AVCajipe's Blog

Web Developer’s Resolutions for 2010.

2010-01-01T01:29:00Z

1. Break down in .css files in a logical manner so I can better locate a selector / class id instead of using the ubiquitous CTRL-F.

2. Get into the habit of using LINQ and Entity Framework more.

3. Create more logical business objects with properties and methods instead of directly using data tables and datasets directly from the database. Stricter separations of concern.

4. Continuous refactoring and unit testing.

5. Review JavaScript functions and see which items can be executed easier in JQuery. Makes learning css selectors easier too.

6. Start using ADO.net Data Services / Astoria.

7. Practice using Code Rush Xpress, learn more features.

8. Blog more. Share more cool links like this one.

9. Listen to more podcasts for continuing education.

10. Do more performance tuning, debugging and other tracing techniques.

Group Chat 2007 SDK Development Environment Setup Demystified

2009-12-28T06:16:00Z

Microsoft’s Office Communications Server 2007 API exposes properties and methods in building custom chartrooms on top of Group Chat R2. However from the download site, they state that in order to install the API you should have the following Operating Systems: Windows Server 2003; Windows Server 2008. They also mentioned to:

Refer to the documentation contained within UCGCSdk.msi for additional information regarding system prerequisites.

Which is I think is quite funny because of the supposed need to install the .msi package without mentioning these additional system requirements before-hand.

Furthermore, this MSI package cannot install in a development machine not running a server operating system. Oh the hilarity.

Saving you from further headaches and as a public service, here are the requirements and steps we took in setting up the development environment for the Group Chat 2007 API. These are the requisites that the download website does not tell you:

1.     Operating System – At least Windows Server 2003 64-bit. Yes, that’s sixty-four-bit. We attempted installing the .msi package on a 32-bit server OS but that didn’t work out. We had to request for a virtual server for our development team in order to progress.
2.     Install Visual Studio 2008 in the server development machine.
3.     Install the Group Chat API.

Accessing our server development machine was funny because we had to remote desktop the server every time we need to start on chat room development. Yes, the hassle.
As of now my development team is starting to learn how to use and build on top of OCS2007 R2 platform for our BOT systems. I’ll write more on this as we go along.

For those interested, I’ve included the Group Chat .dlls and the compiled html help file so that you can take a quick look at the contents without the need experiencing the hassle of setting up the above mentioned environments.

Ideas that keep me going

2009-12-01T14:20:00Z

Flash, WPF, SilverLight?

Variation...

On Feature Management

2009-12-01T03:25:00Z

I’ve got an existing application. It’s running smoothly and I want to keep it as streamlined as possible. I’ve spent countless hours in testing and debugging. It’s not perfect but I try to make it as close. I do not want it bloated, I find convoluted UIs quite cumbersome, but business users are our main customers, they will continuously want features. They’ll ask for the moon if they could. Where do we draw the line? In my area of work being a developer and also acting as business analyst, I’ve learned to accept feature requests if they would fall under the following requisites:

1. If the request will not just be used "once off" or just for one time usage,
e.g. "create a feature that can generate me the end of year financial statement for 2009"

2. If the request is requested by more than one person in more than one different team. It doesn’t mean that if a stakeholder keeps asking for the same feature over a period of time that it’s a good idea.

3. Feature request is useful to everyone that uses the system, not just useful to a handful of people. Maximize ROI.

4. Does not slow down or complicate any existing feature.

Practical C#9 - Export/Import to and from CSVs & DataTables

2009-09-22T14:21:00Z

Another one of the ultra handy helper classes I use for exporting and importing CSVs. End users can perform all sorts of bulk web form requests via a simple text/csv file thereby improving a website’s usability.

Highlighted are the method signatures. This function however, cannot work in the context of an <asp:UpdatePanel because of the usage of the HttpResponse object. If however, lots of web controls or given that the page itself is encapsulated in the update panel, we have to set the buttons to behave synchronously.

