Page 1 of 1

288 - Arithmetic Operations With Large Integers

Posted: Sat Apr 27, 2002 8:43 pm
by Christian Schuster
Hello World,

I tried to get AC for that one several times, but no way...

I'm currently using term splitting, scanning for "+" and "-" from the right (they associate left), afterwards for "*" from the right (dito) and for "**" (which I replaced before) from the left (right associative). I'm then calling that thing recursively. If no operator is found, it's a simple number, where I remove leading zeroes and return.

My program seems to work quite correct for all of my test cases, so I'm sure there's some tricky input. BTW: What's the result of 0**0?

Could someone help me? Any hints?

Here are some (in my opinion) critical samples:
3**4**2 => 43046721
000333-0001000 => -667
3-5+2 => 0
2**9999-4**4999*2 => 0

Posted: Sat Apr 27, 2002 9:09 pm
by Adrian Kuegel
I have to add some questions:
What about 0**0**0? And 1-2**2?
And I hope that I did understand it right that there is not a input like -2*-2.
I also print the result in a line, not in several lines like in the sample output, because the description says: "Print each test case in a different line."
I remove leading zeros all time, even directly before output, and I print 0 if the result is 0.
What could be the reason that I get Wrong Answer?

Posted: Sat Apr 27, 2002 9:18 pm
by Adrian Kuegel
Now I got Accepted. I forgot to store the sign correctly for the result after addition. If the mentioned cases are in the input, then:
0**0 = 1
0**0**0 = 0
1-2**2 = -3
and the other results are like Christian wrote.

Posted: Sat Apr 27, 2002 10:30 pm
by Christian Schuster
I finally got AC - forgot to reset the sign flag in multiplication... :-?

Thanks for the inspiration... :D

Posted: Sat May 04, 2002 1:32 am
by ha_tran
I can't understand why I got it right with all of your results above, however, it's still not AC.
Is there any trick in the input?

Posted: Mon May 27, 2002 9:11 am
by bluna
I don't know what to do!
On my computer it works just fine, but when I send it, it always, and I mean ALWAYS, come back the SIGSEGV error:
Invalid memory reference!!

Have you got any ideas, what is wrong??


Posted: Mon May 27, 2002 11:52 am
by Adrian Kuegel
I think, you are solving the problem recursively. Therefore you have to try that your recursion depth is as small as possible to avoid stack overflow.

Posted: Tue May 28, 2002 12:51 pm
by bluna
I had a solution with recurision, but when I got runtime error message, I changed it into non-recursion way.

Posted: Tue May 28, 2002 12:51 pm
by bluna
I had a solution with recurision, but when I got runtime error message, I changed it into non-recursion way.
10x anyway for suggestion

288 - compile error

Posted: Wed Jul 31, 2002 7:24 pm
by bbio_bibo
I reveived the reply telling me "Compile Error"
,but those errors occur in some *.h files
I don't quite understand how this could happen.

I have tried using "operator overload" or "member function only"
but all have error like - "const BigInteger" etc.

partial messages are as below:
erator.h: In instantiation of `iterator_traits<BigInteger>':
go.h:257: instantiated from here
erator.h:104: no type named `iterator_category' in `class BigInteger'
erator.h:105: no type named `value_type' in `class BigInteger'
erator.h:106: no type named `difference_type' in `class BigInteger'
erator.h:107: no type named `pointer' in `class BigInteger'
erator.h:108: no type named `reference' in `class BigInteger'
go.h: In function `int main(...)':
go.h:257: no type named `difference_type' in `struct iterator_traits<BigInteger>

288 - Runtime Error

Posted: Sat Nov 01, 2003 5:31 pm
by potato
This problem seems to be quite easy. But my program kept getting Rutime error.
I wonder if there are some tricky input cases?

Posted: Thu May 27, 2004 4:00 am
by windows2k
Sorry, I tried to solve the problem and get WA all the time.
I have passed all the input/output above.
Could someone give more tircky Input/Output? Thx :D

Posted: Thu May 27, 2004 11:51 am
by Adrian Kuegel
Perhaps the following information helps you:
There may be more than 2500 digits in a (intermediate) result, but never more than 3000 digits.
Each input line contains at most 200000 characters.
And there may be leading zeros.

Posted: Sun Mar 26, 2006 6:53 am
by Ryan Pai
So I keep getting WA. I implemented, but didn't post, a class Z which is meant to be an arbitrary precision integer class (it is meant to act like int, overloaded operaters etc.). I'm pretty sure the +,* and - work with the Z class, since I've used it on other problems. ** also seems to work (I tested it with quite a few powers of 2). Could there be a problem with the parsing?

Code: Select all

string parse(string);
Z expression(istream&);
Z term(istream&);
Z factor(istream&);

int main(){
  string s;
    istringstream sin(parse(s));
    ostringstream sout;
    //while(s.size()>75){ cout<<s.substr(0,75)<<"\\\\\n"; s=s.substr(75); }

Z expression(istream& in){
  Z a=term(in);
  char op;
    Z b=term(in);
  return a;

Z term(istream& in){
  Z a=factor(in);
    Z b=factor(in);
  return a;

Z factor(istream& in){
  Z a;
    Z b=factor(in);
  return a;

string parse(string s){
  string t;
  for(int i=0;i<s.size();i++)
    if(s[i]=='*' && i+1<s.size() && s[i+1]=='*'){
      t+='^'; i++;
  return t;

Posted: Mon Mar 27, 2006 3:16 pm
by Adrian Kuegel
I think your parsing looks good. The only thing which I can see which may cause some problems is a leading unary -, for example:
I don't know if something like this occurs in the judge input, but my program handles it.
Edit: I just read that such a case would be invalid.