|
I have the following code that converts a lower case alphabetical character (a-z) to upper-case (A-Z):
if (letter >= 97 && letter <= 122)
{
ascii -= 32;
letter = (char)ascii;
}
The problem is I think the cast for "letter" is affecting the performance as I am looping through a text of around 5 million characters.
Someone suggested a "lookup table" but I do not know what it is , possibly a Dictionary type?
Can anyone offer some advise for this problem?
Thanks! 
|
|
|
|
|
Why not just use ToUpper ?
I know the language. I've read a book. - _Madmatt
|
|
|
|
|
.ToString().ToChar() could takes up allot of performance?
Thanks anyway :P
|
|
|
|
|
I didn't say to use ToString.
Char.ToUpper[^]
I know the language. I've read a book. - _Madmatt
|
|
|
|
|
System.Char.ToUpper(x)
Can't really get any faster than that. No need to convert to string first.
Hope this helps
|
|
|
|
|
Ok I will use that, beats the casting anyway :P
|
|
|
|
|
thebuzzwright wrote: Someone suggested a "lookup table" but I do not know what it is , possibly a Dictionary type?
An array of chars, indexed by the char you want to uppercase
Also, I thought that comparisons on chars were perfectly valid?
The cast for letter shouldn't matter that much. But why are "ascii" and "letter" not the same variable?
|
|
|
|
|
Ah thanks for the description of a lookup table 
I have changed the code to:
if (_letter < 65 || _letter > 122) continue;
if (_letter >= 97 && _letter <= 122)
{
_letter = char.ToUpper(_letter);
}
Performance wise its slightly better as expected, I didn't know char had a "ToUpper()"
Thanks for the help
|
|
|
|
|
You don't need the ifs at all; ToUpper performs them itself so you're performing them twice -- and thereby hurting performance.
"The uppercase equivalent of c, or the unchanged value of c, if c is already uppercase or not alphabetic."
|
|
|
|
|
Snap!  You should never use standby on an elephant. It always crashes when you lift the ears. - Mark Wallace
C/C++ (I dont see a huge difference between them, and the 'benefits' of C++ are questionable, who needs inheritance when you have copy and paste) - fat_boy
|
|
|
|
|
| Performance wise, it may be worth taking out the tests to see if you need to do the conversion - it may well be that the .NET ToUpper is more efficient than yours, and since it will be done anyway whenever ToUpper is executed, you may find it quicker without testing - particularly if you are looking at normal text where nearly all characters are lower case to start with. You should never use standby on an elephant. It always crashes when you lift the ears. - Mark Wallace
C/C++ (I dont see a huge difference between them, and the 'benefits' of C++ are questionable, who needs inheritance when you have copy and paste) - fat_boy
|
|
|
|
|
The code shown translates to a few instructions only, so it takes a fraction of a microsecond to execute; if your text handling seems slow, look at the larger picture (and explain and show us more). Unless you need complex calculations and decisions, file input/output is likely to dominate the performance.
 Luc Pattyn [Forum Guidelines] [Why QA sucks] [My Articles]
I only read code that is properly formatted, adding PRE tags is the easiest way to obtain that. All Toronto weekends should be extremely wet until we get it automated in regular forums, not just QA.
