c++ - reading input from keyboard -

i required read keyboard(stdin), following text. pl note entered user keyboard in format only.

 #the total size of physical memory (units b, kb, mb, gb)     512mb   2       #the following memory allocations     {             abc = alloc(1kb);              {                 y_ = alloc(128mb);                 x1= alloc(128mb);                 y_ = alloc(32mb);                (i = 0; < 256; i++) abc[i] =alloc(512kb);                   x1 = alloc(32mb); x2 = alloc(32mb); x3 = alloc(32mb);                x1.next = x2, x2.next = x3, x3.next = x1;     }     abc = alloc(256mb);     } 

so let me break down. line beginning # sign considered comment , ignored. first 2 allocations physical memory size , number of generations. global bracket opened. , may followed line called

abc = alloc(1kb); 

where abc object name, 1kb memory size allocated. x1.next = x2, here x1 points x2.

for (i = 0; < 256; i++) abc[i] =alloc(512kb); 

the loop os entered in format , can have same line command or can have nested loops.

i have following code takes care of somewhat. want know improve on it. please help.

and code this:

#include <iostream> #include <algorithm> #include <string> #include <iomanip> #include <limits> #include <stdio.h> #include <sstream>    using namespace std; using std::stringstream;  string pmem,sgen, comment,val,input,input_for,id_size,id,init_str1, init_str2, inc_str, id_dummy,s_out,sss, id_dummy1; int gen=0, pmem_int=0,i=0, gbrckt =0,cbrckt=0, obrckt=0, id_size_int,v1,v2, for_obrckt=0,for_cbrckt=0,y=0, y1=0, g=0; unsigned long pmem_ulong =0, id_size_ulong; char t[20], m[256], init1[10],init2[10],inc[10]; unsigned pos_start, pos,pos_strt=0,pos_end=0;  string extract(string pmem_extract); unsigned long tobyte(int pmem_int_func, string val); void commentignore(string& input); void func_insert(); void func_insert_for(); stringstream out; void commentignore_for(string& input_for);  int main()  {   /* reading input main memory , num of generations */   /* ignoring comment line */   cin >> pmem;   if(pmem == "#") {     cin.clear();     pmem.clear();     getline(cin,comment);     cin >> pmem;   }   if(pmem == "#") {     cin.clear();     pmem.clear();     getline(cin,comment);     cin >> pmem;   }   if(pmem == "#") {     cin.clear();     pmem.clear();     getline(cin,comment);     cin >> pmem;     }   /* reading input generations */   cin>> sgen;   if(sgen == "#") {     cin.clear();     sgen.clear();     getline(cin,comment);     cin >> sgen;   }   if(sgen == "#") {     cin.clear();     sgen.clear();     getline(cin,comment);     cin >> sgen;   }   if(sgen == "#") {     cin.clear();     sgen.clear();     getline(cin,comment);     cin >> sgen;   }   /* convert sgen , physical memory int , report error if not number */   gen = atoi(sgen.c_str());   if(gen ==0) {     cerr << "generation must number"<<endl;       exit(0);   }   pmem_int = atoi(pmem.c_str());   //  cout<< gen<<" "<<pmem_int<<endl;    /* number pmem removed, unit b,mb,kb */   extract(pmem); /* returns val(string) */    /* convert given physical memory byte. input: pmem_int*/   tobyte(pmem_int, val); /*  return(pmem_ulong)*/   // move pmem_ulond location keep address intact   /* read rest of inputs  */   /* ignore comment lines before global bracket */    cin >> input;   if(input == "#"){     cin.clear();     input.clear();     getline(cin,comment);     cin >> input;   }   if(input == "#"){     cin.clear();     input.clear();     getline(cin,comment);     cin >> input;   }   if(input == "#"){     cin.clear();     input.clear();     getline(cin,comment);     cin >> input;   }    if(input.compare("{") ==0)     gbrckt=1;    else {     cerr<< "syntax error\n";     exit(0);   }    /* clearing input stream next input */   cin.ignore(numeric_limits<streamsize>::max(), '\n');   cin.clear();   input.clear();   //cout<<"input: "<<input<<endl;   while( getline(cin,input)) {      if(input == "ctrl-d")       break;      commentignore(input);     //cout<<"inputloop: "<<input<<endl;      /* if input = '{' or '}'*/     if(input.compare("{") ==0)       obrckt = obrckt + 1;       if (input.compare("}") ==0)       cbrckt = cbrckt + 1;       if (((input.find("alloc"))!= string::npos) && (input.find("alloc") < input.find("for"))) {        func_insert();        //call allocate function here name: id, size: id_size_ulong      }       if ((input.find("for")) != string::npos) {         sscanf(input.c_str(), "for (%s = %d; %s < %d; %[^)])", init1, &v1, init2, &v2, inc);     init_str1 = init1, init_str2 = init2, inc_str = inc;      cout<<init1<<" ="<< v1<<" "<<init_str1<<" < " << v2<< " "<< inc_str<<endl;     cout << input <<endl;      if(init_str1 != init_str2) {       cerr << "error!\n";       exit(0);     }      if ((input.find("alloc"))!= string::npos) {       // unsigned pos = (input.find("alloc"));        if((input.find(";")) != string::npos) {           pos_start = (input.find(")")+1);         string alloc_substr  = input.substr(pos_start);         cout<<"substring alloc: "<< alloc_substr<<endl;          func_insert();         //call allocate function here name: id, size: id_size_ulong       }       else {         cerr << "error: syntax\n";         exit(0);       }       }     //  cin.ignore();     while(getline(cin,input_for)) {       commentignore_for(input_for);        if ((input_for.find("{") != string::npos)) {         pos = input_for.find("{");         for_obrckt = for_obrckt+1;         string for_brckt = input_for.substr(pos,pos);         cout<< "found: " << for_obrckt<<endl;       }         if ((input_for.find("}") != string::npos)) {          pos = input_for.find("}");         for_cbrckt = for_cbrckt+1;         string for_brckt = input_for.substr(pos,pos);         cout<< "found: " << for_cbrckt<<endl;       }        if (((input_for.find("alloc"))!