Hi Community,
I'll be presenting the following MSDN Webcast (Saturday, 28th November 2009 @ 7:00 AM - Australia Time)
I look forward to seeing you there 
Regards,
Angel
Hola Comunidad,
Estaré presentando el siguiente MSDN Webcast (Viernes, 27 de Noviembre de 2009 (05:00 p.m. - Hora Colombia) - (05:30 p.m. - Hora Venezuela))
Espero verlos ese día
Saludos,
Angel
Hi Community,
Grabbed from Channel 9
Regards,
Angel
Hola Comunidad,
Tomado de Channel 9
Saludos,
Angel
Part I
Part II
Hi Community,
I just finished my presentation on dnrTV, thanks to Carl Franklin and Scott Hanselman. This session was about "Using Native Code in .NET" and we covered
-
Hooks
-
Memory Mapped Files
However, I'm planning to organize some other sessions to cover a bit more about this topic (e.g.: Virtual Memory, Device I/O, Error Reporting and Application Recovery and more). I'll keep you updated on this.
Please, feel free to download the source code attached to this post
Regards,
Angel
Hola Comunidad,
Acabo de terminar mi presentación para dnrTV, gracias a Carl Franklin y Scott Hanselman. El tópico de esta sesión fue "Using Native Code in .NET" y cubrimos
-
Hooks
-
Memory Mapped Files
Sin embargo, estoy planeando organizar otras sesiones para hablar un poco más de este tema (por ejemplo, Virtual Memory, Device I/O, Error Reporting and Application Recovery and more). Los mantendré informados al respecto.
Por favor, siéntanse libres de descargar el código adjunto a este post
Saludos,
Angel
A couple of nights ago I was reading about some of the new features in Windows 7 SDK and I can honestly say “They’re a lot”, as a matter of fact, I mentioned some of them in one of my previous Tech-Ed session this year. Despite of all these new features, there’s one that got my attention from the beginning and it was WWSAPI, because there wasn’t any support for Web Service from native code, except for a couple of existing toolkits including gSOAP that can be implemented on Windows, Linux and Mac OSX. WWSAPI was first introduced at PDC 2008 and it will be formally released along Windows 7, although previous versions of the operating system starting from XP SP2 can get it via Windows Update when becomes available. I first tested it when Windows 7 was still RC but I didn’t post about it because I was expecting for Windows 7 to be released or about to be released.
Having said that, last weekend I created a solution to demo WWSAPI in conjunction with .NET. The solution structure is depicted below
-
NativeTester: Console application (C++) which calls our Dynamic Link Library (DLL)
-
RSSFeedService: Web Service (C#) which retrieves RSS feeds, parse them and returns a well-formed XML
-
RSSViewer: WPF application (C#) which consumes the Web Service plus invokes our Dynamic Link Library (DLL)
-
Tester: WinForm application (C#) which consumes the Web Service plus invokes our Dynamic Link Library (DLL)
-
WWSAPIDemo: Dynamic Link Library (C++) which implements WWSAPI and it’s invoked from .NET via Interop
Along with the code you can find MSDN.xml which contains the RSS feeds from MSDN Australia, even when the code references http://localhost/MSDN.xml, you can change the Url and retrieve any given RSS, in my case I deployed the previously mentioned file to my local IIS.
Many of you might be wondering about, how can I generate a proxy based on the Web Service via C++? and the answer is quite simple, the new Windows 7 SDK provides us with an utility that does this for us, WSUTIL.exe. After executing it, we have as a result two files (an .H and a .CPP). The strings are treated as WCHAR* by default which in turn is Unicode and every single string in .NET are Unicode as well, I’ll comment a bit more about this later.
Having the Web Service published already on IIS, we better start working on the Dynamic Link Library which implements WWSAPI. Below we can see its header file
#include "stdafx.h"
#define EXPORT extern "C" __declspec(dllexport)
#define MAX_SIZE 128000 //128Kb
#define TRIM_SIZE 512
EXPORT WCHAR* GetFeeds(WCHAR* szUrl, int cbResults);
EXPORT void GetItem(WCHAR* szUrl, WCHAR* szItemName);
And this is the method responsible for connecting against the Web Service and retrieves the RSS feeds
// Get feeds from a given Url through WWSAPI
EXPORT WCHAR* GetFeeds(WCHAR* szUrl, int cbResults) {
WCHAR* temp = NULL;
WS_HEAP* heap = NULL;
WCHAR retval[MAX_SIZE]; // 128Kb (It should be enough to avoid WS_E_QUOTA_EXCEEDED error)
WS_ERROR* error = NULL;
ULONG propertiesCount = 0;
WS_SERVICE_PROXY* proxy = NULL;
WS_CHANNEL_PROPERTY channelProps[2];
WS_STRING serviceUrl = WS_STRING_VALUE(L"http://localhost/DemoSvc/RSSFeedService.asmx");
WS_ENDPOINT_ADDRESS endpoint = { serviceUrl};
WS_ENVELOPE_VERSION soapVersion = WS_ENVELOPE_VERSION_SOAP_1_1; // Our Webservice is WsiProfiles.BasicProfile1_1 compliant
WS_ADDRESSING_VERSION addressingVersion = WS_ADDRESSING_VERSION_TRANSPORT;
// Set channel's properties
channelProps[propertiesCount].id = WS_CHANNEL_PROPERTY_ENVELOPE_VERSION;
channelProps[propertiesCount].value = &soapVersion;
channelProps[propertiesCount].valueSize = sizeof(soapVersion);
propertiesCount++;
// Set addressing's properties
channelProps[propertiesCount].id = WS_CHANNEL_PROPERTY_ADDRESSING_VERSION;
channelProps[propertiesCount].value = &addressingVersion;
channelProps[propertiesCount].valueSize = sizeof(addressingVersion);
propertiesCount++;
// Can we create an WsError and WsHeap objects?
if (SUCCEEDED(WsCreateError(NULL, NULL, &error)) && SUCCEEDED(WsCreateHeap(MAX_SIZE, TRIM_SIZE, NULL, NULL, &heap, error))) {
// Can we create a proxy based on the service?
if (SUCCEEDED(WsCreateServiceProxy(WS_CHANNEL_TYPE_REQUEST, WS_HTTP_CHANNEL_BINDING, NULL, NULL, NULL,
channelProps, propertiesCount, &proxy, error))) {
// Can we open the proxy object?
