not defining and expr not in defined:
        #print '---',prefix+expr
        if prefix+expr not in undefined and (prefix == '' or prefix[0] != '.') :
            undefined.append(prefix+expr)
        return expr
    if defining:
        if prefix+expr in undefined:
            undefined.remove(prefix+expr)
        if expr not in defined:
            defined.append(expr)
    return get_name(expr)

def fix_names(line):
    for cmp in ['**=','//=','==','!=','<>','<=','>=','+=','-=','*=','/=','%=','//','**','<<','>>','<','>','+','-','*','/','%','&','|','^','~',':',',','.','=','(',')','[',']','{','}']:
        if cmp in line:
            ns = ''
            for sub in line.split(cmp):
                ns += fix_names(sub)+cmp
            return ns[:-len(cmp)]
    if line in defined and line not in vars.values():
        return vars[line]
    return line

def first_pass(file):
    lines_firstpass = []
    for line in file:
        if line.strip() == '':
            continue
        indent = 0
        for char in line:
            if not char in string.whitespace:
                break
            indent += 1

        if multiline != '':
            line_string = ml_last
        else:
            line_string = ''
            #line_string = '\t'*indent
        line = multiline + line.strip()

        newline = ''
        while line:
            char = line[0]
            if char in ['"',"'"]:
                limit=char; i=0
                inside = ''
                escape = False
                while True:
                    i+=1; char = line[i]
                    if escape:
                        escape = False
                        inside += char
                        continue
                    if char == '\\':
                        escape = True
                    elif char == limit:
                        break
                    inside += char
                strings.append(inside)
                newline += limit*2
                line = line[i+1:]
            else:
                if char == '#':
                    break
                newline += char
                line = line[1:]
        line = newline

        if line == '':
            continue 
        if ' ' not in line:
            first = ''
        else:
            first,line = line.split(' ',1)  

        if first in ['class','def']:
            name = line.split('(')[0].split(':')[0].strip()
            line_string += first+' '
            defined.append(name)
            line_string += get_name(name)
            if '(' in line:
                line_string += '('
                inner = line.split('(',1)[1]
                inner = ')'.join(inner.split(')')[:-1])
                part = ''
                for char in inner:
                    if char == ',' and part.count('(') == part.count(')'):
                        line_string += replace_names(part,True)+','
                        part = ''
                    else:
                        part += char
                line_string += replace_names(part,True)+')'
            line_string += ':'

        to_import = ''
        importing = []
        if first == 'from':
            module,rest = line.split('import')
            module = module.strip()
            #imported.append(module)
            first,line = 'import',rest
            to_import += 'from '+module+' '

        if first == 'import':
            to_import += 'import '
            for module in line.split(','):
                module = module.strip()
                #imported.append(module)
                to_import += module+','
            to_import = to_import[:-1]
            line_string += to_import

        if line_string.strip() == '':
            r = replace_names(first+' '+line)
            if multiline != '':
                ml_last = line_string + r
                continue
            line_string += r
            ml_last = ''
        lines_firstpass.append((indent,line_string))
        #print '\t'*indent+line_string

    for i,(indent,line) in enumerate(lines_firstpass):
        lines_firstpass[i] = (indent,fix_names(line))
    return lines_firstpass

def second_pass(firstpass):
    lines = []
    current_line = ''
    current_line_indent = 0
    last_indent = 0
    for i,(indent,line) in enumerate(firstpass):
        for kw in keyword.kwlist:
            if line[:len(kw)] == kw:
                first = kw
                line = line[len(kw):]
                break
        else:
            first = ''
        limit=';'
        for kw in ['import','global']:
            if first == kw and current_line.startswith(kw):
                first = ''
                line = line.strip()
                limit=','

        if first not in ['if','elif','else','while','for','def','class','try','except','finally'] and indent == last_indent:
            current_line += limit*(current_line != '') + first + line
        else:
            lines.append((current_line_indent,current_line))
            current_line = first + line
            current_line_indent = indent
        last_indent = indent
    lines.append((current_line_indent,current_line))

    new_lines = []
    i = 0
    while i < len(lines):
        indent,line = lines[i]
        if i != len(lines)-1 and lines[i+1][0] == indent + 1 and (i == len(lines)-2 or lines[i+2][0] <= indent):
            new_lines.append((indent,line+lines[i+1][1]))
            i += 1
        else:
            new_lines.append((indent,line))
        i += 1
    return new_lines

def third_pass(lines):
    new_definitions = ''
    for u in sorted(undefined,key=lambda s:-s.count('.')):
        #print u
        parts = u.split('.')
        if parts[0] in vars.values():
            continue
        c = 0
        for indent,line in lines:
            if line.startswith('import'):
                continue
            c += line.count(u)
        if c > 1:
            new_definitions += ';'*(new_definitions!='')+get_name(u)+'='+u
            for ind,(indent,line) in enumerate(lines):
                if line.startswith('import'):
                    continue
                nline = ''
                cur_word = ''
                i = 0
                while i < len(line):
                    char = line[i]
                    if char not in string.letters+string.digits+'_.':
                        if cur_word == u:
                            nline += get_name(u)
                        else:
                            nline += cur_word
                        cur_word = ''
                        nline += char
                        i += 1
                        continue
                    if char in '"\'':
                        nline += char
                        limit = char
                        escape = False
                        while True:
                            i += 1
                            char = line[i]
                            nline += char
                            if escape:
                                escape = False
                                continue
                            if char == '\\':
                                escape = True
                            if char == limit:
                                break
                        i += 1
                        continue
                    cur_word += char
                    i += 1
                lines[ind] = (indent,nline+cur_word)

    return [lines[0]]+[(0,new_definitions)]+lines[1:]

def golf(filename):
    file = open(filename)
    write_file = open('golfed.py','w')
    for indent,line in third_pass(second_pass(first_pass(file))):
        write_file.write('\t'*(indent/2)+' '*(indent%2)+line+'\n')
    file.close()
    write_file.close()