= string::npos) && (input_for.find("alloc") < input_for.find("for"))) {         func_insert_for();         //call allocate function here name: id, size: id_size_ulong       }           if(for_obrckt == for_cbrckt)         break;       }     cout<<"out of break"<<endl;      }        if (((input.find(".next"))!= string::npos) && (input.find(".next") < input.find("for"))) {        func_insert();        //call allocate function here name: id, size: id_size_ulong      }        if(((cbrckt-obrckt)) == gbrckt)        break;   }  }  /*---------------------- function definitions --------------------------------*/ /* function extract string part of physical memory */ string extract(string pmem_extract) {   i=0;   const char *p = pmem_extract.c_str();   for(i=0; i<=(pmem_extract.length()); i++) {     if (*p=='0'|| *p=='1'|| *p=='2'|| *p=='3'|| *p =='4'|| *p=='5'|| *p=='6'|| *p=='7'|| *p=='8'|| *p=='9')       *p++;     else {       val = pmem_extract.substr(i);       return(val);     }   } }  /* convert physical memory bytes. return(pmem_ulong);*/ unsigned long tobyte(int pmem_int_func, string val) {   if (val == "kb")     pmem_ulong =  (unsigned long) pmem_int_func * 1024;   else if (val == "b")     pmem_ulong = (unsigned long) pmem_int_func;   else if (val == "gb")     pmem_ulong = (unsigned long) pmem_int_func * 1073741824;   else if (val == "mb")     pmem_ulong = (unsigned long) pmem_int_func * 1048576;   else {     cerr<<"missing value in memory, b, kb, mb, gb\n";     exit(0);   }    return(pmem_ulong); }   /*ignoring comment line*/ void commentignore(string& input) {   unsigned found = input.find('#');    if (found!=std::string::npos)    input= input.erase(found);    else     return;   return; }   void func_insert() {  sscanf(input.c_str(), "%s = alloc(%[^)]);", t, m);        id =t;        id_size =m;        cout<<"tag: "<<id <<"  memory: "<<id_size<<endl;        extract(id_size); /* separates b,mb,kb , gb of input, returns val*/        id_size_int = atoi(id_size.c_str());        /* convert object size b */        tobyte(id_size_int, val); /* return(pmem_ulong) */        id_size_ulong = pmem_ulong;  }  void func_insert_for() {   sscanf(input_for.c_str(), "%s = alloc(%[^)]);", t, m);   id =t;   id_size =m;   if(!((id.find("[")) && (id.find("]")) != string::npos)) {     cout<<"tag: "<<id <<"  memory: "<<id_size<<endl;     extract(id_size); /* separates b,mb,kb , gb of input, returns val*/     id_size_int = atoi(id_size.c_str());     /* convert object size b */     tobyte(id_size_int, val); /* return(pmem_ulong) */     id_size_ulong = pmem_ulong;     // allocate here     return;   }   else {     if(inc_str.find("++"))       y1 =1;     if(inc_str.find("="))       {     sss = inc_str.substr(inc_str.find("+") +1);     y1 = atoi(sss.c_str());     cout<<"y1:"<<y1<<endl;        }     pos_strt = id.find("[");     pos_end = id.find("]") -1;     cout<<"positions start , ebd: " << pos_strt<<pos_end<<endl;     id_dummy = id.substr(0,pos_strt);     id = id_dummy;     cout<<"tag: "<<id_dummy <<"  memory: "<<id_size<<endl;     extract(id_size); /* separates b,mb,kb , gb of input, returns val*/     id_size_int = atoi(id_size.c_str());     /* convert object size b */     tobyte(id_size_int, val); /* return(pmem_ulong) */     id_size_ulong = pmem_ulong;     //allocate here     cout<<"v1: " << v1 << " " << v2<<endl;     // g = 0;     for(y = v1; y < v2; y= y+y1) {       // allocate here     }   }   return; }  void commentignore_for(string& input_for) {   unsigned found = input_for.find('#');    if (found!=std::string::npos)    input_for= input_for.erase(found);    else     return;   return; } 

also required make whitespace compatible. means input can entered in 1 line well. 2 allocations in 1 line. have not been able taken care. need that.

my suggestion write proper tokenizer - piece of code understands belongs (such "words"), , split (e.g. "(, ), {, }"). tokenizer return enum, this:

enum token {     token_unknown,   // error indiciation.      token_leftparen,     token_rightparen,     token_leftbracket,     token_rightbracket,     token_comma,     token_semicolon,      token_equal,     token_word,        // sequence of alphanumerics }; 

once have token, need understand token is. may make sense have table of "known" tokens (aka "keywords"), such "mem", "alloc", "for", , on. if isn't keyword, it's name of symbol, such variable. store in table, can refer them later.

you have use sort of stack of are, can came when finish.

writing generic parser isn't terribly hard, , end lot less code current code in doing so.

certainly, you'd able rid of the:

  if(input == "#"){     cin.clear();     input.clear();     getline(cin,comment);     cin >> input;   } 

just let parser check if there '#' input, , skip on end of line (completing current token if have one, if not, keep going).