if (SUCCEEDED(WsOpenServiceProxy(proxy, &endpoint, NULL, error))) {
// If we're able to invoke the service then copy the results to another variable because if
// we don't we lose the response after freeing the heap
if (SUCCEEDED(RSSFeedServiceSoap12_RetrieveFeeds(proxy, szUrl, cbResults, &temp, heap, NULL, NULL, NULL, error)))
wcscpy_s(retval, temp);
}
}
}
// Deallocate and free resources
if (error != NULL)
WsFreeError(error);
if (proxy != NULL) {
WsCloseServiceProxy(proxy, NULL, error);
WsFreeServiceProxy(proxy);
}
if (heap != NULL)
WsFreeHeap(heap);
return retval;
}
Please note the following:
-
We must specify the version of SOAP to use (otherwise, the client is going to complain about it, because it will use 1.2, as evidence check the RSSFeedServiceSoap12 method name)
-
I return a WCHAR* and it works without issues, even when the right way to do it is returning an HRESULT or an integer, accept a pointer as an argument which at the same time it’s an output parameter. There’s a document about best practices for creating DLLs which can be downloaded from here
Our implementation from the console application is shown below
#include "stdafx.h"
typedef WCHAR* (*myCallback) (WCHAR*, int);
int _tmain(int argc, _TCHAR* argv[]) {
HINSTANCE hInstance;
myCallback ptrToFunc;
WCHAR* results = NULL;
if ((hInstance = LoadLibrary(L"C:\\Users\\angel.hernandez\\Desktop\\WWSAPI\\WWSAPIDemo\\x64\\Debug\\WWSAPIDemo.dll")) != NULL) {
if ((ptrToFunc = (myCallback) GetProcAddress(hInstance, "GetFeeds")) != NULL) {
results = ptrToFunc(L"http://localhost/MSDN.xml", -1);
wprintf(L"\n%ls\n", results);
FreeLibrary(hInstance);
printf("\n\nPress any key to exit...\n");
_getch();
}
}
return 0;
}
After compiling, linking and executing our NativeTester application, we can see how the RSS feeds are displayed on the console.
GetItem is another exported function in our DLL which displays a found element (if any) after executing an XPath query and using the starts-with function (as if it was the Like % operator). GetItem retrieves the RSS feeds by previously calling the GetFeeds function.
// Execute XPATH Query based on feeds retrieved through WWSAPI
EXPORT void GetItem(WCHAR* szUrl, WCHAR* szItemName) {
WCHAR retval[TRIM_SIZE];
WCHAR* results = NULL;
BSTR nodeContent = NULL;
WCHAR xPathQueryBuffer[TRIM_SIZE];
IXMLDOMDocumentPtr docPtr = NULL;
IXMLDOMNodePtr selectedNode = NULL;
if ((results = GetFeeds(szUrl, -1)) != NULL && wcslen(results) > 0) {
CoInitialize(NULL);
docPtr.CreateInstance("Msxml2.DOMDocument.6.0");
wsprintf(xPathQueryBuffer, L"/rss/items/item[starts-with(@title,'%ls')]", szItemName);
if (SUCCEEDED(docPtr->loadXML(_bstr_t(results), NULL))) {
if (SUCCEEDED(docPtr->selectSingleNode(_bstr_t(xPathQueryBuffer), &selectedNode))) {
nodeContent = SysAllocString(retval);
selectedNode->get_xml(&nodeContent);
MessageBox(NULL, nodeContent, L"XPath Query Results", NULL);
SysFreeString(nodeContent);
}
}
}
}
So far we’ve been able to implement and use WWSAPI from native code, however we haven’t tested our Dll’s functionality from .NET, in that case the first thing to do is, “to import” the function we’re interested on by specifying the DllImport attribute.
[DllImport(@"C:\Users\angel.hernandez\Desktop\WWSAPI\WWSAPIDemo\x64\Debug\WWSAPIDemo.dll", CharSet = CharSet.Unicode)]
public static extern IntPtr GetFeeds(IntPtr szUrl, int cbResults);
The GetFeeds function returns a WCHAR* which in turn is interpreted by .NET as an IntPtr. We also need to use the Marshal class to convert it from WCHAR* to an Unicode string and to pass a WCHAR* as parameter to the function
private void btnTestWWSAPI_Click(object sender, EventArgs e) {
IntPtr szUrl = Marshal.StringToBSTR("http://localhost/MSDN.xml");
MessageBox.Show(Marshal.PtrToStringUni(GetFeeds(szUrl, -1)));
Marshal.FreeBSTR(szUrl);
}
and Voila!!! I have my WWSAPI implementation being consumed and used from .NET, well.. from a WinForm application, but what about a WPF application? The answer is… the implementation remains the same except for the XAML code required plus the databinding process is quite simple if an XmlDataProvider is used because the data is extracted and assigned based on an XPath syntax.
<!-- Data Provider-->
<Window.Resources>
<XmlDataProvider x:Key="xmlFeeds" IsAsynchronous="True" XPath="/rss/items/item"/>
</Window.Resources>
<Grid Name="grdFeeds" Margin="0,75,0,12">
<!-- Binding Source-->
<ListView Margin="11.432,6" Name="lstFeeds"
ItemsSource="{Binding Source={StaticResource xmlFeeds} }" MouseDoubleClick="lstFeeds_MouseDoubleClick">
<ListView.View>
<GridView x:Name="grdFeedItems">
<GridViewColumn Header="Title" x:Name="grcTitle" Width="100" DisplayMemberBinding="{Binding XPath=@title}"/>
<GridViewColumn Header="Publishing Date" x:Name="grcPublishingDate" Width="100" DisplayMemberBinding="{Binding XPath=@publishingDate}"/>
<GridViewColumn Header="Url" x:Name="grcUrl" Width="100" DisplayMemberBinding="{Binding XPath=@url}" />
</GridView>
</ListView.View>
</ListView>
</Grid>
The Button Click event handler is shown below
private void btnExecuteOperation_Click(object sender, RoutedEventArgs e) {
XmlDocument feeds = new XmlDocument();
IntPtr szUrl = IntPtr.Zero, szItemName = IntPtr.Zero;
XmlDataProvider xmlFeeds = TryFindResource("xmlFeeds") as XmlDataProvider;
try {
switch (cboOperations.SelectedIndex) {
case 0:
using (RSSFeedService proxy = new RSSFeedService())
feeds.LoadXml(proxy.RetrieveFeeds("http://localhost/MSDN.xml", -1));
xmlFeeds.Document = feeds;
break;
case 1:
szUrl = Marshal.StringToBSTR("http://localhost/MSDN.xml");
feeds.LoadXml(Marshal.PtrToStringUni(GetFeeds(szUrl, -1)));
Marshal.FreeBSTR(szUrl);
xmlFeeds.Document = feeds;
break;
case 2:
szUrl = Marshal.StringToBSTR("http://localhost/MSDN.xml");
szItemName = Marshal.StringToBSTR(!string.IsNullOrEmpty(txtItemTitle.Text) ? txtItemTitle.Text : "Windows");
GetItem(szUrl, szItemName);
Marshal.FreeBSTR(szUrl);
Marshal.FreeBSTR(szItemName);
break;
}
} catch (Exception ex) {
MessageBox.Show(string.Format("Oops! Something wrong just occurred\n{0}", ex.Message),
"Exception caught", MessageBoxButton.OK, MessageBoxImage.Information);
}
}
Attached to this post you can find the code, please feel free to download it, modify it and play with it. I hope you consider this information useful and always remember one thing… The only possible way to learn is by playing with the technology and feel confident about taking risks.