<asp:UpdatePanel ID="updatePanel_CSV" runat="server">

<%--Response object must be synchronous--%>

<asp:PostBackTrigger ControlID="but_importCSV" />

<asp:PostBackTrigger ControlID="but_exportToCSV" />

</Triggers>

<div runat="server" id="div_upload"...

using System;

using System.Collections.Generic;

using System.Data;

using System.IO;

using System.Text.RegularExpressions;

using System.Web;

public class CXUtil_CSVParser

{

public void ExportToCSV(string CSV, string fileName)

{

HttpContext.Current.Response.Clear();

HttpContext.Current.Response.ClearHeaders();

HttpContext.Current.Response.ClearContent();

HttpContext.Current.Response.AddHeader("content-disposition",

"attachment; filename=" + fileName + ".csv");

HttpContext.Current.Response.AddHeader("Pragma", "public");

HttpContext.Current.Response.ContentType = "text/csv";

HttpContext.Current.Response.Write(CSV);

HttpContext.Current.Response.End();

}

public string ConvertDataTableToCSV(DataTable table)

{

string result = "";

for (int x = 0; x < table.Columns.Count; x++)

{

result += x == 0 ? "" : ",";

result += GetDataTableCell(table.Columns[x].ColumnName);

}

result += Environment.NewLine;

foreach (DataRow row in table.Rows)

{

for (int x = 0; x < table.Columns.Count; x++)

{

result += x == 0 ? "" : ",";

result += GetDataTableCell(row[x].ToString());

}

result += Environment.NewLine;

}

return result;

}

public DataTable ConvertCSVtoDataTable(Stream stream)

{

DataTable table = new DataTable();

try

{

StreamReader reader = new StreamReader(stream);

List<string> header = GetCsvRow(reader);

foreach (string column in header)

{

table.Columns.Add(column);

}

while (!reader.EndOfStream)

{

List<string> cells = GetCsvRow(reader);

if (cells.Count > 0)

{

DataRow row = table.NewRow();

bool isRowEmpty = true;

for (int x = 0; x < cells.Count && x < header.Count; x++)

{

row[header[x]] = cells;

if (!string.IsNullOrEmpty(cells[x]))

{

isRowEmpty = false;

}

if (!isRowEmpty)

{

table.Rows.Add(row);

}

catch (Exception ex)

{

throw new Exception("Cannot parse IO stream: " + ex.Message);

}

return table;

}

public bool IsValidString(String strToCheck)

{

bool charIsValid = true;

string[] reserved_words = {

"select", ">",

"select * from",

"drop table",

"insert into",

"exec", "insert", "insert into", "update", "delete", "drop"};

for (int x = 0; x < reserved_words.Length; x++)

{

if (strToCheck.IndexOf(reserved_words[x]) != -1)

{

charIsValid = false;

}

return charIsValid;

}

public bool IsValidDate(string dateString)

{

bool result = true;

try

{

Convert.ToDateTime(dateString);

}

catch (FormatException fe)

{

result = false;

}

return result;

}

#region .. private methods ..

private List<string> GetCsvRow(StreamReader reader)

{

List<string> row = new List<string>();

bool inQuote = false;

string cell = "";

int i = reader.Read();

while (i != -1)

{

char c = Convert.ToChar(i);

if (!inQuote && c == ',')

{

row.Add(cell.Trim().Trim('\"'));

cell = "";

}

else if (!inQuote && c == Environment.NewLine[0])

{

string newLineCheck = "" + c;

bool isNewLine = true;

for (int j = 1; j < Environment.NewLine.Length; j++)

{

int newLineCharInt = reader.Read();

if (newLineCharInt == -1)

{

isNewLine = false;

break;

}

char newLineChar = Convert.ToChar(newLineCharInt);

newLineCheck += newLineChar;

if (newLineChar != Environment.NewLine[j])

{

isNewLine = false;

break;

}

if (isNewLine)

{

row.Add(cell.Trim().Trim('\"'));

break;

}

cell += newLineCheck;

}

else

{

if (c == '\"')

{

inQuote = !inQuote;

}

cell += c;

}

i = reader.Read();

}

return row;

}

private string GetDataTableCell(string content)

{

string result = content.Trim();

if (result.IndexOf(',') > -1 ||
result.IndexOf(Environment.NewLine) > -1)

{

result = '\"' + result + '\"';

}

return result;

}

#endregion

}

Laptop Envy

2009-09-21T00:22:00Z

I was in Toronto, Canada last May 2008 out to buy a new laptop @ the Future Shop.