|
|
|
|
|
using System;
using System.Text;
using System.IO;
namespace CharacterReader
{
unsafe class Program
{
static readonly char[] Letter = new char[26];
static string _fileName = "lol.txt";
[STAThread]
static void Main()
{
using (System.Windows.Forms.OpenFileDialog open = new System.Windows.Forms.OpenFileDialog())
{
if (open.ShowDialog().Equals(System.Windows.Forms.DialogResult.OK))
{
_fileName = open.FileName;
}
}
uint *value = stackalloc uint[26];
Load();
Console.WriteLine("Reading File...");
ReadFile(_fileName);
Console.WriteLine("Analysing file...");
AnalyseFile(value);
Console.WriteLine("Done!");
GetTotal(value);
Console.ReadKey();
}
static int _index;
static int _i;
static void Load()
{
for (_i = 65; _i < 91; _i++)
{
Letter[_index] = (char)_i;
_index++;
}
_index = 0;
}
const int BufferSize = 512;
static readonly StringBuilder Contents = new StringBuilder(99999);
static private void ReadFile(string filename)
{
FileStream fileStream = new FileStream(filename, FileMode.Open, FileAccess.Read);
StreamReader streamReader = new StreamReader(fileStream);
char[] fileContents = new char[BufferSize];
int* ptr = stackalloc int[1];
*ptr = streamReader.Read(fileContents, 0, BufferSize);
while (*ptr > 0)
{
Contents.Append(fileContents);
*ptr = streamReader.Read(fileContents, 0, BufferSize);
}
streamReader.Close();
fileStream.Close();
}
static void AnalyseFile(uint* value)
{
char* charPtr = stackalloc char[1];
for (_i = 0; _i < Contents.Length; _i++)
{
for (int k = 0; k < Letter.Length; k++)
{
*charPtr = Contents[_i];
if (*charPtr < 65 ) continue;
if (*charPtr >= 97 && *charPtr <= 122)
{
*charPtr = char.ToUpper(*charPtr);
}
if (!(*charPtr).Equals(Letter[k])) continue;
(value[k]) += 1;
break;
}
}
}
static void GetTotal(uint* value)
{
uint* biggestValue = stackalloc uint[1];
for (_i = 0; _i < 26; _i++)
{
if (*biggestValue >= value[_i]) continue;
*biggestValue = value[_i];
}
for (_i = 0; _i < 26; _i++)
{
if (!(*biggestValue).Equals(value[_i])) continue;
_index = _i;
break;
}
Console.WriteLine("The most common char is {0} with {1} hits", Letter[_index], *biggestValue);
}
}
}
Has anyone got some tips to make it run faster, it seems the "AnalyseFile" method seems to be slower than the file input 
|
|
|
|
|
That looks horribly inefficient. Please replace CODE tags by PRE tags (so it gets formatted properly), then I will look into it.
Luc Pattyn [Forum Guidelines] [Why QA sucks] [My Articles]
I only read code that is properly formatted, adding PRE tags is the easiest way to obtain that. All Toronto weekends should be extremely wet until we get it automated in regular forums, not just QA.
|
|
|
|
|
Here is an updated version of the code, it runs very fast which is suprising, the small project was actually quite a good learning experiance for me (being a student), one of the things I learnt was that higher performance in code seems to be achieved by decrementing coding standards
using System;
using System.Text;
using System.IO;
using System.Diagnostics;
namespace CharacterReader
{
unsafe class Program
{
static readonly char[] Letter = new char[26];
static string _fileName = "lol.txt";
[STAThread]
static void Main()
{
using (System.Windows.Forms.OpenFileDialog open = new System.Windows.Forms.OpenFileDialog())
{
if (open.ShowDialog().Equals(System.Windows.Forms.DialogResult.OK))
{
_fileName = open.FileName;
}
}
Stopwatch stopWatch = new Stopwatch();
uint* value = stackalloc uint[26];
Load();
Console.WriteLine("Reading File...");
stopWatch.Start();
ReadFile(_fileName);
Console.WriteLine("Analysing file...");
AnalyseFile(value);
Console.WriteLine("Done!");
GetTotal(value);
stopWatch.Stop();
Console.WriteLine(stopWatch.Elapsed.TotalMilliseconds);
Console.ReadKey();
}
static int _index;
static int _i;
static void Load()
{
for (_i = 97; _i < 123; _i++)
{
Letter[_index] = (char)_i;
_index++;
}
_index = 0;
}
const int BufferSize = 512;
static readonly StringBuilder Contents = new StringBuilder(99999);
static private void ReadFile(string filename)
{
FileStream fileStream = new FileStream(filename, FileMode.Open, FileAccess.Read);