God bless you
Regards,
Angel
Hace un par de noches atrás estaba leyendo sobre las características nuevas del SDK de Windows 7 y puedo decir que son bastantes, de hecho en una de las sesiones que tuve en Tech-Ed este año mencioné algunas de ellas. Sin embargo, una que llamó mi atención fue WWSAPI pues hasta ahora no existía soporte para código nativo, excepto por un par de Toolkits existentes entre ellos gSOAP que puede utilizarse con Windows, Linux y Mac OSX. WWSAPI fue presentado en el PDC del año pasado (2008) y será liberado formalmente con Windows 7 aunque versiones anteriores del sistema operativo, a partir de XP SP2. Yo lo probé inicialmente cuando Windows 7 estaba en RC sin embargo no había posteado al respecto por esperar que Windows 7 fuese liberado o cerca de serlo.
Una vez dicho esto, durante el fin de semana creé una solución para demostrar WWSAPI en conjunto con .NET. La solución tiene la estructura mostrada a continuación
-
NativeTester: Aplicación de Consola (C++) que llama a la biblioteca de enlace dinámico (DLL)
-
RSSFeedService: Servicio Web (C#) que recupera entradas RSS, las parsea y regresa el resultado como XML
-
RSSViewer: Aplicación basada en WPF (C#) que consume el servicio Web e invoca a la biblioteca de enlace dinámico (DLL)
-
Tester: Aplicación basada en WinForm (C#) que consume el servicio Web e invoca a la biblioteca de enlace dinámico (DLL)
-
WWSAPIDemo: Biblioteca de enlace dinámico (C++) que implementa WWSAPI y es invocada desde .NET a través de Interop
Así mismo con el código pueden encontrar MSDN.xml que son las entradas RSS de la página de MSDN Australia, aunque el código apunta a http://localhost/MSDN.xml, ustedes pueden cambiar dicho Url y apuntar al RSS que gusten, en mi caso guardé el archivo de entradas RSS y lo pusé en my IIS local.
Muchos de ustedes se preguntarán, ¿cómo genero el proxy del servicio Web desde C++? Y la respuesta es muy simple, el nuevo SDK de Windows 7 trae consigo un utilitario que lo hace por nosotros, WSUTIL.exe. La ejecución nos da como resultado dos archivos (uno .H y otro .CPP). Las cadenas por defecto son tratadas como WCHAR* que es Unicode y todas las cadenas en .NET son interpretadas como Unicode, más adelante comentaré un poco de esto.
Con el servicio Web ya publicado en IIS, entonces nos queda comenzar a trabajar en la biblioteca de enlace dinámico que será cliente e implementará WWSAPI. A continuación el archivo de cabecera de la biblioteca
#include "stdafx.h"
#define EXPORT extern "C" __declspec(dllexport)
#define MAX_SIZE 128000 //128Kb
#define TRIM_SIZE 512
EXPORT WCHAR* GetFeeds(WCHAR* szUrl, int cbResults);
EXPORT void GetItem(WCHAR* szUrl, WCHAR* szItemName);
así como el método que se conecta al servicio Web y recupera las entradas de RSS
// Get feeds from a given Url through WWSAPI
EXPORT WCHAR* GetFeeds(WCHAR* szUrl, int cbResults) {
WCHAR* temp = NULL;
WS_HEAP* heap = NULL;
WCHAR retval[MAX_SIZE]; // 128Kb (It should be enough to avoid WS_E_QUOTA_EXCEEDED error)
WS_ERROR* error = NULL;
ULONG propertiesCount = 0;
WS_SERVICE_PROXY* proxy = NULL;
WS_CHANNEL_PROPERTY channelProps[2];
WS_STRING serviceUrl = WS_STRING_VALUE(L"http://localhost/DemoSvc/RSSFeedService.asmx");
WS_ENDPOINT_ADDRESS endpoint = { serviceUrl};
WS_ENVELOPE_VERSION soapVersion = WS_ENVELOPE_VERSION_SOAP_1_1; // Our Webservice is WsiProfiles.BasicProfile1_1 compliant
WS_ADDRESSING_VERSION addressingVersion = WS_ADDRESSING_VERSION_TRANSPORT;
// Set channel's properties
channelProps[propertiesCount].id = WS_CHANNEL_PROPERTY_ENVELOPE_VERSION;
channelProps[propertiesCount].value = &soapVersion;
channelProps[propertiesCount].valueSize = sizeof(soapVersion);
propertiesCount++;
// Set addressing's properties
channelProps[propertiesCount].id = WS_CHANNEL_PROPERTY_ADDRESSING_VERSION;
channelProps[propertiesCount].value = &addressingVersion;
channelProps[propertiesCount].valueSize = sizeof(addressingVersion);
propertiesCount++;
// Can we create an WsError and WsHeap objects?
if (SUCCEEDED(WsCreateError(NULL, NULL, &error)) && SUCCEEDED(WsCreateHeap(MAX_SIZE, TRIM_SIZE, NULL, NULL, &heap, error))) {
// Can we create a proxy based on the service?
if (SUCCEEDED(WsCreateServiceProxy(WS_CHANNEL_TYPE_REQUEST, WS_HTTP_CHANNEL_BINDING, NULL, NULL, NULL,
channelProps, propertiesCount, &proxy, error))) {
// Can we open the proxy object?