Nothing left to say save that they’re cheap and they’re fast! CAN$1199 + 12% tax (Php 57,878.12) already earns you an HP Pavilion 17” Centrino Core2 Duo with 3GB of Memory; 320HD and some cutting edge graphics NVIDIA GeForce 8400M GS Graphics card. Oh, and did I already mentioned that this was back in 2008? By the way, for our conversion, 1CAN$ = Php43.10.

This was the laptop I bought.

This machine also comes with its own remote control (dubbed as “entertainment pc”) and w/ it’s built-in Altec Speakers. Here are the other specs.

I cashed in another $40 something dollars (Php1,724) for an additional 1GB ram to give my machine a 4GB memory bringing the total laptop cost to Php59,602.12.

And oh and did I mention this was still in 2008?

Fast forward to September 2009. Here is the closest laptop that I came across villman’s site.

Nice machine, great hardware art, wish mine had that too but that’s right kids, it's Php88,944.00.
Checking on the specs:

To be honest I saw nothing fancy save maybe for the Blu-Ray ROM and the newer chipset, for a price difference of an estimate of Php30,000? Was the art expensive? Sounds to me you can already buy another machine with that. I'm now absolutely convinced that laptops being sold here are ridiculously expensive, and to make things worse are stripped down (reduced hard disk and reduced memory, etc). Never mind the plane fares. With due respect, I don’t think I’ve really made a fair comparison because 2009 should be compared to 2009.

But then again, I would be comparing 18.4” laptops with 8GB memory, 64-bit OS, TB capacity hard drives to the ones I see now in Villaman’s, PCExpress, etc. Common people, where’s the justice?

Oh and did I mention that my laptop was still in 2008?

Practical C#8 - Displaying Timezones

2009-08-18T08:11:00Z

Writing a truly global application would at times require the developer in displaying the different time zones in plain text format.

My previous implementation was to capture the offset in the client and pass this value to a server-side hidden field control which can be read from the server.

The disadvantage of this is that I have to perform offset value calculation on the server side to obtain the local time zone where the page can possibly be viewed.

protected void Button1_Click(object sender, EventArgs e)
    {
        DateTime utcNow = DateTime.UtcNow;
        Int32 offsetValue = -(Convert.ToInt32(hf_offsetValue.Value));
        label_globalTime.Text =
           utcNow.AddMinutes(offsetValue).ToShortDateString() + " "
            + utcNow.AddMinutes(offsetValue).ToShortTimeString();

}

Furthermore, this code is difficult to test and validate without someone testing it for you residing in a different time zone.

However, this blog does a great job in explain the differences (theoretical and practical) of the TimeZone and the TimeZoneInfo namespaces in net.
Building up from the concepts learned in the blog, I’ve written a sample page that conveniently displays and enumerates the different time zones relative to your location using the little known namespace TimeZoneInfo.

using System;
using System.Collections.Generic;
using System.Collections;
using System.Collections.ObjectModel;
using System.Linq;
using System.Web;
using System.Web.UI;
using System.Web.UI.WebControls;

public partial class _Default : System.Web.UI.Page
{
    protected void Page_Load(object sender, EventArgs e)
    {
        displayTimeZones();
    }

    void displayTimeZones()
    {
        DateTimeOffset nowDateTime = DateTimeOffset.Now;
        Response.Write("<strong>Now in Manila: </strong>" + nowDateTime);
        Response.Write("<br /><br /><br />");

