|
Djavid j wrote: Hoi,
Nog es bedankt en ik heb jou een mailtje gestuurd. Zou je die even lezen aub!?
I'm sorry, but you seem to have got your keyboard a bit jammed up here. Perhaps you could keep this board operating in it's default English only setting. Anything else causes the hamsters to go into attack mode.
|
|
|
|
|
That is the whole idea, gremlins make the site run faster. 
Luc Pattyn [Forum Guidelines] [My Articles]
This month's tips:
- before you ask a question here, search CodeProject, then Google;
- the quality and detail of your question reflects on the effectiveness of the help you are likely to get;
- use PRE tags to preserve formatting when showing multi-line code snippets.
|
|
|
|
|
Luc Pattyn wrote: gremlins make the site run faster.
Screaming in terror I assume.
|
|
|
|
|
I presume your config file is in the same directory as the executable? Do you tell it this, or does it just "find" the file the first time?
The issue here is that you're relying on the config being in the applications "current" directory (which it is on startup, since the current directory is defaulted to the exe's folder)
When you use OpenFileDialog it changes the current directory to the location of the file you browsed to, meaning the config can't be found any more.
My suggestion would be to use System.IO.GetCurrentDirectory[^] on startup, and use this when loading the config.
e.g. something like:
m_ApplicationDirectory = System.IO.GetCurrentDirectory();
m_Configfilename = m_ApplicationDirectory + "Config.xml";
...
XmlDocument configfile = new XmlDocument();
configfile.Load( m_Configfilename );
|
|
|
|
|
Hi i'm new in remoting and since that have any common question about
1) How i can understand in server part what some client is expired on client part(detroyed by gc)
2) How my client (on client part) can understand what server is no longer exists (i.e. remote instance of this client to what one delegate ones calls no longer exists)???
TANHK A LOT
|
|
|
|
|
I have never seen either happen but it seems likely that Exceptions would be thrown in both cases
|
|
|
|
|
Good morning,
Say I have various types of messages to send across the network, but obviously I can only send them as an object. How do I filter where the various messages are sent when they get to the server?
I've been going around in circles, and the obvious way (in my head) to handle this is to have some sort of header (using an interface), and then do a switch depending on what's in the header, and cast the object to the specific type.
But something's telling me this is a bad way of doing it, it doesn't feel very OO. any suggestions?
Alex
Sorry if this is in the wrong forum 
|
|
|
|
|
I am building a C# control for charting and i want to implement zooming functionality in it. After hitting zoom button in my control, i want to show new form with ennlarged selected area. For this, i need to get data from usercontrol into a new window form. Please tell me how to access data from usercontrol into windows form or is there any other way to implement zooming functionality?
Thanking in advance......
Dhok
|
|
|
|
|
If you are already displaying the chart using the System.Drawing.Graphics, you can use a ScaleTransform[^] for zooming. The two parameters to the ScaleTransform method are the amount of scaling to be done along the x and y axes, respectively.
|
|
|
|
|
Hi,
i'm using the windows media player as activeX object in a C#
application. I want to change the "Video border color"
(Windows Media Player->Menu Tools->Options->Performance->Advanced)
in my application. I cannot find the right options in the msdn library
to change the color from my application.
It's because i want to change the background color of the media player
if a video is changed during playback.
Can please someone tell me how to to this?
regards,
Lucas
|
|
|
|
|
Hi
I was wondering if there are any way to count the occurences of a bitpattern within a textstring using C#. If ANYONE can help me out with this one I would be eternal grateful.
For instance. If I have the text in a string named str1, I want to check this for every occurence of 11's or perhaps 00's.
|
|
|
|
|
int pos = str1.IndexOf("1");
int pos = str1.indexOf("1",5)
A combination of those two methods should have you finding each occurance of 1 (or 0) in no time
|
|
|
|
|
I think you misinterpreted my question. I am not after every occurence of the letter "1" but the bitpattern...like one letter is 8 or 16 bits.
|
|
|
|
|
If you want to search for a particular bit pattern within a string, you will first need to convert string to a byte array, this can be done using System.Text.ASCIIEncoding object.