#print first_pass(["for u in sorted(undefined,key=lambda s:-s.count('.')):"])
golf('golfer.py')

Testowane na starej fraktalnej programie graficznym miałem ( 4672 do 1889 ):

Oryginalny:

import pygame
import math
import os
import colorsys
from decimal import *

#two = Decimal(2)
#half = Decimal(0.5)

def fractal_check_point(function,x,y):
    #n = (Decimal(0),Decimal(0))
    n = (0,0)
    i = 0
    last_dist = 0
    while n[0]**2 + n[1]**2 <= 16 and i < max_iter:
        nr,ni = function(n)
        n = (nr+x,ni+y)
        i+=1
    if i == max_iter:
        return False

    #extra = math.log(math.log( (n[0]**two + n[1]**two)**half )/math.log(300),2)
    extra = math.log(math.log( (n[0]**2 + n[1]**2)**0.5 )/math.log(300),2)

    #prev = math.sqrt(last_dist)
    #final = math.sqrt(n.real**2+n.imag**2)

    return i - extra

def f((r,i)):
    return (r**2 - i**2, 2*r*i)

screen_size = (500,500)
try: screen = pygame.display.set_mode(screen_size)
except pygame.error:
    print 'Too large to draw to window...'
    screen = pygame.Surface(screen_size)

#pixels = pygame.PixelArray(screen)

#xmin = Decimal(- 2.2)
#xmax = Decimal(.8)
#
#ymin = Decimal(- 1.5)
#ymax = Decimal(1.5)

max_iter = 50

xmin = -2.2
xmax = 0.8
ymin = -1.5
ymax = 1.5


def draw_fractal():
    print repr(xmin),repr(xmax)
    print repr(ymin),repr(ymax)
    print

    xlist = []
    ylist = []

    for x in range(screen_size[0]):
        #xlist.append(Decimal(x)*(xmax-xmin)/Decimal(screen_size[0])+xmin)
        xlist.append(x*(xmax-xmin)/screen_size[0]+xmin)

    for y in range(screen_size[1]):
        #ylist.append(Decimal(y)*(ymax-ymin)/Decimal(screen_size[1])+ymin)
        ylist.append(y*(ymax-ymin)/screen_size[1]+ymin)

    xi = 0
    for x in xlist:
        yi = 0
        for y in ylist:
            val = fractal_check_point(f,x,y)
            if val == False:
                screen.set_at((xi,yi),(0,0,0))
                #pixels[xi][yi] = (0,0,0)
            else:
                r,g,b = colorsys.hsv_to_rgb(val/10.0 % 1, 1, 1)
                screen.set_at((xi,yi),(r*255,g*255,b*255))
                ##screen.set_at((xi,yi),(0,(val/300.0)**.25*255,(val/300.0)**.25*255))
                #pixels[xi][yi] = (0,(val/300.0)**.25*255,(val/300.0)**.25*255)

            yi += 1
        xi += 1
        pygame.event.get()
        pygame.display.update((xi-1,0,1,screen_size[1]))
    save_surface('F:\FractalZoom\\')

def save_surface(dirname):
    i = 0
    name = '%05d.bmp' % i
    while name in os.listdir(dirname):
        i += 1
        name = '%05d.bmp' % i
    pygame.image.save(screen,dirname+name)
    print 'saved'

x_min_step = 0
x_max_step = 0
y_min_step = 0
y_max_step = 0

savefail = 0

def zoom(xmin_target,xmax_target,ymin_target,ymax_target,steps):
    global xmin
    global xmax
    global ymin
    global ymax

    xc = (xmax_target + xmin_target)/2
    yc = (ymax_target + ymin_target)/2

    d_xmin = ((xc-xmin_target)/(xc-xmin))**(1.0/steps)
    d_xmax = ((xc-xmax_target)/(xc-xmax))**(1.0/steps)
    d_ymin = ((yc-ymin_target)/(yc-ymin))**(1.0/steps)
    d_ymax = ((yc-ymax_target)/(yc-ymax))**(1.0/steps)

    for s in range(steps):
        xmin = xc-(xc-xmin)*d_xmin
        xmax = xc-(xc-xmax)*d_xmax
        ymin = yc-(yc-ymin)*d_ymin
        ymax = yc-(yc-ymax)*d_ymax

        draw_fractal()
        save_dir = 'D:\FractalZoom\\'
        global savefail
        if not savefail:
            try:
                save_surface(save_dir)
            except:
                print 'Warning: Cannot save in given directory '+save_dir+', will not save images.'
                savefail = 1

#zoom(.5,.6,.5,.6,10)
#zoom(-1.07996839017,-1.07996839014,-0.27125861927,-0.27125861923,100)

