Shoutbox

Looks complex

( 6 + 4 ) * 10 has 7 important tokens.
6 4 + 10 * has five important tokens.
Taking into account the stack-implentation of RPN, it is slightly simpler, especially from a program's POV.
* RaceProUK may use Reverse Polish for any maths questions on his next exam...

i don't see whats wrong with createing a quick text /calc
to be a /run link to the MS calculator

quote:

---

Originally posted by Orkblutt
/xcalc 666/77*8
=1.08                             ==> FALSE !!!

/xcalc (666/77)*8          ==> TRUE !!!
=69.20

---

That makes me wonder how you do the calculation and what method you use... because first of all in many language of today there is already an evaluation function for such things (dunno about C++ though). If you use that, you only need to pass the string and you would never have problems like that....

If you do seperate the numbers yourself then, again, this is some basic flaw in your algorithm, as it would be very easy to do such things in a home-made function:

* use a recursive function
* get the deepest linked parenthesis
* calculate the result of that:
  "*" and "/" coming first, then "+" and "-"
  Also calculate from LEFT to RIGHT!!! <== the error you made ( * or / doesn't take precedence over each other, just like + and - doesnt....)
    1) 666 / 77
    2) 8.64 * 8
  not:
    1) 77 * 8
    2) 666 / 616

* update the "string"
* exit the function and go up 1 level

This is a realy realy basic way in doing it (without optimazations), and you would never have a problem with parenthesis...

note about using "-": it is best that you don't implement substractions, but only additions! If a minus is used, it should be considered part of the number, not a math-function. If you do this you also wont have problems with: "8 - 5" because your function would calculate "8 + -5", but most importantly, you can do "8 * -5" without any additional coding...


So... I'm realy interested in the method you use...

PS:
http://www.codeproject.com/cpp/FastMathParser.asp
http://www.codeproject.com/cpp/RPNcalculator.asp
http://www.programmersheaven.com/zone3/cat415/30915.htm
http://www.informit.com/articles/article.asp?p=21484
http://trubetskoy1.narod.ru/english/ppne.html <= implement this and you also have a PPN calculator...

quote:

---

That makes me wonder how you do the calculation and what functions you use... because in most language of today (VB, C++, etc... ) there is already an evaluation function for such things. If you use that, you only need to pass the string and you would never have problems like that....

---

C++, VB do that ?

quote:

---

note about using "-": it is best that you don't implement substractions, but only additions! If a minus is used, it should be considered part of the number, not a math-function. If you do this you also wont have problems with: "8 - 5" because your function would calculate "8 + -5", but most importantly, you can do "8 * -5" without any additional coding...

---

have you test mine ?

quote:

---

So... I'm realy interested in the method you use...

---

humm it's a little bit complex to explain (I'm too bad in english!!)

Using a little more advanced parser tree method  than http://www.informit.com/articles/article.asp?p=21484 but it's the same principle...

Hey nice plugin works great

P.S. srry for the earlier confusion

you don't need to register it.

quote:

---

Originally posted by CookieRevised
http://trubetskoy1.narod.ru/english/ppne.html

---

This is probably the best of the links, as it describes an algorithm for converting infix to RPN, which you can then evaluate. That way, users can use the familiar infix, and you can implement the simpler RPN calculator.

thank you!
I'll see if it's necessary to use RPN (if I've too more difficulties with Infix)
thanks for this link