C / C++ / MFC

Writing the actual addresses of a struct in memory???

9-Jul-03 4:47

Assuming you are not wanting to read the entire file at once, there is still a maximum of 3GB of address space available. Just a guess, but I'd say that you'll need to read the file in chunks.

johnstonsk9-Jul-03 4:22

johnstonsk

9-Jul-03 4:22

Let me see if I can explain this with shortness and clearity.

I am having to hard code the layout of a struct in memory.

Meaning that I have to hard code the actual memory location of each element in the struct.

Lets just say for simplicity that I have to start my first element at 0x0000.

The struct is layed out like this:

struct data<br />
{<br />
   double min[4];<br />
   double max[4];<br />
   char  name[4];<br />
   int   count;<br />
<br />
}

In order to create the struct in memory I think I need to create the layout of a byte(short word) like so:

typedef union u_do_word{<br />
  unsigned short word;<br />
  struct do_word_bits{<br />
    unsigned pad  : 8;<br />
    unsigned bit7 :1;<br />
    unsigned bit6 :1;<br />
    unsigned bit5 :1;<br />
    unsigned bit4 :1;<br />
    unsigned bit3 :1;<br />
    unsigned bit2 :1;<br />
    unsigned bit1 :1;<br />
    unsigned bit0 :1;<br />
  }bits;<br />
} do_word_t;

And create the pointer and variables for the elements
I put the hex address in the name.

start of the min array

<br />
  static do_word_t word_0x00;<br />
  static unsigned short *vme_word_0x00; //1st double in the min array<br />
  static do_word_t word_0x20;<br />
  static unsigned short *vme_word_0x20; //2nd double in the min array<br />
  static do_word_t word_0x40;<br />
  static unsigned short *vme_word_0x40; //3rd double in the min array<br />
  static do_word_t word_0x80;<br />
  static unsigned short *vme_word_0x80; //4th double in the min array

start of the max array

<br />
  static do_word_t word_0xA0;<br />
  static unsigned short *vme_word_0xA0; //1st double in the max array<br />
  static do_word_t word_0xC0;<br />
  static unsigned short *vme_word_0xC0; //2nd double in the max array<br />
  static do_word_t word_0xE0;<br />
  static unsigned short *vme_word_0xE0; //3rd double in the max array<br />
  static do_word_t word_0x100;<br />
  static unsigned short *vme_word_100; //4th double in the max array

start of the char array

This is the part that I am not sure about.
I believe that an unsigned short will give me the correct size for a double, but how should I do the char array.
Is this correct???

static do_word_t word_0x120;<br />
  static unsigned char *vme_word_0x120; //1st char in the name array <br />
<br />
  static do_word_t word_0x?????;   (??  0x124 ??)<br />
  static unsigned char *vme_word_0x?????; //2nd char in the name array <br />
<br />
  static do_word_t word_0x?????;    (??  0x128 ??)<br />
  static unsigned char *vme_word_0x?????; //3rd char in the name array <br />
<br />
  static do_word_t word_0x?????;     (??  0x12C ??)<br />
  static unsigned char *vme_word_0x?????; //4th char in the name array

I am not sure how to step through the memory and create the correct size for the char array or the int at the end

Thanks for taking the time to read my winded message,
Steven

Re: Writing the actual addresses of a struct in memory???

John M. Drescher9-Jul-03 4:35

John M. Drescher

9-Jul-03 4:35

The size of a double is 8 bytes. Use the sizeof(double) to see that your self. I have no clue why you would want to do this. Also if you don't pack the data using #pragma pack(push,1) before your struct and #pragma pack(pop) after your struct you will get unexpected results.

John

type "double" question

DaveE9th9-Jul-03 4:11

9-Jul-03 4:11

I have used type "double" for the elements in my array. My numbers are 10 digits long, so I can't use the int type. My numbers don't have any decimals in them, but the output from my program is displaying decimals on the monitor. eg. number 2647584936 will display as 2647584936.000000

How do I get rid of the decimal and the six zeros?

Thanks Hmmm | :|

BoudewijnEctor9-Jul-03 4:14

BoudewijnEctor

9-Jul-03 4:14

No idea,

But i'd use the Int64 type instead of int. The max size of Int64 is 2^64 --> Seems big enough.

DaveE9th9-Jul-03 4:25

9-Jul-03 4:25

I tried to enter the "int64" data type into my VC++ 6.0 compiler, but it doesn't turn blue as do other valid prototypes. How do I implement the int64 data type? It sounds very interesting.

Thanks

Cedric Moonen9-Jul-03 4:30

Cedric Moonen

9-Jul-03 4:30

Use long integer (long int) instead! This should work: range is:
–2,147,483,648 to 2,147,483,647 . Or you can even use unsigned long int if you don't need negative numbers.

David Crow9-Jul-03 4:34

9-Jul-03 4:34

It's a Microsoft-specific type, so you must preface it with two underscores.

keegan9-Jul-03 4:19

keegan

9-Jul-03 4:19

look for SetPrecision on the msdn library.

or look here.[^]
*.*

Trollslayer9-Jul-03 4:23

Trollslayer

9-Jul-03 4:23

User the correct type specifier in your print, e.g.

<br />
str.Format("%5.0f", d)

where str is a CString and d is the double variable.