ReadOnlyCollection<TimeZoneInfo> timeZones =
TimeZoneInfo.GetSystemTimeZones();

        Response.Write("<strong>ALL TIMEZONES: </strong>");
        foreach (TimeZoneInfo indiTimeZoneInfo in timeZones)
        {
            DateTimeOffset newDateTime =
                 TimeZoneInfo.ConvertTime(nowDateTime,
                     TimeZoneInfo.FindSystemTimeZoneById(indiTimeZoneInfo.Id));
            Response.Write("<br />");
            Response.Write("<strong>Display Name</strong>: " +
                    indiTimeZoneInfo.DisplayName + "<br />");
            Response.Write("<strong>Now in " + indiTimeZoneInfo.StandardName + ":
                    </strong>" + newDateTime);
            Response.Write("<br />");
        }
    }
}

With the resulting page as:

Why engineers and sales don't mix

2009-08-18T07:36:00Z

dilbert.com

Practical Javascript 8: Extending Checkboxes

2009-08-04T02:52:00Z

Input checkboxes can be used for all sorts of selection and deselection activities. They can also become extremely useful when used within a grid for multiple item selection. In this example, checked items (or rather rows) are coloured differently when selected. The colouring happens on the client to do away with needless postbacks.

A client-side JavaScript was attached to each checkbox as the GridView was rendered. The rendering happens in the gv_RowDataBound event of the gridview. The client-side function was written this way:

function colorCheckBoxOnCheck(CB, originalColor, proposedColor, rowIndex, GV)

{

var checkBox = document.getElementById(CB.id);

if (document.getElementById(GV).rows[rowIndex] != null) {

var cells = document.getElementById(GV).rows[rowIndex].cells;

var targetColor = "";

if (checkBox.checked == true)

targetColor = proposedColor;

else

targetColor = originalColor;

for (var i = 0; i < cells.length; i++) {

cells[ i].style.background = targetColor;

}

The parameters are the checkbox object, the original and proposed colour(the new colour of the row when checked). The 0-based row index was also used to target the specific row the checkbox belonged to and the GridView object itself. Targeting the specific row (composed of columns/cells) on the client would be making this call:

var cells = document.getElementById(GV).rows[rowIndex].cells;

Wherein we made use of the GridView object and row index. The GridView declaration is straightforward:

<asp:GridView

SkinID="RX_AdminTableNoHeader"

ID="gv"

Width="97%"

runat="server"

onrowdatabound="gv_RowDataBound">

<asp:TemplateField>

<asp:CheckBox

id="cb"

runat="server" />

</ItemTemplate>

</asp:TemplateField>

<asp:TemplateField>

<asp:Label ID="lbl_someData"

runat="server"

Text='<%#bind("SOME_DATA") %>'/>

</div>

</ItemTemplate>

</asp:TemplateField>

</Columns>

</asp:GridView>

Finally tying the client-side code on the GridView server event onRowDataBound:

protected void gv_RowDataBound(object sender, GridViewRowEventArgs e)

{

if (e.Row.DataItem != null)

{

DataRowView drv = e.Row.DataItem as DataRowView;

switch (e.Row.RowState)

{

case DataControlRowState.Normal:

case DataControlRowState.Normal | DataControlRowState.Alternate:

CheckBox cbTaskGroup = e.Row.FindControl("cb")
as CheckBox;

string gvClientID = ((GridView)sender).ClientID;

string originalColor = "";

if (e.Row.RowIndex % 2 == 0)

originalColor = "#EFF3FB";

else

originalColor = "#FFFFFF";

string proposedColor = "#D9E9C5";

string RowIndex = Convert.ToString(e.Row.RowIndex);

string cb_jsCall = "return colorCheckBoxOnCheck(this," +

" '" + originalColor + "', '" + proposedColor + "', "

+ RowIndex + ", '" + gvClientID + "');";

cb.Attributes["onClick"] = cb_jsCall;

break;

}

The required parameters are conveniently provided as server-side properties. The row index is represented by the reference of “e” as the instance of the event handler. Since the GridView has alternating colours per row, we capture their initial colours alternatively. After composing our JavaScript call we hook it up to the onClick client-Side event of the checkbox.