#n = 1
#while 1:
#    pygame.display.update()
#    pygame.event.get()
#    
#    def f(x):
#        if x == 0:
#            return 0
#        else:
#            return x**n
#    draw_fractal()
#    n += .0001

draw_fractal()
zooming = 0
#firstx = Decimal(0)
#firsty = Decimal(0)
firstx = firsty = 0

clicking = 0
while 1:
    pygame.display.update()
    pygame.event.get()
    mx, my = pygame.mouse.get_pos()
    rx, ry = pygame.mouse.get_rel()

#   mx = Decimal(mx)
#   my = Decimal(my)
#   sx = Decimal(screen_size[0])
#   sy = Decimal(screen_size[1])
    sx = screen_size[0]
    sy = screen_size[1]

    if pygame.mouse.get_pressed()[0]:
        if clicking == 0:
            clicking = 1
        if zooming and clicking == 1:
            secondx = mx*(xmax-xmin)/sx+xmin
            secondy = my*(ymax-ymin)/sy+ymin

            firstx = firstx*(xmax-xmin)/sx+xmin
            firsty = firsty*(ymax-ymin)/sy+ymin

            if secondx < firstx:
                xmin = secondx
                xmax = firstx
            else:
                xmin = firstx
                xmax = secondx

            if secondy < firsty:
                ymin = secondy
                ymax = firsty

            else:
                ymin = firsty
                ymax = secondy

            screen.fill((0,0,0))
            screen.lock()
            draw_fractal()
            screen.unlock()
            zooming = 0 

        elif clicking == 1:
            firstx = mx
            firsty = my

            zooming = 1
            screen.set_at((firstx,firsty),(255,255,255))

        if clicking:
            clicking = 2

    else:
        clicking = 0

Gra w golfa:

import pygame,math,os,colorsys;from decimal import *
ai=pygame.event.get;aj=pygame.display.update;ak=math.log;al=pygame.mouse;am=False;an=pygame;ao=repr;ap=range;aq=os
def a(b,c,d):
 e=(0,0);f=0;g=0
 while e[0]**2+e[1]**2<=16 and f<o:h,i=b(e);e=(h+c,i+d);f+=1
 if f==o:return False
 j=ak(ak((e[0]**2+e[1]**2)**0.5)/ak(300),2);return f-j
def k((l,f)):return(l**2-f**2,2*l*f)
m=(500,500)
try:n=pygame.display.set_mode(m)
except pygame.error:print'Too large to draw to window...';n=pygame.Surface(m)
o=50;p=-2.2;q=0.8;r=-1.5;s=1.5
def t():
 print ao(p),ao(q);print ao(r),ao(s);print;u=[];v=[]
 for c in ap(m[0]):u.append(c*(q-p)/m[0]+p)
 for d in ap(m[1]):v.append(d*(s-r)/m[1]+r)
 w=0
 for c in u:
    x=0
    for d in v:
     y=a(k,c,d)
     if y==am:n.set_at((w,x),(0,0,0))
     else:l,z,A=colorsys.hsv_to_rgb(y/10.0%1,1,1);n.set_at((w,x),(l*255,z*255,A*255))
     x+=1
    w+=1;ai();aj((w-1,0,1,m[1]))
 B('F:\FractalZoom\\')
def B(C):
 f=0;D='%05d.bmp'%f
 while D in os.listdir(C):f+=1;D='%05d.bmp'%f
 pygame.image.save(n,C+D);print'saved'
E=0;F=0;G=0;H=0;I=0
def J(K,L,M,N,O):
 global p,q,r,s;P=(L+K)/2;Q=(N+M)/2;R=((P-K)/(P-p))**(1.0/O);S=((P-L)/(P-q))**(1.0/O);T=((Q-M)/(Q-r))**(1.0/O);U=((Q-N)/(Q-s))**(1.0/O)
 for V in ap(O):
    p=P-(P-p)*R;q=P-(P-q)*S;r=Q-(Q-r)*T;s=Q-(Q-s)*U;t();W='D:\FractalZoom\\';global I
    if not I:
     try:B(W)
     except:print'Warning: Cannot save in given directory '+W+', will not save images.';I=1
t();X=0;Y=Z=0;_=0
while 1:
 aj();ai();aa,ab=pygame.mouse.get_pos();ac,ad=pygame.mouse.get_rel();ae=m[0];af=m[1]
 if pygame.mouse.get_pressed()[0]:
    if _==0:_=1
    if X and _==1:
     ag=aa*(q-p)/ae+p;ah=ab*(s-r)/af+r;Y=Y*(q-p)/ae+p;Z=Z*(s-r)/af+r
     if ag<Y:p=ag;q=Y
     else:p=Y;q=ag
     if ah<Z:r=ah;s=Z
     else:r=Z;s=ah
     n.fill((0,0,0));n.lock();t();n.unlock();X=0
    elif _==1:Y=aa;Z=ab;X=1;n.set_at((Y,Z),(255,255,255))
    if _:_=2
 else:_=0