if (SUCCEEDED(WsOpenServiceProxy(proxy, &endpoint, NULL, error))) {
// If we're able to invoke the service then copy the results to another variable because if
// we don't we lose the response after freeing the heap
if (SUCCEEDED(RSSFeedServiceSoap12_RetrieveFeeds(proxy, szUrl, cbResults, &temp, heap, NULL, NULL, NULL, error)))
wcscpy_s(retval, temp);
}
}
}
// Deallocate and free resources
if (error != NULL)
WsFreeError(error);
if (proxy != NULL) {
WsCloseServiceProxy(proxy, NULL, error);
WsFreeServiceProxy(proxy);
}
if (heap != NULL)
WsFreeHeap(heap);
return retval;
}
Por favor, nótese lo siguiente:
-
Debemos especificar la versión de SOAP (de lo contrario, el cliente se va a quejar al respecto, porque va a utilizar 1.2. Prueba de esto, es el nombre del método RSSFeedServiceSoap12)
-
Regreso un WCHAR* y funciona sin problemas, aunque la manera correcta debería ser es regresar un HRESULT o un entero, tomar un puntero como parámetro que al mismo tiempo sirve de valor de retorno. Un documento sobre las mejores prácticas de desarrollo de DLLs puede encontrarse aquí
Nuestra implementación desde la aplicación de consola es mostrada a continuación
#include "stdafx.h"
typedef WCHAR* (*myCallback) (WCHAR*, int);
int _tmain(int argc, _TCHAR* argv[]) {
HINSTANCE hInstance;
myCallback ptrToFunc;
WCHAR* results = NULL;
if ((hInstance = LoadLibrary(L"C:\\Users\\angel.hernandez\\Desktop\\WWSAPI\\WWSAPIDemo\\x64\\Debug\\WWSAPIDemo.dll")) != NULL) {
if ((ptrToFunc = (myCallback) GetProcAddress(hInstance, "GetFeeds")) != NULL) {
results = ptrToFunc(L"http://localhost/MSDN.xml", -1);
wprintf(L"\n%ls\n", results);
FreeLibrary(hInstance);
printf("\n\nPress any key to exit...\n");
_getch();
}
}
return 0;
}
Al compilar y ejecutar nuestra aplicación de prueba NativeTester, podemos ver como se muestran las entradas recuperadas en la consola.
La otra función contenida en la biblioteca de enlace dinámico es GetItem, que muestra el elemento encontrado tras la ejecución de una consulta de XPath y haciendo uso de la función starts-with (como si fuera el operador Like %). La función GetItem recupera las entradas RSS al llamar previamente a la función GetFeeds.
// Execute XPATH Query based on feeds retrieved through WWSAPI
EXPORT void GetItem(WCHAR* szUrl, WCHAR* szItemName) {
WCHAR retval[TRIM_SIZE];
WCHAR* results = NULL;
BSTR nodeContent = NULL;
WCHAR xPathQueryBuffer[TRIM_SIZE];
IXMLDOMDocumentPtr docPtr = NULL;
IXMLDOMNodePtr selectedNode = NULL;
if ((results = GetFeeds(szUrl, -1)) != NULL && wcslen(results) > 0) {
CoInitialize(NULL);
docPtr.CreateInstance("Msxml2.DOMDocument.6.0");
wsprintf(xPathQueryBuffer, L"/rss/items/item[starts-with(@title,'%ls')]", szItemName);
if (SUCCEEDED(docPtr->loadXML(_bstr_t(results), NULL))) {
if (SUCCEEDED(docPtr->selectSingleNode(_bstr_t(xPathQueryBuffer), &selectedNode))) {
nodeContent = SysAllocString(retval);
selectedNode->get_xml(&nodeContent);
MessageBox(NULL, nodeContent, L"XPath Query Results", NULL);
SysFreeString(nodeContent);
}
}
}
}
Hasta ahora hemos logrado implementar y utilizar WWSAPI desde código nativo, sin embargo aún no hemos probado la funcionalidad de nuestra biblioteca de enlace dinámico desde .NET, en ese caso lo primero que debemos hacer es importar la función que nos interesa a través de DllImport
[DllImport(@"C:\Users\angel.hernandez\Desktop\WWSAPI\WWSAPIDemo\x64\Debug\WWSAPIDemo.dll", CharSet = CharSet.Unicode)]
public static extern IntPtr GetFeeds(IntPtr szUrl, int cbResults);
La función GetFeeds retorna un WCHAR* que es traducido a NET como un IntPtr, así mismo hacemos uso de la clase Marshal para traducir el WCHAR* a una cadena Unicode y para pasar un WCHAR* a la función, como es mostrado a continuación
private void btnTestWWSAPI_Click(object sender, EventArgs e) {
IntPtr szUrl = Marshal.StringToBSTR("http://localhost/MSDN.xml");
MessageBox.Show(Marshal.PtrToStringUni(GetFeeds(szUrl, -1)));
Marshal.FreeBSTR(szUrl);
}
Y Voila!!! Tengo mi implementación de WWSAPI siendo utilizada desde .NET, bueno al menos desde una aplicación Windows Form, pero ¿cómo será con una aplicación WPF? Pues… la respuesta es será igual, a diferencia de que debes escribir el XAML y que hacer binding a los objetos es más sencillo si se utiliza un XmlDataProvider pues los datos se extraen y asignan utilizando una sintaxis basada en XPath.
<!-- Data Provider-->
<Window.Resources>
<XmlDataProvider x:Key="xmlFeeds" IsAsynchronous="True" XPath="/rss/items/item"/>
</Window.Resources>
<Grid Name="grdFeeds" Margin="0,75,0,12">
<!-- Binding Source-->
<ListView Margin="11.432,6" Name="lstFeeds"
ItemsSource="{Binding Source={StaticResource xmlFeeds} }" MouseDoubleClick="lstFeeds_MouseDoubleClick">
<ListView.View>
<GridView x:Name="grdFeedItems">
<GridViewColumn Header="Title" x:Name="grcTitle" Width="100" DisplayMemberBinding="{Binding XPath=@title}"/>
<GridViewColumn Header="Publishing Date" x:Name="grcPublishingDate" Width="100" DisplayMemberBinding="{Binding XPath=@publishingDate}"/>
<GridViewColumn Header="Url" x:Name="grcUrl" Width="100" DisplayMemberBinding="{Binding XPath=@url}" />
</GridView>
</ListView.View>
</ListView>
</Grid>
El manejador del evento Click del botón para llamar a nuestra biblioteca de enlace dinámico es mostrado a continuación
private void btnExecuteOperation_Click(object sender, RoutedEventArgs e) {
XmlDocument feeds = new XmlDocument();
IntPtr szUrl = IntPtr.Zero, szItemName = IntPtr.Zero;
XmlDataProvider xmlFeeds = TryFindResource("xmlFeeds") as XmlDataProvider;
try {
switch (cboOperations.SelectedIndex) {
case 0:
using (RSSFeedService proxy = new RSSFeedService())
feeds.LoadXml(proxy.RetrieveFeeds("http://localhost/MSDN.xml", -1));
xmlFeeds.Document = feeds;
break;
case 1:
szUrl = Marshal.StringToBSTR("http://localhost/MSDN.xml");
feeds.LoadXml(Marshal.PtrToStringUni(GetFeeds(szUrl, -1)));
Marshal.FreeBSTR(szUrl);
xmlFeeds.Document = feeds;
break;
case 2:
szUrl = Marshal.StringToBSTR("http://localhost/MSDN.xml");
szItemName = Marshal.StringToBSTR(!string.IsNullOrEmpty(txtItemTitle.Text) ? txtItemTitle.Text : "Windows");
GetItem(szUrl, szItemName);
Marshal.FreeBSTR(szUrl);
Marshal.FreeBSTR(szItemName);
break;
}
} catch (Exception ex) {
MessageBox.Show(string.Format("Oops! Something wrong just occurred\n{0}", ex.Message),
"Exception caught", MessageBoxButton.OK, MessageBoxImage.Information);
}
}
El post tiene adjunto el código mostrado, pueden descargarlo, modificarlo y jugar con él. Espero que sea de utilidad y recuerden, la única manera de aprender es jugar con la tecnología y no tener miedo para asumir nuevos retos.