Practical Javascript 7: PageRequestManager's Begin and End Events

2009-07-20T08:25:00Z

Scripts handlers has worked for me so far because they define in the page processing life cycle those previously obscure phases where (we wished in the past) we could perhaps execute some client-side scripts. These are the PageRequestManager Events. The MSDN documentation defines all the events but I’d like to give an example how I used begin and end event handlers.

The required elements for enabling this event would be to ajaxify the page a web page using a ScriptManager and an UpdatePanel. The panel would contain the button that would trigger the event.

Coding the scripts would be simply putting this script below the ScriptManager tag.

function requestEndHandler(sender, args)
{

//client-side script that execute after the asynchronous postback

}

function requestBeginHandler(sender, args)

{

//client-side script that execute before the asynchronous postback

}

Sys.WebForms.PageRequestManager.getInstance().add_endRequest(requestEndHandler);

Sys.WebForms.PageRequestManager.getInstance().add_beginRequest(requestBeginHandler);

</script>

A common usage for this I do is to disable all submit buttons (in one loop run of the form elements) on a page while the asynchronous postback is happening.

function requestBeginHandler(sender, args)

{

//client-side script that execute before the asynchronous postback

affectButtons(true, null);

}

function affectButtons(_isDisabled, _exemptedButtons) {

var THE_FORM = document.getElementById('aspnetForm');

for (var x = 0; x < THE_FORM.elements.length; x++)

{

if (THE_FORM.elements[x].type == 'submit') {

if (_exemptedButtons != null) {

for (var j = 0; j < _exemptedButtons.length; j++) {

if (THE_FORM.elements[x].id != _exemptedButtons[j].ToString()) {

document.getElementById(THE_FORM.elements[x].id).disabled
= _isDisabled;

}

else {

//no exemptions, disable all buttons

document.getElementById(THE_FORM.elements[x].id).disabled
= _isDisabled;

}

}//end of for loop

}

Once the script enters requestEndHandler, the buttons are automatically enabled.

Practical Javascript 6: Validating Text Input Length

2009-07-20T07:56:00Z

Another common validation task that can be more successful if implemented on the client is the length restriction of inputs. Off the bat, we have this handy property in an order textbox.

<asp:TextBox ID="TextBox1" runat="server" MaxLength="256" />

The rendering engine would produce:
<input name="TextBox1" type="text" maxlength="256" id=" TextBox1" />

This is all well and good if we use our textbox just the way it is. But hilarity would soon ensue once we start adding this property if we wanted to render a text area.

<asp:TextBox ID="TextBox1" runat="server" MaxLength="256" TextMode="MultiLine" />

Corresponding output would be:
<textarea name="ctl00$TextBox1" rows="2" cols="20" id=" TextBox1"></textarea>

Not only MaxLength="256" wouldn’t work anymore but it’s completely gone. The first HTML was an input form with type=”Text” and the latter was a textarea form element. Maximum length appears not to be a native property of a textarea.

I’ve developed a short client-side script (because I enjoy churning these out) to help in validation of text input lengths applicable to both single and multi-line text fields.

function checkCharacterLimit(TB)

{

var maxCharacters = 256;

var withinCharacterLimit = true;

var ValidateTB = document.getElementById(TB.id);

if (ValidateTB.value.length > maxCharacters)

{

TB.focus();

withinCharacterLimit = false;

} return withinCharacterLimit;

}

The script is attempting to return the cursor (via the focus()) function back to the textfield once the input exceeds the desired maximum length which is 256 characters. Hooking this up to the input form elements, we take advantage of the onKeyUp and the onKeyDown client-side events of keyboard events.

