Re: Multiplatform (32bits/64bits/CE) & Interop - C# Discussion Boards

19-Jun-06 16:16

If I write a pure C# application, only using the standart publi API, no interop, the same binary should work well on32 bit, 64 bits and perhaps on the compact framework as well if I link against it.

Now how could I achieve the same thing if I do interop and I have a managed C++ API?

let say I have a managed C++ module compiled for 3 different target but with always the same interface:
mycppmodule32.dll
mycppmodule64.dll
mycppmoduleCE.dll

let say in the C# code I also do some interop
internal class User32 // & CE ?!
{
[DllImport("USER32")]
public IntPtr GetHdc(IntPtr hWnd); // 4 byte IntPtr
}
internal class User64 // not sure it exists, but I assume....
{
[DllImport("USER64")]
public IntPtr GetHdc(IntPtr hWnd); // 8 byte IntPtr
}

Let suppose these DLLs are included as modules in my csharpdll.dll.

Now if I deploy all these DLLs, how could write a simple nice code in my C# dll which internally target whatever native lib is appropriate?
Is it possible (to keep it simple)?

leppie19-Jun-06 22:01

leppie

19-Jun-06 22:01

Create a C# 'runtime' dll that you can conditionally compile. Then simply swap that out.

How xacc.ide transforms text to colored words on the screen
Intel PentuimM (aka Centrino) undervolting

Super Lloyd19-Jun-06 22:13

19-Jun-06 22:13

er...
thanks for answering!

what I was trying to do in a simple manner (if possible) is to have one library to run them all!

like imagine my program is a click once application, I would like (if possible) to avoid the x32 link, the x64 link, the CE link.
Just one link and it runs whatever the end-user machine.

Or also imagine the user "install" (i.e. copy) the file on one PC, and then he simply copy on an other PC and it sill works!

Conditional compilation is a clean way to have a build for a given platform.
But I wonder if there is a no hassel way to have one build for all platform that the user could simply copy without worrying about his pc's architecture....

you see what I mean?
is it possible? (in a simple way)

Note:
I did something like that on the PocketPC a while ago but it was akward.
Basically it was worked like that:

enum Platform
{
CE,
Win32,
}
public class OSUtil
{
internal static Platform Platform { get {} }

public Value AMethod()
{
switch(Platform)
{
case CE:
return OSUtilCE.AMethod();
case CE:
return OSUtilWin32.AMethod();
}
}
}
class OSUtilCE
{
static Value AMethod() {}
}
class OSUtil32
{
static Value AMethod() {}
}

Super Lloyd19-Jun-06 22:17

19-Jun-06 22:17

No!
It didn't work like that, too simple!
Don't forget runtim verification when the method is launched, it worked like that:

enum Platform
{
CE,
Win32,
}
public class OSUtil
{
internal static Platform Platform { get {} }

public Value AMethod()
{
switch(Platform)
{
case CE:
return OSUtilCE.AMethod();
case CE:
return OSUtilWin32.AMethod();
}
}
}
class OSUtilCE
{
[DllImport("CEDll")]
static Value AMethodImpl();
static Value AMethod() { return AMEthodImpl(); }
}
class OSUtil32
{
[DllImport("Win32Dll")]
static Value AMethodImpl();
static Value AMethod() { return AMEthodImpl(); }
}

Super Lloyd20-Jun-06 5:40

Microsoft Word/Excel Automation Errors [modified]

20-Jun-06 5:40

I know, I could do as I did, but generate the wrapper code automatically Cool | :cool:

XSL:, XMLDOM C#

Ramon Gzz M19-Jun-06 16:10

Ramon Gzz M

19-Jun-06 16:10

Hi All,

I currently have an xslt called from C#, this xslt calls(defines an xslt) that uses some jscript functions.

The jscript functions declares a new (microsoft.xmldom) object, this used to worked before when I had just installed msxml 4.0 sp2 in my server, after I installed sql 2005 which installed msxml6.0. this sentence takes too long to execute, sometimes it timeouts. I've commented this line an the xslt executes pretty fast, does anyone has a clue on what might be happening, this seems to happens only when the xslt is being called from C# code, I have several instances of microsoft.xmldom declaration running without any problems,

Any commments will be appreciated.

Thanks in advance.

dbloch19-Jun-06 15:19

dbloch

19-Jun-06 15:19

Hi All

I have a question relating to the resources/namespaces required when automating either to Excel or Word as I find that namespaces and resources that work perfectly on my machine with Office 2003, do not work with other machines who have Office 2003.

I'm having to type the full Microsoft.Office.Interop.Word/Excel namespace out which in the tutorials and other sample code I've seen, is not the case so I wonder if there is something wrong with my installation.