Que Dios los bendiga.
Saludos,
Angel
Hi Community,
Please find attached a couple of pictures from Microsoft TechEd, they were taken and sent to me by our ANZ MVP Lead (Roseanne Stamell). In Picture 1, my fellow MVP and mate, Paul Glavich and I. Picture 2, it’s just me getting ready to start a Mini Theatre session. .
Regards,
Angel
Hola Comunidad,
Por favor vean adjuntas un par de fotos desde Microsoft TechEd, las cuales fueron tomadas y enviadas por nuestra líder MVP (Roseanne Stamell). En la foto 1, un colega MVP y amigo, Paul Glavich y yo. Foto 2, soy yo preparándome para comenzar una sesión en un mini teatro.
Saludos,
Angel
A couple of weeks ago I was attending and presenting at Microsoft TechEd on Gold Coast, when I got a phone call from our Sales Manager. The mission: Develop a new webpart for an existing B2B portal which integrates SAP, MQ, CRM and webMethods through SharePoint. The challenge: Write some Javascript code to do some processing on the server-side based on the user’s input.
I’ve been doing some SharePoint development for a while and if you know how to do things on ASP.NET then is not that hard to leverage those skills and use them on SharePoint. The requirements were:
-
Allow user input and perform some validation on the browser-side
-
Add AJAX capabilities to deliver a responsive UI to the users
-
Grab information stored on the client-side (browser) from the server-side and do some processing against CRM
So let’s start from the beginning explaining how I got this done
Properties required
private DataTable SearchResults {
get {
return (ViewState["block_orders"] != null ?
(DataTable)ViewState["block_orders"] : null);
}
set {
ViewState["block_orders"] = value;
}
}
private string GridViewSortCondition {
get {
if (ViewState["SortCondition"] == null)
ViewState["SortCondition"] = "sapSearch ASC";
return ViewState["SortCondition"].ToString();
}
set {
ViewState["SortCondition"] = value;
}
}
private SortDirection GridViewSortDirection {
get {
if (ViewState["SortDirection"] == null)
ViewState["SortDirection"] = SortDirection.Ascending;
return (SortDirection)ViewState["SortDirection"];
}
set {
ViewState["SortDirection"] = value;
}
}
private string SortExpression {
get {
return (ViewState["sort_exp"] != null ?
ViewState["sort_exp"].ToString() : string.Empty);
}
set {
ViewState["sort_exp"] = value;
}
}
private string SelectedCheckBoxes {
get {
return (ViewState["selected_check_boxes"] != null ?
ViewState["selected_check_boxes"].ToString() : string.Empty);
}
set {
ViewState["selected_check_boxes"] = value;
}
}
I need to store the search results so the user can perform paging and sorting operations on the result set
Now, let’s proceed to the code required to add AJAX functionality, I usually break down the UI code into pieces, I mean, a method to generate the top, middle and bottom parts of a webpart. Since I’m implementing some AJAX I enclose everything inside a <DIV> so the code required looks like this
private void AddAJAXControls() {
Controls.Add(new Literal() { Text = "<div>", ID = "mainDiv" });
Controls.Add(new ScriptManager() { ID = "scriptManager_OrderBlock" });
Controls.Add(new Literal() { Text = InjectScriptToDisableDoublePostBack() });
Controls.Add(new UpdatePanel() { ID = "updatePanel_OrderBlock" });
}
Please note the call to “InjectScriptToDisableDoublePostback” method, it’s responsible for disabling the control which has triggered the postback (e.g.: a submit button) so the user won’t be able to “re-submit” the form more than once, the code is shown below
private string InjectScriptToDisableDoublePostBack() {
StringBuilder retval = new StringBuilder();
retval.AppendLine("<script type='text/javascript'>");
retval.AppendLine("var pbControl = null;");
retval.AppendLine("var prm = Sys.WebForms.PageRequestManager.getInstance();");
retval.AppendLine("prm.add_beginRequest(BeginRequestHandler);");
retval.AppendLine("prm.add_endRequest(EndRequestHandler);");
retval.AppendLine("function BeginRequestHandler(sender, args) {");
retval.AppendLine("//the control causing the postback");
retval.AppendLine("pbControl = args.get_postBackElement();");
retval.AppendLine("if (pbControl.id.indexOf('btnSearch') > -1 || ");
retval.AppendLine("pbControl.id.indexOf('btnUpdate') > -1 )");
retval.AppendLine("pbControl.disabled = true;");
retval.AppendLine("}");
retval.AppendLine("function EndRequestHandler(sender, args) {");
retval.AppendLine("pbControl.disabled = false;");
retval.AppendLine("pbControl = null;");
retval.AppendLine("}");
retval.AppendLine("</script>");
return retval.ToString();
}
The method responsible for putting all of the pieces together and rendering the webpart is
protected virtual void CreateUI() {
UpdatePanel panel = null;
// Add AJAX controls
AddAJAXControls();
if ((panel = FindControl("updatePanel_OrderBlock") as UpdatePanel) != null) {
// Top Table
panel.ContentTemplateContainer.Controls.Add(GetTopTable());
// Bottom Table
panel.ContentTemplateContainer.Controls.Add(GetBottomTable());
}
// Close Div containing AJAX controls
Controls.Add(new Literal() { Text = "</div>" });
}
This approach gives me a clean and tidy HTML when rendered plus it’s easy to follow, debug and maintain. Now, I have all of the objects required inside a <DIV> and it’s AJAX enabled… The only missing thing is our UpdateProgress template (code is attached) which uses an animated GIF bundled in SharePoint.