TextBox1.Attributes["onKeyUp"] = "return checkCharacterLimit(this);";

TextBox1.Attributes["onKeyDown"] = "return checkCharacterLimit(this);";

More Proof that I can be lazy

2009-07-17T13:55:00Z

Likes: Having a stored procedure that attempts to do everything, as in the whole CRUD salad.
Dislikes: Having to create lots of separate stored procedures if I can the first option is feasible.
Thank God for dynamic queries for that.

* Name : MY_LAZY_SP

* Purpose : multiple crud operations for group/division/department

* History

* Date Version Modified By Comments

* =========== ======= =========== ========

* >> SELECT OPERATION

-------------------

exec dbo.MY_LAZY_SP

@TABLE_affected = 'GROUP',

@OPERATION = 'select',

@REF_ID = 0,

@REF_VALUE = '',

@WILDCARD = 'some wildcard'

>> UPDATE OPERATION

-------------------

exec dbo.MY_LAZY_SP

@TABLE_affected = 'DIVISION',

@OPERATION = 'update',

@REF_ID = 2,

@REF_VALUE = 'my new value',

@WILDCARD = ''

>> INSERT OPERATION

-------------------

exec dbo.MY_LAZY_SP

@TABLE_affected = 'DEPARTMENT',

@OPERATION = 'insert',

@REF_ID = 0,

@REF_VALUE = 'some new value',

@WILDCARD = ''

CREATE PROCEDURE dbo.MY_LAZY_SP

@TABLE_affected VARCHAR(25), --GROUP/DIVISION/DEPARTMENT

@OPERATION VARCHAR(20), --select, update, insert, delete

@REF_ID INT,

@REF_VALUE VARCHAR(256),

@WILDCARD VARCHAR(256)

BEGIN

DECLARE @sql VARCHAR(4000),

@table_name VARCHAR(10)

------------------------------------------------------------------

IF (@OPERATION = 'select')

BEGIN

DECLARE @columnName VARCHAR(5)

IF (@TABLE_affected = 'GROUP')

SET @columnName = 'GRP'

ELSE IF (@TABLE_affected = 'DIVISION')

SET @columnName = 'DIV'

ELSE IF (@TABLE_affected = 'DEPARTMENT')

SET @columnName = 'DEP'

SELECT @sql = ' SELECT '+@TABLE_affected+'_ID, '+@TABLE_affected+'_NAME '

SELECT @sql = @sql + 'FROM RX_REF_'+@TABLE_affected+' '

SELECT @sql = @sql + ' WHERE '+@TABLE_affected+'_ID <> 1'

IF (@WILDCARD <> '')

BEGIN

SELECT @sql = @sql +' AND UPPER('+@TABLE_affected+'_NAME)
like ''%'+ UPPER(@WILDCARD) +'%'' '

END

SELECT @sql = @sql + ' ORDER BY '+@TABLE_affected+'_NAME '

END

------------------------------------------------------------------

ELSE IF (@OPERATION = 'update')

BEGIN

SELECT @sql = ' DECLARE @countExistingUpdate INT '

SELECT @sql = @sql + ' SELECT @countExistingUpdate =
count('+@TABLE_affected+'_NAME) from RX_REF_'+@TABLE_affected+' '

SELECT @sql = @sql + ' WHERE UPPER('+@TABLE_affected+'_NAME) =
UPPER('''+@REF_VALUE+''') '

SELECT @sql = @sql + ' AND '+@TABLE_affected+'_ID <> '+
CONVERT(varchar(20),@REF_ID) +' '

SELECT @sql = @sql + ' IF (@countExistingUpdate = 1) '

SELECT @sql = @sql + ' BEGIN '

SELECT @sql = @sql + ' RAISERROR 100001 '''+@TABLE_affected+' name
already exists.'' '

SELECT @sql = @sql + ' END '

SELECT @sql = @sql + ' ELSE '

SELECT @sql = @sql + ' BEGIN '

SELECT @sql = @sql + ' UPDATE RX_REF_'+@TABLE_affected+' '

SELECT @sql = @sql + ' SET '+@TABLE_affected+'_NAME =
'''+@REF_VALUE+''' '