After converting string to byte array we may search this array for a particular bit pattern. Following is a code snippet which will give you some idea about this-
---------------------Code Start-------------------
byte[] bytes;
byte pattern1 = 111;
string str = "hello there how are you";
System.Text.ASCIIEncoding getencoding = new ASCIIEncoding();
//convert string to byte array
bytes = getencoding.GetBytes(str);
int i;
for (i = 0; i < str.Length; i++)
{
if (bytes.GetValue(i).Equals(pattern1))
{
MessageBox.Show("");
}
}
------------------------Code End---------------
I hope this helps .
-Dave.
Dave Traister,
ComponentOne LLC.
www.componentone.com
|
|
|
|
|
But if I have a byte with 8 bits and another consecutive byte with another 8 bits, how can I check this entire range (16 bits) for only a 11 at any location?
for instance:
11000000 00000000 = 1 time
01110000 00000000 = 2 times
etc...
there is no fixed number representing the 11 since it can move around the byte...
|
|
|
|
|
i guess the crude way would be:
convert your pattern from byte [] to bool []
store the sequence you're searching for as bool[]
loop through the sequence until you match the first value then check the second value of the sequence etc. This isn't going to be quick but it'll work OK for short strings
Russ
|
|
|
|
|
Hmmm its kind of unfortunate that is should be so complicated and slow because its for a (non-profit) research project where performance is everything...
|
|
|
|
|
If you are passing long pieces of binary data about the place as strings you have to be prepared for different people to mean different things by the same number. AFAIK ASCII leaves the choice of numbers above 127 to the person using them. While there are conventions for numbers 127-255 there are more than 1 so it may be that your data gets mangled on the way. .Net languages often try to make sense of this chaos by translating the string from one codepage to another; while this might be good for someone trying to read the string at the end it's the last thing you would want to happen to your data.
If you are going to search the same string repeatedly then maybe you could do something clever with trees, thus putting the effort in at the start when you build the tree and getting payback through faster searching later.
I would use the brute force method to start and make sure it's wrapped up in a tidy method so that you can come back and improve it later or try to find a better way of doing the same thing that doesn't require an exhaustive search of a list of bytes.
Russell
|
|
|
|
|
invictus3 wrote: performance is everything...
then you need to tell more about the application, and I can come up with several suggestions.
furthermore there are some basic questions:
- is the length of the search pattern fixed? small? always 2? always less than or equal to 8?
- how do characters (or bytes) join together, is it most-significant-bit first, so
msb of byte1 connects to lsb of byte0, or the other way around?
- do you need exact location of a match? number of matches? first match only?
- what if the search pattern is periodic, i.e. how often do you recognize 0101 in 01010101
(2 or 3 times?)
BTW: IMO this topic belongs in the math&algo forum, not the C# forum.
Luc Pattyn [Forum Guidelines] [My Articles]
This month's tips:
- before you ask a question here, search CodeProject, then Google;
- the quality and detail of your question reflects on the effectiveness of the help you are likely to get;
- use PRE tags to preserve formatting when showing multi-line code snippets.
modified on Tuesday, January 15, 2008 7:10:53 AM
|
|
|
|
|
Ok here it goes:
- I can not tell more about the application at this point. At this point it is only at the planning stage where we test the possibility. Getting this algorithm right is a prerequisite for any further work.
- For sake of simplicity lets assume that we want to count for all 2-bit (00, 01, 10, 11)(and as a variant perhaps also 3-bit (000, 001, etc...)) combination. It is fixed sized pattern we want to search for, but the string we are searching in is variable sized.
- Whether it is MSB or LSB is irrelevant how I see it. It is all C# strings on a intel laptop, and that is pretty much what I know about that.
- I do not need location: only the count of each pattern
- The pattern you described would be counted 3 times. It will overlap locations already counted.
To be honest I don't want a complete code, but only pointers on how to approach the bit pattern count problem on my own. So far I got 2 possible solutions:
- convert each char into a binary string of 8 bytes where each letter is either 0 or 1 and then to regular string comparison
- use unsafe code and the & operator
Neither seems to be that efficient.
In Java I already solved this problem in a kind of inefficient way: first converting to bytes, then into a binary-string (e.g. each character from the original string takes up 8 chars with either 0/1 in the binary string) and then just do a normal string comparison (e.g. str.equals("01"). However, this is 1) inefficient and 2) in Java when C# is the targeted language. Unfortunately I suck at C# at this point but are reading up on it as we speak.
|
|
|
|
|
Hi,
here is my suggested algorithm:
- looking for a pattern of length P;
- have an accumulator that holds a sliding part of the binary data stream; one way to do
that is to initialize it to zero, then inside a loop shift it left over B positions and
ORing B bits of new data (B could be 8 or 16, the accumulator's bit count must
equal or exceed B+P-1);
- have a method that counts the number of pattern matches in the rightmost B+P-1 bits
of the accumulator, i.e. ignore the higher bits in the accumulator;
doing so makes sure every match will be counted only once.