Uruchom na sobie (tworząc bardzo długi quine) (od 9951 do 5323 ):

import string,keyword,pkgutil;a=__builtins__.__dict__.keys();b={};c='';d='';e=[];f=[];g=[]
aw=string.letters;ax=string.digits;ay=keyword.kwlist;az=False;aA=True;aB=len;aC=enumerate;aD=open
def h(i):
 if i.startswith('__'):b[i]=i;return i
 if i in b:return b[i]
 for j in aw+'_':
    if j not in b.values():b[i]=j;return j
 for k in aw+'_':
    for l in aw+ax+'_':
     if k+l not in b.values():
        if k+l in ay:continue
        b[i]=k+l;return k+l
def m(n,o=az,p='',q=aA):
 if';'in n:
    r=''
    for s in n.split(';'):r+=m(s,q=q)+';'
    return r[:-1]
 global c;n=n.strip()
 if n in['']+ay:return n
 if n=='""':return'"'+e.pop(0)+'"'
 if n=="''":return"'"+e.pop(0)+"'"
 if'='in n and q:
    s=n[:];t=['']
    while'='in s:
     u=s.index('=')
     if s!=''and s[0]=='=':t[-1]+='=';s=s[1:];continue
     if s[u-1]not in'!<>*/+-%'and s[u+1]!='='and(t[-1]+s[:u]).count('(')==(t[-1]+s[:u]).count(')'):t[-1]+=s[:u];s=s[u+1:];t.append('')
     else:t[-1]+=s[:u+1];s=s[u+1:]
    if aB(t)>1:
     t[-1]+=s;r='';v,w=t[:-1],t[-1]
     for x in v:y=m(x,aA,q=q);r+=y+'='
     r+=m(w,q=q);return r
 if n[0]in['(','[','{']:
    try:
     z=n[0];u=0;A=1
     while A>0:
        u+=1;B=n[u]
        if B in'([{':A+=1
        if B in')]}':A-=1
     C=n[1:u];D=n[u+1:];return n[0]+m(C,o,q=az)+n[u]+m(D,o,q=q)
    except IndexError:c=n;return''
 if n.startswith('for')and not n.endswith('in'):
    E='';F=''
    for u,B in aC(n):
     if B in aw+ax+'_':F+=B
     else:
        if F=='in':break
        F=''
     E+=B
    D=n[u:];G=m(E[3:-2],aA,q=q);H=m(D,q=q);return'for'+' '*(G[0]in aw+ax+'_')+G+' '*(G[-1]in aw+ax+'_')+'in'+' '*(H[0]in aw+ax+'_')+H
 if n.startswith('lambda'):I=n.split('lambda',1)[1].split(':')[0];m(I,aA,q=q)
 J=[''if s==-1 else s for s in(n.find(B)for B in['(','[','{'])];K=min(J)
 if K!='':z='([{'[J.index(K)];L,D=n.split(z,1);return m(L,o,q=q)+m(z+D,o,q=q)
 c=''
 if' 'in n:
    r=''
    for M in n.split(' '):
     y=m(M,o,q=q)
     if y=='':continue
     if r!=''and(r[-1]in aw+ax+'_'and y[0]in aw+ax+'_'):r+=' '+y
     else:r+=y
    return r
 for N in['**=','//=','==','!=','<>','<=','>=','+=','-=','*=','/=','%=','//','**','<<','>>','<','>','+','-','*','/','%','&','|','^','~',':',',','.']:
    if N in n:
     r=''
     for M in n.split(N):
        y=m(M,o,p,q=q)+N;r+=y
        if N=='.':p+=y
     return r[:-aB(N)]
 if n[0]in ax:return n
 if not o and n not in f:
    if p+n not in g and(p==''or p[0]!='.'):g.append(p+n)
    return n
 if o:
    if p+n in g:g.remove(p+n)
    if n not in f:f.append(n)
 return h(n)
def O(P):
 for N in['**=','//=','==','!=','<>','<=','>=','+=','-=','*=','/=','%=','//','**','<<','>>','<','>','+','-','*','/','%','&','|','^','~',':',',','.','=','(',')','[',']','{','}']:
    if N in P:
     r=''
     for M in P.split(N):r+=O(M)+N
     return r[:-aB(N)]
 if P in f and P not in b.values():return b[P]
 return P
def Q(R):
 S=[]
 for P in R:
    if P.strip()=='':continue
    T=0
    for B in P:
     if not B in string.whitespace:break
     T+=1
    if c!='':U=d
    else:U=''
    P=c+P.strip();V=''
    while P:
     B=P[0]
     if B in['"',"'"]:
        W=B;u=0;X='';Y=False
        while aA:
         u+=1;B=P[u]
         if Y:Y=az;X+=B;continue
         if B=='\\':Y=True
         elif B==W:break
         X+=B
        e.append(X);V+=W*2;P=P[u+1:]
     else:
        if B=='#':break
        V+=B;P=P[1:]
    P=V
    if P=='':continue
    if' 'not in P:L=''
    else:L,P=P.split(' ',1)
    if L in['class','def']:
     i=P.split('(')[0].split(':')[0].strip();U+=L+' ';f.append(i);U+=h(i)
     if'('in P:
        U+='(';C=P.split('(',1)[1];C=')'.join(C.split(')')[:-1]);Z=''
        for B in C:
         if B==','and Z.count('(')==Z.count(')'):U+=m(Z,aA)+',';Z=''
         else:Z+=B
        U+=m(Z,aA)+')'
     U+=':'
    _='';aa=[]
    if L=='from':ab,D=P.split('import');ab=ab.strip();L,P='import',D;_+='from '+ab+' '
    if L=='import':
     _+='import '
     for ab in P.split(','):ab=ab.strip();_+=ab+','
     _=_[:-1];U+=_
    if U.strip()=='':
     ac=m(L+' '+P)
     if c!='':d=U+ac;continue
     U+=ac;d=''
    S.append((T,U))
 for u,(T,P)in aC(S):S[u]=(T,O(P))
 return S
def ad(ae):
 af=[];ag='';ah=0;ai=0
 for u,(T,P)in aC(ae):
    for aj in ay:
     if P[:aB(aj)]==aj:L=aj;P=P[aB(aj):];break
    else:L=''
    W=';'
    for aj in['import','global']:
     if L==aj and ag.startswith(aj):L='';P=P.strip();W=','
    if L not in['if','elif','else','while','for','def','class','try','except','finally']and T==ai:ag+=W*(ag!='')+L+P
    else:af.append((ah,ag));ag=L+P;ah=T
    ai=T
 af.append((ah,ag));ak=[];u=0
 while u<aB(af):
    T,P=af[u]
    if u!=aB(af)-1 and af[u+1][0]==T+1 and(u==aB(af)-2 or af[u+2][0]<=T):ak.append((T,P+af[u+1][1]));u+=1
    else:ak.append((T,P))
    u+=1
 return ak
def al(af):
 am=''
 for an in sorted(g,key=lambda s:-s.count('.')):
    ao=an.split('.')
    if ao[0]in b.values():continue
    j=0
    for T,P in af:
     if P.startswith('import'):continue
     j+=P.count(an)
    if j>1:
     am+=';'*(am!='')+h(an)+'='+an
     for ap,(T,P)in aC(af):
        if P.startswith('import'):continue
        aq='';ar='';u=0
        while u<aB(P):
         B=P[u]
         if B not in aw+ax+'_.':
            if ar==an:aq+=h(an)
            else:aq+=ar
            ar='';aq+=B;u+=1;continue
         if B in'"\'':
            aq+=B;W=B;Y=False
            while aA:
             u+=1;B=P[u];aq+=B
             if Y:Y=az;continue
             if B=='\\':Y=True
             if B==W:break
            u+=1;continue
         ar+=B;u+=1
        af[ap]=(T,aq+ar)
 return[af[0]]+[(0,am)]+af[1:]
def at(au):
 R=aD(au);av=aD('golfed.py','w')
 for T,P in al(ad(Q(R))):av.write('\t'*(T/2)+' '*(T%2)+P+'\n')
 R.close();av.close()
at('golfer.py')