SELECT @sql = @sql + ' WHERE '+@TABLE_affected+'_ID = '+
CONVERT(varchar(20),@REF_ID) +' '

SELECT @sql = @sql + ' END '

END

------------------------------------------------------------------

ELSE IF (@OPERATION = 'insert')

BEGIN

SELECT @sql = ' INSERT INTO RX_REF_'+@TABLE_affected+' ('+@TABLE_affected+'_NAME)'

SELECT @sql = @sql + ' VALUES ('''+@REF_VALUE+''')'

SELECT @sql = ' DECLARE @countExistingInsert INT '

SELECT @sql = @sql + ' SELECT @countExistingInsert =
count('+@TABLE_affected+'_NAME) from RX_REF_'+@TABLE_affected+' '

SELECT @sql = @sql + ' WHERE UPPER('+@TABLE_affected+'_NAME) =
UPPER('''+@REF_VALUE+''') '

SELECT @sql = @sql + ' IF (@countExistingInsert = 1) '

SELECT @sql = @sql + ' BEGIN '

SELECT @sql = @sql + ' RAISERROR 100001 '''+@TABLE_affected+' name
already exists.'' '

SELECT @sql = @sql + ' END '

SELECT @sql = @sql + ' ELSE '

SELECT @sql = @sql + ' BEGIN '

SELECT @sql = @sql + ' INSERT INTO RX_REF_'+@TABLE_affected+'
('+@TABLE_affected+'_NAME)'

SELECT @sql = @sql + ' VALUES ('''+@REF_VALUE+''')'

SELECT @sql = @sql + ' END '

END

------------------------------------------------------------------

ELSE IF (@OPERATION = 'delete')

BEGIN

SELECT @sql = ' DELETE FROM RX_REF_'+@TABLE_affected+' '

SELECT @sql = @sql + ' WHERE '+@TABLE_affected+'_ID = '+
CONVERT(varchar(20),@REF_ID) +' '

END

exec (@sql)

END

Where the passionate geeks go

2009-07-15T14:37:00Z

May 24, 2008 saturday morning, EnergizeIT 2008. Toronto Congress Center, Canada. Not a rock concert event but a conference for geeks, dorks and everything in between.

C# Quirk 1: Pretty Useless Method

2009-07-14T08:29:00Z

Since asp.net 1.0/1.1, I enjoyed storing my settings in web.config because it’s both fast and convenient.

Reading them is straightforward enough:
ConfigurationManager.AppSettings["activeEnvironment"].ToString();

Since .net 2.0/3.5 I wanted to do more than read settings, I wanted to be able to modify them on the fly.
I thought this method would be convenient:
ConfigurationManager.AppSettings.Set("activeEnvironment", "prod");

The documentation was also simple:
// Summary:
//     Sets the value of an entry in the System.Collections.Specialized.NameValueCollection.
//
// Parameters:
//   name:
//     The System.String key of the entry to add the new value to. The key can be
//     null.
//
//   value:
//     The System.Object that represents the new value to add to the specified entry.
//     The value can be null.
//
// Exceptions:
//   System.NotSupportedException:
//     The collection is read-only.

I developed an administration page that handles system settings modification by
calling ConfigurationManager.AppSettings.Set(key, value)

Since calling this static method, I was able to notice the website behave appropriately
according to the settings that I had set.

I did however notice that the setting value at web.config did not change
as I initially assumed. Were was the method storing it?

Here were also some interesting feedback in using this method. The answer is cache.

Why did the .net framework team design it this way? Pretty useless method don’t you think?

Anyway, ranting obviously isn’t going to solve my problem. Such case cannot work in an enterprise level web application,
so I designed my own and simple Settings.xml file. I wanted it lightweight.