Elaine Rose | [Rose]

The tigress is here Big Grin | :-D

DaveE9th9-Jul-03 4:32

9-Jul-03 4:32

Thanks Elaine, I was hoping you might show me how to apply the code you mentioned into my program below?

<br />
#include "stdio.h"<br />
#include "math.h"<br />
#include "iostream.h"<br />
#include "stdlib.h"<br />
<br />
<br />
double values[] = {<br />
1115364759770,<br />
3748596078661, <br />
3849573800596,<br />
3849573846696,<br />
4615243374995,<br />
4745520008574,<br />
4748020008574,<br />
4847020005584,<br />
4847020007584, <br />
5758020005007,<br />
5758020005287,<br />
7557030006574, <br />
7937030006574,<br />
9500020004758, <br />
9585520004758, <br />
};<br />
<br />
<br />
int compare (const void * a, const void * b)<br />
{<br />
#define ABS(a) ((a) < 0 ? -(a) : (a) ) 	<br />
double _a = *(double*)a;   <br />
double _b = *(double*)b;   <br />
if( ABS(_a - _b) < 0.00000001 )      <br />
return 0;   <br />
else if( _a < _b)<br />
return -1;   <br />
else <br />
return 1;<br />
}<br />
int main ()<br />
{   <br />
double * pItem;  <br />
cout<<"please enter a number"<<endl; <br />
double key;<br />
cin>>key; <br />
pItem = (double*) bsearch (&key, values, 15, sizeof (double), compare);  <br />
if (pItem!=NULL)      <br />
printf ("%lf is in the array\n",*pItem);<br />
else      <br />
printf ("%lf is not in the array\n",key);   <br />
return 0;<br />
}<br />

Dave

David Crow9-Jul-03 4:37

9-Jul-03 4:37

printf("%.0f is in the array\n", *pItem);<br />
printf("%.0f is not in the array\n", key);

Ryan Binns9-Jul-03 4:25

Ryan Binns

9-Jul-03 4:25

Instead of printing them with '%f', print them with '%10.f'. This will get rid of the decimal point and the digits following it.

Hope this helps,

Ryan

"Punctuality is only a virtue for those who aren't smart enough to think of good excuses for being late" John Nichol "Point Of Impact"

DaveE9th9-Jul-03 4:53

Natural Language Processing open source

9-Jul-03 4:53

Thanks everybody, the %10.f and %.0f worked great.

Smile | :)

JoeSox9-Jul-03 3:48

JoeSox

9-Jul-03 3:48

Does anyone know of any Natural Language Processing open source for
C++?

at this point in my search, I will take any suggestions, ~~even a link to a free VB source~~ nah forget that. But maybe Java.
I have posted a question to Google answers, and they are trying to find me one too but no answer yet
Hopefully someone here knows of something I could use for the NLP part for Carl v2 I am working on, currently. Thanks.

sorry for the repost but I thought I might be able to reach some more people

Later,
JoeSox
www.humanaiproject.org
"Dream as if you'll live forever; live as if you'll die tomorrow."
- James Dean(ISTP)

Loading ComboBox within a thread

cycosi9-Jul-03 3:46

cycosi

9-Jul-03 3:46

Hi I have the following problem.
In my application, I have a window in which I have several comboboxes.
When accessing the data on the server to fill them, The application was really slowed down.
So I decided to have all the loading in a thread.
Meanwhile everything is loading, I set the text for the combos.
It is OK for all of them, except one, in which all the data inferior the the text disappear from the combo.
If I load the data out of the thread everything is fine.
Any idea??

Re: Loading ComboBox within a thread

David Crow9-Jul-03 4:00

Custom message in ClassWizzard

9-Jul-03 4:00

So the primary thread, which owns the comboboxes, spawned a secondary thread.. Is that right? Does the secondary thread try to populate the comboboxes (if so, the secondary thread will need a message pump), or does it post a message back to the primary thread (if so, remember to use PostMessage() and not SendMessage())?

To eliminate the secondary thread as a culprit, comment out the code that creates it, and have the primary thread populate the comboboxes. If the problem persists, you know the secondary thread is not to blame.

Cedric Moonen9-Jul-03 3:23

Cedric Moonen

9-Jul-03 3:23

Hi!
I have designed a custom control (a real-time chart) and I want to associated a function handler (in the class of the window wich uses the chart) when the user clicks on the legend. So, in the class of the custom control, I added a handler for the WM_LBUTTONDOWN message and I check there if the mouse is on the legend.
If yes, I send a message to the parent window but I want this custom message appear in the ClassWizzard (like for example: OnLegendClick).
Is such a thing possible and how to implement it ??

Thanks

Easy question on DHTML Editing control

Scozturk9-Jul-03 3:15

Scozturk

9-Jul-03 3:15

Hi!
I am using the DHTML editng control to make a WYSIWYG application.
Well my problem is that I don't know how to insert some HTML tags into the code.
The dhtml editing control gives you some functions for fonts, style,alignement,etc... but the user may need to add tables too.
I want to put a button which will open a table configuration dialog... the dialog will put all the values in a string... ( I can do it until here) and it will put the generated html tags into the html code...
Can anyone help me? Confused | :confused:

Thank you very much for your future answers!

Well... I am a beginner ...

Re: Automatic overwriting in setup

David Crow9-Jul-03 3:15