StreamReader streamReader = new StreamReader(fileStream);
char[] fileContents = new char[BufferSize];
int* ptr = stackalloc int[1];
*ptr = streamReader.Read(fileContents, 0, BufferSize);
while (*ptr > 0)
{
Contents.Append(fileContents);
*ptr = streamReader.Read(fileContents, 0, BufferSize);
}
streamReader.Close();
fileStream.Close();
}
static void AnalyseFile(uint* value)
{
char* charPtr = stackalloc char[1];
for (_i = 0; _i < Contents.Length; _i++)
{
*charPtr = Contents[_i];
if ((*charPtr < 65) || (*charPtr > 90 && *charPtr < 97) || (*charPtr > 122)) continue;
*charPtr = char.ToLower(*charPtr);
value[*charPtr - 97] +=1;
}
}
static void GetTotal(uint* value)
{
uint* biggestValue = stackalloc uint[1];
for (_i = 0; _i < 26; _i++)
{
if (*biggestValue >= value[_i]) continue;
*biggestValue = value[_i];
}
for (_i = 0; _i < 26; _i++)
{
if (!(*biggestValue).Equals(value[_i])) continue;
_index = _i;
break;
}
Console.WriteLine("The most common char is {0} with {1} hits", Letter[_index], *biggestValue);
}
}
}
Thanks for your help :P
|
|
|
|
|
OK, PRE tags have turned that into a readable code snippet.
I see you got rid of the nested for loops; there was indeed no reason to have two nested loops.
Why are you still using that char pointer?
I would give the technique below a try:
static void AnalyseFile(uint* value) {
foreach(char c in Contents) value[(int)c]++;
foreach(char uc in "ABCDEFGHIJKLMNOPQRSTUVWXYZ") {
value[(int)uc.ToLower()]+=value[(int)uc];
}
comments:
1. performance optimization like these may be relevant only for large amounts of data
2. it has a different use of the value array
3. it will fail for non-ANSI text (when character values exceed 255).
 Luc Pattyn [Forum Guidelines] [Why QA sucks] [My Articles]
I only read code that is properly formatted, adding PRE tags is the easiest way to obtain that. All Toronto weekends should be extremely wet until we get it automated in regular forums, not just QA.
|
|
|
|
|
Here is the final version of the code and I feel I have beaten it to death:
using System;
using System.Text;
using System.IO;
namespace CharacterReader
{
unsafe class Program
{
static readonly byte[] Letter = new byte[26];
static string _fileName;
[STAThread]
static void Main()
{
using (System.Windows.Forms.OpenFileDialog open = new System.Windows.Forms.OpenFileDialog())
{
if (open.ShowDialog().Equals(System.Windows.Forms.DialogResult.OK))
{
_fileName = open.FileName;
}
}
uint* value = stackalloc uint[26];
Load();
Console.WriteLine("Reading File...");
ReadFile( _fileName);
Console.WriteLine("Analysing file...");
AnalyseFile(ref value);
Console.WriteLine("Done!");
GetTotal(ref value);
Console.ReadKey();
}
static int _index;
static int _i;
static void Load()
{
for (byte i = 97; i < 123; i++)
{
Letter[_index] = i;
_index++;
}
_index = 0;
}
const int BufferSize = 512;
static readonly StringBuilder Contents = new StringBuilder(99999);
static private void ReadFile(string filename)
{
FileStream fileStream = new FileStream(filename, FileMode.Open, FileAccess.Read);
StreamReader streamReader = new StreamReader(fileStream);
char[] fileContents = new char[BufferSize];
int* ptr = stackalloc int[1];
*ptr = streamReader.Read(fileContents, 0, BufferSize);
while (*ptr > 0)
{
Contents.Append(fileContents);
*ptr = streamReader.Read(fileContents, 0, BufferSize);
}
streamReader.Close();
fileStream.Close();
}
static void AnalyseFile(ref uint* value)
{
byte charPtr;
for (_i = 0; _i < Contents.Length; _i++)
{
charPtr = (byte)Contents[_i];
if (charPtr < 65) continue;
if (charPtr > 122) continue;
if (charPtr > 90 && charPtr < 97) continue;
if (charPtr < 97)
{
value[charPtr - 65] += 1;
}
else
{
value[charPtr - 97] += 1;
}
}
}
static void GetTotal(ref uint* value)
{
uint* biggestValue = stackalloc uint[1];
for (_i = 0; _i < 26; _i++)
{
if (*biggestValue >= value[_i]) continue;
*biggestValue = value[_i];
}
for (_i = 0; _i < 26; _i++)
{
if (!(*biggestValue).Equals(value[_i])) continue;
_index = _i;
break;
}
Console.WriteLine("The most common char is {0} with {1} hits", (char)Letter[_index], *biggestValue);
}
}
}
I read the entire English bible in under 6ms !