Table retval = new Table() {
ID = "tblTopTable", CellPadding = 1,
CellSpacing = 0, Width = Unit.Percentage(100)
};
retval.Rows.AddRange(new TableRow[] {new TableRow(), new TableRow(),
new TableRow(), new TableRow(),
new TableRow(), new TableRow(),
new TableRow(), new TableRow(),
new TableRow()});
retval.Rows[ 8 ].Cells.AddRange(new TableCell[] { new TableCell() {
Width=Unit.Percentage(20)},
new TableCell() {
Width=Unit.Percentage(60)}});
// Update Progress
retval.Rows[ 8 ].Cells[0].ColumnSpan = 2;
retval.Rows[ 8 ].Cells[0].HorizontalAlign = HorizontalAlign.Left;
retval.Rows[ 8 ].Cells[0].Controls.Add(new UpdateProgress() {
ID = "workProgress",
ProgressTemplate = new UpdateProgressTemplate()
});
Ok… so far we’ve got AJAX working but what’s special about that? Well.. not much really but now we need to synchronize user’s selection on the client-side and reflect these changes on the server-side and to accomplish that we need some scripts (code is attached) and a couple of hidden fields which are referenced from our JavaScript as well as from our webpart code and for the sake of keeping things tidy we add them to the controls collection right after creating everything else (as depicted below)
protected override void CreateChildControls() {
CreateUI();
Controls.Add(new HiddenField() { ID = "selected_checkboxes",
EnableViewState = true });
Controls.Add(new HiddenField() { ID = "checkboxes_cleared",
EnableViewState = true });
ChildControlsCreated = true;
}
We also need to add some Javascript to support client-side operations so we do this on the OnPreRender method (AddJavaScriptToWebpart method is attached to the post)
protected override void OnPreRender(EventArgs e) {
if (!Page.ClientScript.IsClientScriptBlockRegistered(JSCRIPT_NAME))
Page.ClientScript.RegisterClientScriptBlock(typeof(string),
JSCRIPT_NAME, AddJavaScriptToWebpart());
}
Up to now we’ve met some of the user’s requirements but what about the GridView? Let’s talk about it then. The GridView is created by the following method
private GridView GetSearchResultsGrid() {
GridView retval = new GridView() {
ID = "grdSearchResults", AutoGenerateColumns = false,
AllowPaging = true, AllowSorting = true, Width = Unit.Percentage(100),
PageSize = 100,
EmptyDataText="<font color='red'><b>No results found</b></font>"
};
// Add Columns to the GridView
retval.Columns.Add(new BoundField() { DataField = "sapSearch",
HeaderText = "SAP Search",
SortExpression = "sapSearch" });
retval.Columns.Add(new BoundField() { DataField = "accountNo",
HeaderText = "Account No.",
SortExpression = "accountNo" });
retval.Columns.Add(new BoundField() { DataField = "accountName",
HeaderText = "Account Name",
SortExpression = "accountName" });
retval.Columns.Add(new BoundField() { DataField = "state",
HeaderText = "State",
SortExpression = "state" });
retval.Columns.Add(new TemplateField() {
ItemTemplate = new CheckBoxTemplate(ListItemType.Item),
HeaderTemplate = new CheckBoxTemplate(ListItemType.Header)
});
// Let's align to the centre the Order Block and State Columns
retval.Columns[3].ItemStyle.HorizontalAlign = HorizontalAlign.Center;
retval.Columns[4].ItemStyle.HorizontalAlign = HorizontalAlign.Center;
// Subscribe to events
retval.Sorting += grdSearchResults_Sorting;
retval.RowCreated += grdSearchResults_RowCreated;
retval.PageIndexChanging += grdResults_PageIndexChanging;
retval.RowDataBound += grdSearchResults_RowDataBound;
return retval;
}
The ticking/unticking of the checkboxes is handled by a Javascript call from CheckBoxTemplate (attached to the post) as shown below
private CheckBox GetTemplateContents() {
CheckBox retval = null;
switch (_type) {
case ListItemType.Header:
retval = new CheckBox() { ID = "chkHeader",
Text = "Account Block",
EnableViewState = true };
retval.Attributes["onclick"] = "BLOCKED SCRIPTcheckUncheckHeader(this);";
break;
case ListItemType.Item:
retval = new CheckBox() {ID = "chkItem_", EnableViewState = true };
retval.Attributes["onclick"] = "BLOCKED SCRIPTcheckUncheckItem(this);";
break;
}
return retval;
As previously mentioned, our Gridview had to support paging and sorting plus display an indicator for the current sorting criteria so I had to handle the RowCreated event as depicted below
private void grdSearchResults_RowCreated(object sender,
GridViewRowEventArgs e) {
int rowIndex = 0;
DataView sorted = null;
string strSortedHeader = string.Empty;
// Header
if (e.Row.RowType == DataControlRowType.Header) {
foreach (TableCell tc in e.Row.Cells) {
if (tc.Controls.Count > 0 && tc.Controls[0].GetType().ToString() ==
"System.Web.UI.WebControls.DataControlLinkButton") {
strSortedHeader = ((LinkButton)tc.Controls[0]).Text;
// Sort indicator (Webdings font does the job for us)
if (((LinkButton)tc.Controls[0]).CommandArgument ==
SortExpression) {
if (GridViewSortDirection == SortDirection.Descending)
((LinkButton)tc.Controls[0]).Text =
strSortedHeader.Replace(strSortedHeader,
strSortedHeader + "<font face='Webdings'>5</font>");
else
((LinkButton)tc.Controls[0]).Text =
strSortedHeader.Replace(strSortedHeader,
strSortedHeader + "<font face='Webdings'>6</font>");
}
}
}
} else if (e.Row.RowType == DataControlRowType.DataRow) { // DataRow
if (e.Row.Cells[e.Row.Cells.Count - 1].Controls.Count > 0 &&
SearchResults != null && SearchResults.Rows.Count > 0) {
sorted = new DataView(SearchResults);
sorted.Sort = GridViewSortCondition;
rowIndex = e.Row.DataItemIndex;
((CheckBox)e.Row.Cells[e.Row.Cells.Count - 1].Controls[0]).ID +=
sorted[e.Row.DataItemIndex]["accountNo"].ToString();
((CheckBox)e.Row.Cells[e.Row.Cells.Count - 1].Controls[0]).
Checked = (bool) sorted[e.Row.DataItemIndex]["isSelected"];
// Is it checked? (Non user interaction)
if (((CheckBox)e.Row.Cells[e.Row.Cells.Count - 1].
Controls[0]).Checked)
SelectedCheckBoxes += string.Format("{0};",
((CheckBox)e.Row.Cells[e.Row.Cells.Count - 1].
Controls[0]).ClientID);
}
}
}
Please note that we make a difference when creating/rendering the rows based on their type, for instance, the header is going to display an arrow to indicate sorting direction, otherwise we grab the values retrieved from the database (CRM) and change the checkboxes’ ID on the fly, by adding the account number to it (this is required to keep track of the user selection).