Here is a little program I created which takes all Word Documents within a directory and merges them into one large Word Document.

using System;<br />
using System.Collections.Generic;<br />
using System.ComponentModel;<br />
using System.Data;<br />
using System.Drawing;<br />
using System.Text;<br />
using System.Windows.Forms;<br />
using System.IO;<br />
using Microsoft.Office;<br />
<br />
 object oMissing = System.Reflection.Missing.Value;<br />
<br />
        public Form1()<br />
        {<br />
            InitializeComponent();<br />
<br />
<br />
        }<br />
<br />
        private void Form1_Load(object sender, EventArgs e)<br />
        {<br />
            lblSelect.Visible = false;<br />
<br />
        }<br />
<br />
<br />
        private void btnOK_Click(object sender, EventArgs e)<br />
        {<br />
            Microsoft.Office.Interop.Word._Document oDoc;<br />
            Microsoft.Office.Interop.Word._Document cDoc;<br />
            Microsoft.Office.Interop.Word.Application wordAppA = new Microsoft.Office.Interop.Word.Application();<br />
            Microsoft.Office.Interop.Word.Application wordAppB = new Microsoft.Office.Interop.Word.Application();<br />
            oDoc = wordAppA.Documents.Add(ref oMissing, ref oMissing, ref oMissing, ref oMissing);<br />
<br />
            folderBrowserDialog1.ShowDialog();<br />
            DirectoryInfo di = new DirectoryInfo(folderBrowserDialog1.SelectedPath.ToString());<br />
            FileInfo[] diFiles = di.GetFiles("*.doc");<br />
<br />
            lblSelect.Text = "Folder currently selected: " + folderBrowserDialog1.SelectedPath.ToString();<br />
            lblSelect.Visible = true;<br />
            <br />
            int count = 0;<br />
            foreach(FileInfo fi in diFiles)<br />
            {<br />
                count++;<br />
            }<br />
<br />
            progressBar1.Maximum = count;<br />
<br />
            count = 0;<br />
            foreach (FileInfo fi in diFiles)<br />
            {<br />
                object file = fi.FullName;<br />
                cDoc = wordAppB.Documents.Open(ref file, ref oMissing, ref oMissing, ref oMissing, ref oMissing, ref oMissing, ref oMissing, ref oMissing, ref oMissing, ref oMissing, ref oMissing, ref oMissing, ref oMissing, ref oMissing, ref oMissing, ref oMissing);<br />
                cDoc.Select();<br />
                wordAppB.Selection.Copy();<br />
                oDoc.Application.Selection.Paste();<br />
                oDoc.Application.Selection.InsertBreak(ref oMissing);<br />
                cDoc.Close(ref oMissing, ref oMissing, ref oMissing);<br />
                count++;<br />
                progressBar1.Value = count;<br />
            }<br />
            wordAppB.Quit(ref oMissing, ref oMissing, ref oMissing);<br />
            wordAppA.Visible = true;<br />
            Form1.ActiveForm.Close();<br />
        }<br />

Now this works absolutely perfectly on my machine without any fail what so ever but when I try to put it on any other machine even though we both have Office 2003, I get this following error

************** Exception Text **************
System.IO.FileNotFoundException: Could not load file or assembly 'Microsoft.Office.Interop.Word, Version=11.0.0.0, Culture=neutral, PublicKeyToken=71e9bce111e9429c' or one of its dependencies. The system cannot find the file specified.
File name: 'Microsoft.Office.Interop.Word, Version=11.0.0.0, Culture=neutral, PublicKeyToken=71e9bce111e9429c'

I can get around this by having it copy the Microsoft.Office.Interop.Word.dll into the source directory of my application and it will then work but this is a pain because I shouldn't need to do it.

Anyone got any ideas?

Thanks so much

David Bloch
Houston Medical

-- modified at 21:19 Monday 19th June, 2006

User settings for application

Mike0919-Jun-06 14:27

Mike09

19-Jun-06 14:27

I'm working on a windows application, in the main form I have a few custom controls, which are contained on a different project. In the main form a few user settings are set, and I need the controls to get those settings, but I can't get them. Is there a way that I can share user settings on different projects that belongs to the same application? I-m using visual studio 2005 and C#

Thanks in advance

Re: User settings for application

Mike Poz19-Jun-06 15:03

Mike Poz

19-Jun-06 15:03

Are they persisted settings? If so, have you considered setting/reading registry key values in HKCU\Software\{mycompany}\{mysettings\{setting a}, etc? Or if the settings are persisted in an INI file local to the application exe, reading in that (assuming the INI is updated real time) and getting the settings from there?