BrainFuck - 489 znaków

Usuwa wszystkie znaki niewykonywalne. Szanuje komentarze od # do końca linii.

,[>--[<++>+++++++]<+[>+>+<<-]>[<+>-]+>[<->[-]]<[[,----------]]<>>--[>+
<++++++]<<--------[>+>+<<-]>[<+>-]+>[<->[-]]<[->>.<<]>>+<<<-[>+>+<<-]>
[<+>-]+>[<->[-]]<[->>.<<]>>+<<<-[>+>+<<-]>[<+>-]+>[<->[-]]<[->>.<<]>>+
<<<-[>+>+<<-]>[<+>-]+>[<->[-]]<[->>.<<]>>++++++++++++++<<<------------
--[>+>+<<-]>[<+>-]+>[<->[-]]<[->>.<<]>>++<<<--[>+>+<<-]>[<+>-]+>[<->[-
]]<[->>.<<]>++++[>+++++++<-]>+<<++++[<------->-]<-[>+>+<<-]>[<+>-]+>[<
->[-]]<[->>.<<]>>++<<<--[>+>+<<-]>[<+>-]+>[<->[-]]<[->>.<<]>>[-]<<<,]

Oczywiście przebiegnij przez to źródło:

,[

#subtract #
>--[<++>+++++++]<+

#strip comments
[>+>+<<-]>[<+>-]+>[<->[-]]<[[,----------]]<

#put '+' in 4th cell
>>--[>+<++++++]<<
#+
--------
[>+>+<<-]>[<+>-]+>[<->[-]]<[->>.<<]>>+<<<
#,
-
[>+>+<<-]>[<+>-]+>[<->[-]]<[->>.<<]>>+<<<
#-
-
[>+>+<<-]>[<+>-]+>[<->[-]]<[->>.<<]>>+<<<
#.
-
[>+>+<<-]>[<+>-]+>[<->[-]]<[->>.<<]>>++++++++++++++<<<
#<
--------------
[>+>+<<-]>[<+>-]+>[<->[-]]<[->>.<<]>>++<<<
#>
--
[>+>+<<-]>[<+>-]+>[<->[-]]<[->>.<<]>++++[>+++++++<-]>+<<
#[
++++[<------->-]<-
[>+>+<<-]>[<+>-]+>[<->[-]]<[->>.<<]>>++<<<
#]
--
[>+>+<<-]>[<+>-]+>[<->[-]]<[->>.<<]>>[-]<<<
,]

Brainfuck golfista w Bash (v3)

To jest praca w toku, będę ją aktualizować, jeśli będę mógł.

Odczytuje z pliku (nazwa pliku powinna być pierwszym argumentem wiersza poleceń).