|
|
|
|
|
how to run a php code from your program?
like just execute an already existing php code from a remote server just like being a normal internet user requesting that the php is executed but without the need of any php-generated interface,just taking advantage of the server-side characteristics of php.
i wonder,can c# take some details that could be returned by php,maybe picking up info from cookies left on the host machine? or maybe some session details?
and another thing: can c# force a server to run a code generated right from the c# interface,maybe transferred through ftp?
thanks alot in advance!
|
|
|
|
|
Don't ask the same question twice. Continue this on the previous thread.
But, since you added some other questions:
andyxfun wrote: just like being a normal internet user requesting that the php is executed but without the need of any php-generated interface
You don't need a php-generated interface. The php script can be executed from a web page.
andyxfun wrote: ,just taking advantage of the server-side characteristics of php.
php doesn't know about, nor care about, if it's running on a server or a workstation. The machine it's running on looks the same as any other.
andyxfun wrote: i wonder,can c# take some details that could be returned by php,maybe picking up info from cookies left on the host machine?
Without a web server, there are no cookies.
andyxfun wrote: or maybe some session details?
No web server = no session.
andyxfun wrote: can c# force a server to run a code generated right from the c# interface,maybe transferred through ftp?
No. As I said in your previous thread, running code on a remote box is a VERY HIGH SECURITY RISK. This is not something you normally would do.
|
|
|
|
|
ok,yeah,altmost the same question but the details make the difference: in this one i take the idea of any general php script,with the possibility of adding some new scripts using your application.
the cookies and sessions thing i asked there was referring to the fact that when you visit a webpage, this can put some cookies and some session info on your computer's temp folder and i know for a fact that both php and javascript are able to read/write these cookies and i was curious if c# may be capable of the same thing,giving me a good medium for transferring data between the php files present on the server and my project from my local computer.
as you advised me in your other comment, i will try google-ing it and see what comes!
thanks a lot, you even raised my moral!
|
|
|
|
|
Hi
I want to know all Operating System's usernames who are working under a particular domain.
I need to bind it to Dropdown..
Thanks in Advance..
|
|
|
|
|
And?? I assuming this is Active Directory, correct? You have to provide a bit more detail than what you already have. What code have you tried? Where did you run into a problem? Any error messages?
|
|
|
|
|
hi!
please tell me, can i request the running of a .php file located on a server via a C# program and maybe return some info some way?
i am really good at php but still suck at c# and need to finish a project for school and this would save alot of time...
thanks a lot in advance!!!
|
|
|
|
|
It's possible, so long as the .php code you're running does not put up a user interface or require user interaction at all. Look into the WMI Win32_Process class and you'll find that you can launch remote processes, with some restrictions. Running code on a remote box is a HUGE security risk, so it's not going to be a simple little process to do.
|
|
|
|
|
it is all for a school project so security is not one of my problems...
anyway,i only need php to do some basic operations on serverside like connecting to databases, adding-retrieving information and so on,and i understand that i cannot use any user interface.
thanx for the info, awaiting some more exact instructions.
i am grateful,though!
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