<?xml version="1.0" encoding="utf-8"?>
<Settings>
<setting>
<id>activeEnvironment</id>
<value>dev</value>
</setting>
</Settings>

The next piece of the puzzle was to design a helper class that would read and write the value, another opportune time for me to practice Linq to Xml.
Since I am lazy and wanted to take advantage of intellisense support, I created an enumeration object for the settings key.

public enum WebConfigSettings
{
activeEnvironment
}

What follows is the helper class definition:

using System;
using System.Collections.Generic;
using System.Linq;
using System.Web;
using Resources;
using System.Xml;
using System.Xml.Linq;

/// <summary>
/// Reads and updates XML settings migrated from web.config
/// </summary>

public static class xmlSettings
{
private static string settingsXmlFile =
HttpContext.Current.Server.MapPath("~/Settings.xml");

    public static string readSetting(WebConfigSettings s)
    {
        string setting = "";
        IEnumerable<string> xmlQuery =
               from AllSettings in XElement.Load(settingsXmlFile).Elements("setting")
               where (string)AllSettings.Element("id") == s.ToString()
               select (string)AllSettings.Element("value");

        foreach (string xmlSetting in xmlQuery)
            setting = xmlSetting;
        return setting;
     }

public static void saveSetting(WebConfigSettings s, string proposedValue)
{
        string xPath = "/Settings/setting[id='" + s.ToString() + "']/value";
        XmlDocument xmlDocument = new XmlDocument();
        xmlDocument.Load(settingsXmlFile);
        XmlNode node = xmlDocument.SelectSingleNode(xPath);

        //remove existing value
        node.ParentNode.RemoveChild(node);

        //save in correct node
        string xPathSettingsNode = "/Settings/setting[id='" + s.ToString() + "']";
        XmlNode settingsNode = xmlDocument.SelectSingleNode(xPathSettingsNode);

        //create new value
        XmlElement ElementSetting = xmlDocument.CreateElement("value");
        ElementSetting.InnerText = proposedValue;
        settingsNode.AppendChild(ElementSetting);

        //save Xml file
        xmlDocument.Save(settingsXmlFile);
    }
}

XPath is a query language best learned by example. XPath is a rich language in its own right, with a bezy of functions.

My usage for xPath was:
string xPath = "/Settings/setting[id='" + s.ToString() + "']/value";

This means I wanted to traverse the node “setting” and then get the “value” element given the “id” element as a parameter.
<setting>
<id>activeEnvironment</id>
<value>dev</value>
</setting>

Calling the read static method:
string activeEnvironment = xmlSettings.readSetting(WebConfigSettings.activeEnvironment);

Thanks to the enum, I do not have to remember the setting’s identifier.

Calling the save static method:
xmlSettings.saveSetting(WebConfigSettings.activeEnvironment,TextBox1_settings.Text);

This time, my custom settings are now stored in disk.

Practical C#9 - Simple Asynchronous Implementations

2009-07-11T22:56:00Z

2 ways I came up with implementing “fire-and-forget” asynchronous calls:

Method 1:

using System.Threading;
...

//might be binded in a button-click

protected void but_click(object sender, eventArgs e)

{

if (ThreadPool.QueueUserWorkItem(new WaitCallback(myLongRunningMethod), "abc"))

txtResult.Text = "sucess";

}

private void myLongRunningMethod(object sender)

{

//long running method, could be a long database query/update, etc

}

Method 2:

<%@ Page
Language="C#"

< ... >
Async="True"
%>

protected delegate void myAsyncTaskDelegate();
private myAsyncTaskDelegate _func;

public IAsyncResult CustomBeginHandler(object sender, EventArgs e, AsyncCallback cb, object state)
{
_func = new myAsyncTaskDelegate(longMethod);
IAsyncResult ar = _func.BeginInvoke(cb, state);
return ar;
}

private void longMethod()
{
//..long operation here
}

public void CustomEndHandler(IAsyncResult ar)
{
_func.EndInvoke(ar);
}

Calling the asynchronous task delegates would mean making a function invocation call AddOnPreRenderCompleteAsync(new BeginEventHandler(CustomBeginHandler), new EndEventHandler(CustomEndHandler)); probably in the Page_Load or a Button_ClickEvent.

The method is documented here and with further references.