Na razie to wszystko

• Usuń wszystkie znaki, które nie są <>+-.,[]
• Usuń ciągi dwuznakowych które robią nic użytecznego, na przykład <>, ><, +-,-+
• Po zakończeniu powtarza całą procedurę, więc >>>><<<<<zostaje zredukowana do<

Kod

#!/bin/bash

#if the file exists, take input from it
if [ -f $1 ]; then
input=`cat $1`
else
#complain to STDERR and exit
echo "File not found. Exiting with status 1.">&2
exit 1
fi

original=$input #save the original input
code=$input #save original input to another variable which we will be golfing
code=`echo $code|grep -o "[][<>.,+-]"|tr -d " \n"` #remove non-executable chars
output=$code; #this will be output
hits=-1; #Count the number of golfing operations carried out every time the code loops. When this is 0, the code will be completely golfed.

until [ $hits = 0 ]; do
hits=0
#we will be processing the optimised version from last time
code=$output
output=""

#Keep taking characters off from $code until it is empty
until [ m$code = m ]; do
#examine the first two chars
c1=${code:0:1}
code=`echo $code|cut -c2-`
c2=${code:0:1}
#if this is 1, the two characters now being read will be removed
ignore=0

if [ $c1$c2 = "<>" ] ; then
#set the second character to be taken off as well and not saved to output
ignore=1
elif [ $c1$c2 = "><" ] ; then
ignore=1
elif [ $c1$c2 = "+-" ] ; then
ignore=1
elif [ $c1$c2 = "-+" ] ; then
ignore=1
else
#save the char we took off to output, we aren't removing it
output=$output$c1;
fi

if [ $ignore = 1 ]; then
#ignore the second character and save no chars to output
code=`echo $code|cut -c2-`
#another hit
hits=`expr $hits + 1`
fi

#end inner until loop
done
#end main loop
done

#done, print output
echo $output;
exit 0;

Jak to działa

Po usunięciu wszystkich znaków niewykonywalnych wykonuje następujące czynności. Licznik trafień jest ustawiony -1na początku - zlicza liczbę operacji golfowych wykonanych za każdym razem, gdy biegnie zewnętrzna pętla.

1. Jeśli golfowany kod jest pusty, przejdź do kroku 5.
2. Przeczytaj dwa pierwsze znaki z kodu i usuń pierwszy znak.
3. Jeśli są <>, ><, +-lub -+, dodać 1 do hitslicznika i wróć do kroku 1.
4. Jeśli nie, zapisz pierwszy znak do wysłania i przejdź do kroku 1.
5. Jeśli licznik trafień wynosi 0, wydrukuj i wyjdź.
6. Jeśli nie, zresetuj licznik trafień na 0, ustaw kod golfowy jako zmienną wyjściową, zresetuj dane wyjściowe do pustego ciągu i przejdź do kroku 1.

HQ9 + golfista w Bash (v3)

Wiem, że HQ9 + jest bezużyteczny, ale równie dobrze mogę przesłać pięciowierszowy. Czyta zstandardowe wejścieplik. Ścieżka do pliku powinna być pierwszym argumentem wiersza poleceń.

cechy

• Usuwa wszystkie znaki komentarza (cokolwiek oprócz HhQq9+)
• Usuwa +(zwiększa liczbę, ale nie ma możliwości wydrukowania tej liczby)
• Konwertuje hqna wielkie litery (nie gra w golfa)

Kod

#!/bin/bash
if [ -f $1 ]; then
input=`cat $1`
else exit 1; fi
echo $input|tr "[[:lower:]]" "[[:upper:]]"|grep -o "[HQ9]"|tr -d ' \n'
exit 0

Java z Javą

Pobiera nazwę pliku jako argument wiersza poleceń i edytuje plik na swoim miejscu.

• Usuwa komentarze
• Skraca identyfikatory (w tym klasy, metody, zmienne, parametry metody i parametry wyrażenia lambda)
• Przycina przestrzenie
• Skraca import
• Usuwa niepotrzebne nawiasy klamrowe wokół instrukcji jednowierszowych
• Konwertuje while(true)nafor(;;)
• Usuwa niepotrzebne modyfikatory, takie jak privateifinal

Po uruchomieniu programu jego rozmiar zmniejsza się z 7792 do 4366.

import java.io.File;
import java.io.IOException;
import java.nio.file.Files;
import java.nio.file.Path;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.List;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
import java.util.stream.Collectors;
import java.util.stream.IntStream;

public class Golfer {

    public static void main(String[] args) throws IOException {
        Path path = new File(args[0]).toPath();
        String program = Files.readAllLines(path).stream().collect(Collectors.joining("\n"));
        Golfer golfer = new Golfer(program);
        System.out.println(golfer.toString().length() + " characters");
        System.out.println(golfer.toString().getBytes().length + " bytes");
        golfer.golf();
        String str = golfer.toString();
        Files.write(path, Arrays.asList(str));
        System.out.println(golfer);
        System.out.println(golfer.toString().length() + " characters");
        System.out.println(golfer.toString().getBytes().length + " bytes");
    }

    private String program;

    public Golfer(String program) {
        this.program = program;
    }

    public void golf() {
        doUnicodeSubstitutions();
        protectStrings();
        removeComments();
        removeDuplicateSpaces();
        removeExcessSpaces();
        removeUnnecessaryModifiers();
        simplifyImports();
        shortenIdentifiers();
        improveControlStructures();
        unprotectStrings();
    }