Mike Poz

How to get distinct items from an array?

mejobloggs19-Jun-06 13:34

mejobloggs

19-Jun-06 13:34

I have an ArrayList with numerical values in.

I want to be able to only have distinct items, either by getting rid of any duplicates, or copying single values over to a new array (whatever works).

e.g.

I have: 1, 2, 3, 3, 4, 5, 5, 5, 8, 9, 9
I want: 1, 2, 3, 4, 5, 8, 9

Unfortunately I can't figure out how to do this. Can someone please help?

Alexander Wiseman19-Jun-06 14:04

Alexander Wiseman

19-Jun-06 14:04

There a couple of ways you could attack this, depending on speed and space considerations. Here they are theoretically:

1) Sort the array list, then go through and call ArrayList.RemoveAt to chop out any duplicates in a simple loop through the sorted array.

2) Copy the ArrayList over to another ArrayList item by item, but each time, before you add an item, call ArrayList.Contains to check if the NEW array already contains the value you are trying to add from the OLD array.

3) Copy the ArrayList over to a Hashtable item by item, making the key the items in your array, and each time call Hashtable.ContainsKey to check if you've already added a value from the array.

These were a couple of the ways I could think of. There are probably others and more efficient, but these will get the job done. I wouldn't recommend using number 3 unless time is a real consideration. Number 2 is going to take much longer if you have a long array, because this will be looping through the array as many times as you have items. Number 1 can be longer or shorter depending on the sort algorithm you have.

Hope that helps!

Sincerely,
Alexander Wiseman

mejobloggs19-Jun-06 15:57

mejobloggs

19-Jun-06 15:57

Thanks!

For 2, I found a thing called ArrayList.BinarySearch. Would that make it more efficient?

E.g. Something like
myArrayListOne = original data;
myArrayListTwo = distinct data;

for (int i=0; i < myArrayListOne.Count-1; i++)
{
if (myArrayListTwo.BinarySearch(myArrayListOne[i] > 0))
{
myArrayListTwo.Add(myArrayListOne[i]);
}
}

Which probably won't work straight off, as it takes me a few tries to get things to work, but you get the idea.

So, would using BinarySearch like that be more efficient?
Also, are ArrayLists very effecient? Or should I be using, or converting the ArrayList to an Array as soon as possible?

leppie19-Jun-06 22:07

leppie

19-Jun-06 22:07

Alexander Wiseman wrote:
I wouldn't recommend using number 3 unless time is a real consideration.

What do you mean by that?

Alexander Wiseman wrote:
3) Copy the ArrayList over to a Hashtable item by item, making the key the items in your array, and each time call Hashtable.ContainsKey to check if you've already added a value from the array.

You dont need to use ContainsKey when adding in this case. You can simply use SetItem (indexer) for all the array elements.

How xacc.ide transforms text to colored words on the screen
Intel PentuimM (aka Centrino) undervolting

Alexander Wiseman20-Jun-06 5:43

Alexander Wiseman

20-Jun-06 5:43

leppie,

Thanks for your comments. I said I wouldn't recommend number 3 unless time was a consideration because I was unsure what overhead one might be pulling in by using a Hashtable to store the information. It seemed to me to be a roundabout way of doing it, and I wasn't comfortable with suggesting it as the most robust solution. On the other hand, it also seemed to me that using a Hashtable would be very fast, since looking up keys in a Hashtable is efficient. Please tell me if I'm not understanding something fundamental about Hashtables or whether you have a different opinion.

As regards your second point, I stand corrected. I should have realized that, in adding items to the Hashtable, one need not check if they are already there, because the same key cannot be repeated in the Hashtable!

Thanks again.

Sincerely,
Alexander Wiseman

Leslie Sanford19-Jun-06 16:43

Leslie Sanford

19-Jun-06 16:43

Do you have control over the items as they are being added to the ArrayList? If so, you could check the Contains method first to see if the ArrayList already contains the item.

Or sort the ArrayList and add the items to a new ArrayList. As you add the items, you sift out duplicates (untested):

public static ArrayList Filter(ArrayList items)
{
    #region Guard

    if(items.Count == 0)
    {
        return items;
    }
  
    #endregion

    items.Sort();

    object previousItem = items[0];
    ArrayList result = new ArrayList();

    result.Add(previousItem);

    foreach(object obj in items)
    {
        if(!previousItem.Equals(obj))
        {
            result.Add(obj);
        }

        previousItem = obj;
    }

    return result;   
}

mejobloggs19-Jun-06 17:09

mejobloggs

19-Jun-06 17:09

Ahh, true, thanks. I keep getting stuck in one little 'thought process', and don't think outside it sometimes.