There are basically two ways to do the count method:
1. inside a loop say for(x=0; x<b; x++),="" mask="" away="" all="" but="" p="" bits="" and="" compare="" those<br="" mode="hold"> with the pattern; so have a mask initially equal to (1<<P)-1, AND with accumulator,
compare with pattern; then shift left mask and pattern and compare again, etc.
2. if B+P-1 is small, say less than 16, you can precalculate an array with the
count value for every possible accumulator value, so the count method is just
a single array lookup.
Remarks:
- even if you think of your data as 16-bit items (say Unicode characters), you can still
work with B=8 if you choose to do so, the algorithm does not care what the bit
stream means;
- the sliding accumulator works best if the most-significant-bit of data item 1
connects to the least-significant-bit of data item 0; otherwise it will take
more complex load-and-shift operations.
- the algorithm requires B+P-1 not to exceed the length of the largest available integer
type; otherwise "big integer" techniques become necessary;
- precalculating an array is a common way to drastically improve performance provided
there is sufficient data to process; here the array's dimension is only 1<<(B+P-1)
which is 512 for B=8 and P=2.
- obviously when B+P-1 becomes large, allocating and precalculating an array may become
prohibitive and the loop-implementation may be prefered.
- when the pattern size increases AND the count of all possible pattern values is
required, it may not be obvious what the best order of execution would be:
either first loop over the pattern values, and scan the data multiple times,
or allocate 1<<P arrays and do all pattern values in parallel, scanning the data
only once.
Luc Pattyn [Forum Guidelines] [My Articles]
This month's tips:
- before you ask a question here, search CodeProject, then Google;
- the quality and detail of your question reflects on the effectiveness of the help you are likely to get;
- use PRE tags to preserve formatting when showing multi-line code snippets.
|
|
|
|
|
I started reading up on bit-operators in C# and came up with this solution:
static int[] countbits(string s)<br />
{<br />
int[] vector = { 0, 0, 0, 0 };<br />
<br />
for (int i = 0; i < s.Length; i++)<br />
{<br />
for (int j = 6; j >= 0; j--)<br />
vector[((s[i] >> j) & 0x03)]++;
if (i < s.Length - 1)<br />
vector[(((s[i] << 1) | (s[i + 1] >> 7)) & 0x03)]++;<br />
}<br />
<br />
return vector;<br />
}
Not sure if its the best way, but it works!
Any suggestions for improvement are welcome. Have optimized it some, but I wouldnt be surprised if there are better ways to solve this :p
Mvh Sverre
|
|
|
|
|
Hi,
from your approach I see you want to get the count for each possible 2-bit pattern at once,
which was not mentioned before.
I think the larger accumulator approach is more elegant and performant than the explicit
join statetement(s) your approach needs.
And I am confident an array-based solution (with precalculated counts) has the highest
performance, assuming the amount of data warrants the initial investment.
Both were introduced in my previous reply.
Luc Pattyn [Forum Guidelines] [My Articles]
This month's tips:
- before you ask a question here, search CodeProject, then Google;
- the quality and detail of your question reflects on the effectiveness of the help you are likely to get;
- use PRE tags to preserve formatting when showing multi-line code snippets.
|
|
|
|
|
Actually I was considering the suggestion about array of pre-calculated values, but I couldn't think of a good way of making the bits between the bytes (last bit from one byte and the first one from the next) count.
Not sure if I understand your larger accumulator approach though...
|
|
|
|
|
The simplest way to look at it is to have an accumulator that holds two consecutive characters,
which you achieve by shift-and-add or shift-and-or. The trick is to AND away the extraneous
bits that would cause some positions to be counted twice, hence the P+B-1 stuff.
Luc Pattyn [Forum Guidelines] [My Articles]
This month's tips:
- before you ask a question here, search CodeProject, then Google;
- the quality and detail of your question reflects on the effectiveness of the help you are likely to get;
- use PRE tags to preserve formatting when showing multi-line code snippets.
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