    void removeDuplicateSpaces() {
        program = program.replaceAll("\\s+", " ");
    }

    void removeExcessSpaces() {
        program = program.trim();
        program = program.replaceAll("([.,;:?!@%^&*()\\[\\]{}=|/<>~]) ", "$1");
        program = program.replaceAll(" ([.,;:?!@%^&*()\\[\\]{}=|/<>~])", "$1");
        program = program.replaceAll("([^+]) \\+", "$1+").replaceAll("\\+\\+ \\+", "+++").replaceAll("\\+ ([^+])", "+$1");
        program = program.replaceAll("([^-]) -", "$1-").replaceAll("-- -", "---").replaceAll("- ([^-])", "-$1");
    }

    void removeUnnecessaryModifiers() {
        program = program.replaceAll("private |final |@Override ", "");
    }

    void simplifyImports() {
        int startImports = program.indexOf("import ");
        List<String> imports = new ArrayList();
        Matcher importMatcher = Pattern.compile("import [A-Za-z0-9$_.*]*;").matcher(program);
        while (importMatcher.find()) {
            imports.add(importMatcher.group());
        }
        for (int i = 0; i < imports.size(); i++) {
            if (!imports.get(i).endsWith("*;")) {
                imports.set(i, imports.get(i).replaceFirst("\\.[A-Za-z0-9$_]*;", "\\.*;"));
            }
            imports = imports.stream().distinct().collect(Collectors.toList());
            program = program.replaceAll("import [A-Za-z0-9$_.*]*;", "");
            program = program.substring(0, startImports) + String.join("", imports) + program.substring(startImports);
        }
    }

    private List<Character> unusedCharacters;

    void shortenIdentifiers() {
        unusedCharacters = IntStream.concat(IntStream.rangeClosed('a', 'z'), IntStream.rangeClosed('A', 'Z'))
                .mapToObj(i -> (char) i)
                .filter(c -> !Pattern.compile("[^A-Za-z0-9$_']" + c + "[^A-Za-z0-9$_]").matcher(program).find())
                .collect(Collectors.toList());
        shortenIdentifiers("(class|interface|enum) ([A-Za-z0-9$_]{2,})( extends [A-Za-z0-9$_]+)?( implements [A-Za-z0-9$_,]+)?\\{",
                           2, "[^A-Za-z0-9$_]", "[^A-Za-z0-9$_]");
        shortenIdentifiers("([A-Za-z0-9$_]+(\\[+\\]+| )|<[A-Za-z0-9$_,]+>+\\[+\\]+)(?<!implements |else |package |import |return )([A-Za-z0-9$_]{2,})(?<!null)[=,;)]",
                           3, "[^A-Za-z0-9$_]", "[^A-Za-z0-9$_(]");
        shortenIdentifiers("([A-Za-z0-9$_]+ |<[A-Za-z0-9$_,]+>+)(?<!new |else )([A-Za-z0-9$_]{2,})(?<!main|toString|compareTo|equals|hashCode|paint|repaint)\\(",
                           2, "[^A-Za-z0-9$_]", "\\(", "::", "[^A-Za-z0-9$_]");
        shortenIdentifiers("[^A-Za-z0-9$_]([A-Za-z0-9$_]{2,})(,[A-Za-z0-9$_]+)*\\)?->",
                           1, "[^A-Za-z0-9$_]", "[^A-Za-z0-9$_(]");
    }

    void shortenIdentifiers(String pattern, int groupNumber, String... afficesForReplacement) {
        Pattern compiledPattern = Pattern.compile(pattern);
        Matcher matcher;
        while ((matcher = compiledPattern.matcher(program)).find()) {
            String identifer = matcher.group(groupNumber);
            char newIdentifier;
            if (unusedCharacters.remove((Character) identifer.charAt(0))) {
                newIdentifier = identifer.charAt(0);
            } else if (Character.isUpperCase(identifer.charAt(0))
                       && unusedCharacters.remove((Character) Character.toLowerCase(identifer.charAt(0)))) {
                newIdentifier = Character.toLowerCase(identifer.charAt(0));
            } else if (Character.isLowerCase(identifer.charAt(0))
                       && unusedCharacters.remove((Character) Character.toUpperCase(identifer.charAt(0)))) {
                newIdentifier = Character.toUpperCase(identifer.charAt(0));
            } else if (unusedCharacters.size() > 0) {
                newIdentifier = unusedCharacters.remove(0);
            } else {
                System.err.println("out of identifiers");
                break;
            }
            for (int i = 0; i < afficesForReplacement.length; i += 2) {
                program = program.replaceAll('(' + afficesForReplacement[i] + ')' + identifer + "(?=" + afficesForReplacement[i + 1] + ')', "$1" + newIdentifier);
            }
        }
    }

    void improveControlStructures() {
        program = program.replaceAll("while\\(([^()]+)\\)\\{", "for(;$1;){");
        program = program.replaceAll("for\\(([^()]*);true;([^()]*)\\)", "for($1;;$2)");
        while (!program.equals(removeBrackets(program))) {
            program = removeBrackets(program);
        }
    }