At this moment you’ll be wondering about, where is this guy grabbing the values set already from the client side and calling CRM? I’m pretty sure that many of you know the answer… The button click event handler.
private void btnUpdate_Click(object sender, EventArgs e) {
int rowIndex = 0;
int selectedCell = 0;
DataView sorted = null;
string[] splitData = null;
string selectedControl = string.Empty;
HiddenField clientSideSelection = null;
StringBuilder userSelection = new StringBuilder();
GridView grdSearchResults = FindControl("grdSearchResults") as GridView;
HiddenField checkboxes_cleared = FindControl("checkboxes_cleared") as
HiddenField;
// Is there any data to continue?
if ((SearchResults != null && SearchResults.Rows.Count > 0) &&
(grdSearchResults != null && grdSearchResults.Rows.Count > 0)) {
// Let's create a view to deal with the data as it is displayed
sorted = new DataView(SearchResults);
sorted.Sort = GridViewSortCondition;
// Let's sync both server and client checkboxes selection
ManageDeselectedItemsOnClientSide();
// Should we combine client-side selection with data from the DB?
if ((clientSideSelection = FindControl("selected_checkboxes") as HiddenField) != null
&& !string.IsNullOrEmpty(clientSideSelection.Value)) {
splitData = clientSideSelection.Value.Split(';');
var includeQuery = from checkBoxName in splitData.ToList()
.Where(controlName => !string.IsNullOrEmpty(controlName) &&
SelectedCheckBoxes.IndexOf(controlName) == -1 &&
controlName.IndexOf("_REMOVED") == -1)
select checkBoxName;
// Is there any item we must include?
if (includeQuery.ToList().Count > 0)
includeQuery.ToList().ForEach(controlToInclude =>
userSelection.Append(string.Format("{0};",
controlToInclude.Substring(controlToInclude.IndexOf("chkItem")))));
// Let's combine it with the information coming from the DB
userSelection.Append(SelectedCheckBoxes);
} else if (checkboxes_cleared != null && // Has the user interacted with the UI?
string.IsNullOrEmpty(checkboxes_cleared.Value))
userSelection.Append(SelectedCheckBoxes);
// Accounts to update
splitData = userSelection.ToString().Split(';');
// We loop through the control names collection
// (by default, we unblock those unselected checkboxes)
foreach (string checkBoxId in splitData) {
rowIndex = 0;
foreach (GridViewRow selectedRow in grdSearchResults.Rows) {
selectedCell = selectedRow.Cells.Count - 1;
// Is it the right one?
if (selectedRow.Cells[selectedCell].Controls.Count > 0 &&
selectedRow.Cells[selectedCell].Controls[0] is CheckBox) {
selectedControl = ((CheckBox)selectedRow.Cells[selectedCell].Controls[0]).ID;
if (selectedControl.Equals(checkBoxId)) {
// Account to block
sorted[rowIndex]["isSelected"] = true;
break;
} else if (!splitData.Contains(selectedControl)) {
// Account to unblock (default behaviour)
sorted[rowIndex]["isSelected"] = false;
}
}
rowIndex++;
}
}
// Block/Unblock accounts based on the user's selection
var accountsToProcess = from currentSelection in sorted.Table
.AsEnumerable()
select new {
AccountId = currentSelection["accountId"].ToString(),
IsSelected = (bool)currentSelection["isSelected"]
};
accountsToProcess.ToList().ForEach(account =>
OrderBlockManagement.updateAccount(account.AccountId,
account.IsSelected));
// Let's clear DB selection information
SelectedCheckBoxes = string.Empty;
// Let's reflect the changes on the UI
GridBindingHelper();
}
}
The final result is shown below
Our webpart integrating different systems as mentioned at the beginning of this post plus providing a responsive UI to the user.
Regards,
Angel
Hace un par de semanas atrás estaba atendiendo y presentando en Microsoft TechEd en Gold Coast, cuando de repente recibí una llamada del Gerente de Mercadeo y Ventas. La misión: Desarrollar un nuevo webpart para un portal B2B existente que integra SAP, MQ, CRM y webMethods a través SharePoint. El reto: Escribir código en Javascript para luego realizar operaciones del lado del servidor en base a la selección realizada por el usuario.
He estado desarrollando para SharePoint por un rato ya, pero si sabes como hacer las cosas con ASP.NET entonces no es díficil utilizar ese conocimiento y aplicarlos con SharePoint. Los requerimientos eran :
-
Permitir la entrada del usuario y validar del lado del navegador
-
Agregar AJAX para ofrecer una interfaz rápida a los usuarios
-
Obtener información almacenada del lado del cliente (navegador) desde el lado del servidor para realizar operaciones contra CRM
Así que comencemos por el principio explicando como lo hice
Propiedades requeridas
private DataTable SearchResults {
get {
return (ViewState["block_orders"] != null ?
(DataTable)ViewState["block_orders"] : null);
}
set {
ViewState["block_orders"] = value;
}
}
private string GridViewSortCondition {
get {
if (ViewState["SortCondition"] == null)
ViewState["SortCondition"] = "sapSearch ASC";
return ViewState["SortCondition"].ToString();
}
set {
ViewState["SortCondition"] = value;
}
}
private SortDirection GridViewSortDirection {
get {
if (ViewState["SortDirection"] == null)
ViewState["SortDirection"] = SortDirection.Ascending;
return (SortDirection)ViewState["SortDirection"];
}
set {
ViewState["SortDirection"] = value;
}
}
private string SortExpression {
get {
return (ViewState["sort_exp"] != null ?
ViewState["sort_exp"].ToString() : string.Empty);
}
set {
ViewState["sort_exp"] = value;
}
}
private string SelectedCheckBoxes {
get {
return (ViewState["selected_check_boxes"] != null ?