    String removeBrackets(String string) {
        return string.replaceAll("((if|while|do)\\([^{;]*\\))\\{([^;}]*;)\\}", "$1$3")
                .replaceAll("else\\{([^;}]*;)\\}", "else $1")
                .replaceAll("(for\\([^{;]*;[^{;]*;[^{;]*\\))\\{([^;}]*;)\\}", "$1$2");
    }

    void protectStrings() {
        adjustStrings(1000);
    }

    void unprotectStrings() {
        adjustStrings(-1000);
    }

    void adjustStrings(int n) {
        char[] chars = new char[program.length()];
        for (int i = 0; i < program.length(); i++) {
            chars[i] = program.charAt(i);
            if (chars[i] == '"' && chars[i - 1] != '\'') {
                for (i++; chars.length > i + 1 && program.charAt(i) != '"'; i++) {
                    chars[i] = (char) (program.charAt(i) + n);
                }
                chars[i] = program.charAt(i);
            }
        }
        program = new String(chars);
    }

    void removeComments() {
        program = program.replaceAll("(?s)/\\*.*\\*/", "");
        program = program.replaceAll("//.*\n", "");
    }

    void doUnicodeSubstitutions() {
        List<Character> chars = new ArrayList();
        for (int i = 0; i < program.length(); i++) {
            if (program.charAt(i) != '\\' || program.charAt(i + 1) != 'u') {
                chars.add(program.charAt(i));
            } else {
                chars.add((char) Integer.parseInt(program.substring(i + 2, i + 6), 16));
                i += 5;
            }
        }
        char[] charArray = new char[chars.size()];
        for (int i = 0; i < charArray.length; i++) {
            charArray[i] = chars.get(i);
        }
        program = new String(charArray);
    }

    @Override
    public String toString() {
        return program;
    }
}

Perl, części 1–2

(usuwa komentarze i ignoruje #znaki w podwójnych cudzysłowach)

(usuwa wszystkie białe znaki po nawiasach i =znakach)

Nie próbowałem golfa tym kodem. Może po zakończeniu może sam zagrać w golfa.

$/="\n\n";
chomp($prog = <>);
for(1..length($prog)){
 $rev.=chop $prog;
}
#print$rev;
for(1..length($rev)){
 $temp = chop$rev;
 if ($temp eq '"'){
  if($quote == 0){$quote = 1}
  else{$quote = 0}
 }
 if($temp eq '#' && $quote == 0){$comment = 1}
 if($temp eq "\n"){
  $comment = 0;
 }
 if($comment != 1){
  $prog.=$temp;
 }
}
for(1..length($prog)){
 $rev.=chop $prog;
}
for(1..length($rev)){
 $temp = chop$rev;
 if ($temp eq ";" || $temp eq "}" || $temp eq ")" || $temp eq "=" || $temp eq "{"){
  $ws = 2;
 }
 if(($temp eq "\n" || $temp eq " ") && $ws != 0){$ws = 1}elsif($ws==1){$ws=0}
 if($ws != 1){
  $prog.=$temp;
 }
}
print$prog;

Przykładowe dane wejściowe

for(1..10) {
 print "Hello, World!";#prints hello world
 print ($n = <>); #prints "#"
}

Wydajność

for(1..10){print "Hello, World!";print ($n =<>);}

Następnie wyeliminuje spacje między symbolami i znakami alfanumerycznymi.

Golfista Java w Perlu

WIP w tej chwili, chociaż teraz robi się całkiem niezły kod.

Cechy:

• Usuwa wszystkie komentarze
• Usuwa niepotrzebne białe znaki
• Usuwa deklaracje pakietów (w każdym razie jest to jeden plik!)

Kod

#!/usr/bin/env perl
use strict;
use warnings;

my $string_re = qr/"(?:\\[btnfr"'\\]|\\(?:[0-3][0-7][0-7]|[0-7]{1,2})|\\u+[0-9a-fA-F]{4}|[^\r\n\\"])*"/;
my @lines = <>;
my @strings = ();

# First, replace strings with placeholders. Strings suck REALLY hard when replacing.
map {
  while (m/$string_re/) {
    s//"\032".@strings."\032"/e;
    push @strings, $&;
  }
} @lines;

# Remove comments
my $in_comment = 0;
my $partial_line;

sub remove_comments {
  my $line = $_[0];
  start:
  if ($in_comment) {
    if ((my $mulc_end_index = index $line, '*/') != -1) {
      $line = $partial_line . substr $line, $mulc_end_index + 2;
      $in_comment = 0;
      goto no_comment;
    }
    return 0;
  }

  no_comment:
  my $eolc_idx = index $line, '//';
  my $mulc_idx = index $line, '/*';

  if ($mulc_idx == -1 || ($eolc_idx != -1 && $eolc_idx < $mulc_idx)) {
    $line =~ s;//.*$;;;
  } elsif ($mulc_idx != -1) {
    $in_comment = 1;
    $partial_line = substr $line, 0, $mulc_idx;
    goto start;
  }
  $_ = $line;
  return 1
}

@lines = grep {&remove_comments($_)} @lines;

# Remove empty lines, remove line ends
@lines = grep(!/^\s*$/, @lines);
map {chomp;s/^\s*//;s/\s*$//} @lines;

# Remove unnecessary whitespace
map {s/\s*([][(){}><|&~,;=!+-])\s*/$1/g} @lines;

# Remove unnecessary package declaration
$lines[0] =~ s/^package [^;]+;//;

# Finally, put strings back.
map {s/\032(\d+?)\032/$strings[$1]/g} @lines;

print @lines;