ViewState["selected_check_boxes"].ToString() : string.Empty);
}
set {
ViewState["selected_check_boxes"] = value;
}
}
Necesito guardar los resultados de la búsqueda para que el usuario pueda paginar y ordenar estos
Ahora, prosigamos con el código requerido para agregar AJAX, en mi caso siempre separo/distribuyo el código que genera la interfaz de usuario en partes, es decir, tengo un método para generar la parte superior, media e inferior del webpart y como estoy implementando AJAX me gusta encerrar todo dentro de un <DIV> por lo que el código requerido es como éste
private void AddAJAXControls() {
Controls.Add(new Literal() { Text = "<div>", ID = "mainDiv" });
Controls.Add(new ScriptManager() { ID = "scriptManager_OrderBlock" });
Controls.Add(new Literal() { Text = InjectScriptToDisableDoublePostBack() });
Controls.Add(new UpdatePanel() { ID = "updatePanel_OrderBlock" });
}
Por favor nótese la llamada al método “InjectScriptToDisableDoublePostback”, el cuál es responsable de deshabilitar el control que generó el postback (por ejemplo, un botón enviar) así evito que el usuario lo presione más de una vez, el código es mostrado abajo
private string InjectScriptToDisableDoublePostBack() {
StringBuilder retval = new StringBuilder();
retval.AppendLine("<script type='text/javascript'>");
retval.AppendLine("var pbControl = null;");
retval.AppendLine("var prm = Sys.WebForms.PageRequestManager.getInstance();");
retval.AppendLine("prm.add_beginRequest(BeginRequestHandler);");
retval.AppendLine("prm.add_endRequest(EndRequestHandler);");
retval.AppendLine("function BeginRequestHandler(sender, args) {");
retval.AppendLine("//the control causing the postback");
retval.AppendLine("pbControl = args.get_postBackElement();");
retval.AppendLine("if (pbControl.id.indexOf('btnSearch') > -1 || ");
retval.AppendLine("pbControl.id.indexOf('btnUpdate') > -1 )");
retval.AppendLine("pbControl.disabled = true;");
retval.AppendLine("}");
retval.AppendLine("function EndRequestHandler(sender, args) {");
retval.AppendLine("pbControl.disabled = false;");
retval.AppendLine("pbControl = null;");
retval.AppendLine("}");
retval.AppendLine("</script>");
return retval.ToString();
}
El método responsable de poner todas las piezas juntas y dibujar el webpart es
protected virtual void CreateUI() {
UpdatePanel panel = null;
// Add AJAX controls
AddAJAXControls();
if ((panel = FindControl("updatePanel_OrderBlock") as UpdatePanel) != null) {
// Top Table
panel.ContentTemplateContainer.Controls.Add(GetTopTable());
// Bottom Table
panel.ContentTemplateContainer.Controls.Add(GetBottomTable());
}
// Close Div containing AJAX controls
Controls.Add(new Literal() { Text = "</div>" });
}
Este método me genera un HTML limpio y ordenado cuando el webpart es dibujado además es fácil de seguir, depurar y mantener. Ahora, todos los objetos requeridos están dentro del <DIV> y ya agregué AJAX… La única cosa que me falta es nuestra plantilla de progreso (código adjunto) que utiliza un GIF animado que viene con SharePoint.
Table retval = new Table() {
ID = "tblTopTable", CellPadding = 1,
CellSpacing = 0, Width = Unit.Percentage(100)
};
retval.Rows.AddRange(new TableRow[] {new TableRow(), new TableRow(),
new TableRow(), new TableRow(),
new TableRow(), new TableRow(),
new TableRow(), new TableRow(),
new TableRow()});
retval.Rows[ 8 ].Cells.AddRange(new TableCell[] { new TableCell() {
Width=Unit.Percentage(20)},
new TableCell() {
Width=Unit.Percentage(60)}});
// Update Progress
retval.Rows[ 8 ].Cells[0].ColumnSpan = 2;
retval.Rows[ 8 ].Cells[0].HorizontalAlign = HorizontalAlign.Left;
retval.Rows[ 8 ].Cells[0].Controls.Add(new UpdateProgress() {
ID = "workProgress",
ProgressTemplate = new UpdateProgressTemplate()
});
Bueno… hasta ahora tenemos el AJAX trabajando con nuestro webpart pero, ¿qué es lo especial de eso? Bien... no mucho en realidad pero ahora necesitamos sincronizar la selección realizada por el usuario del lado del cliente y reflejar estos cambios del lado del servidor, así que para lograr eso necesitamos escribir unos scripts (código adjunto) y un par de campos ocultos que son referenciados desde nuestro código en JavaScript y del webpart también, por lo que en aras de mantener las cosas ordenadas y simples, agregamos estos a la colección de controles del webpart una vez que ya hemos creado todo lo demás (como se muestra abajo)
protected override void CreateChildControls() {
CreateUI();
Controls.Add(new HiddenField() { ID = "selected_checkboxes",
EnableViewState = true });
Controls.Add(new HiddenField() { ID = "checkboxes_cleared",
EnableViewState = true });
ChildControlsCreated = true;
}
También necesitamos agregar JavaScript para soportar operaciones del lado del cliente por lo que sobrescribimos el método OnPreRender (El método AddJavaScriptToWebpart está adjunto a este post)
protected override void OnPreRender(EventArgs e) {
if (!Page.ClientScript.IsClientScriptBlockRegistered(JSCRIPT_NAME))
Page.ClientScript.RegisterClientScriptBlock(typeof(string),
JSCRIPT_NAME, AddJavaScriptToWebpart());
}
Hasta ahora hemos satisfecho algunos de los requerimientos solicitados por el usuario, pero ¿qué pasó con el GridView? Hablemos del GridView entonces. El GridView es creado por el siguiente método
private GridView GetSearchResultsGrid() {
GridView retval = new GridView() {
ID = "grdSearchResults", AutoGenerateColumns = false,
AllowPaging = true, AllowSorting = true, Width = Unit.Percentage(100),
PageSize = 100,
EmptyDataText="<font color='red'><b>No results found</b></font>"
};
// Add Columns to the GridView
retval.Columns.Add(new BoundField() { DataField = "sapSearch",
HeaderText = "SAP Search",
SortExpression = "sapSearch" });
retval.Columns.Add(new BoundField() { DataField = "accountNo",
HeaderText = "Account No.",
SortExpression = "accountNo" });
retval.Columns.Add(new BoundField() { DataField = "accountName",
HeaderText = "Account Name",
SortExpression = "accountName" });
retval.Columns.Add(new BoundField() { DataField = "state",
HeaderText = "State",
SortExpression = "state" });
retval.Columns.Add(new TemplateField() {
ItemTemplate = new CheckBoxTemplate(ListItemType.Item),
HeaderTemplate = new CheckBoxTemplate(ListItemType.Header)
});
// Let's align to the centre the Order Block and State Columns
retval.Columns[3].ItemStyle.HorizontalAlign = HorizontalAlign.Center;
retval.Columns[4].ItemStyle.HorizontalAlign = HorizontalAlign.Center;
// Subscribe to events
retval.Sorting += grdSearchResults_Sorting;
retval.RowCreated += grdSearchResults_RowCreated;
retval.PageIndexChanging += grdResults_PageIndexChanging;
retval.RowDataBound += grdSearchResults_RowDataBound;
return retval;
}
La selección/de-selección de las casillas de verificación es manejado por una llamada a una función en JavaScript, realizada desde CheckBoxTemplate (adjunto al post) como es mostrado abajo
private CheckBox GetTemplateContents() {
CheckBox retval = null;
switch (_type) {
case ListItemType.Header:
retval = new CheckBox() { ID = "chkHeader",
Text = "Account Block",
EnableViewState = true };
retval.Attributes["onclick"] = "BLOCKED SCRIPTcheckUncheckHeader(this);";
break;
case ListItemType.Item:
retval = new CheckBox() {ID = "chkItem_", EnableViewState = true };
retval.Attributes["onclick"] = "BLOCKED SCRIPTcheckUncheckItem(this);";
break;
}
return retval;
Como mencionamos previamente, nuestro GridVied tenía que soportar paginación y ordenamiento además de mostrar un indicador para el criterio de ordenamiento actual así que tuve que manejar el evento RowCreated event como se muestra a continuación
private void grdSearchResults_RowCreated(object sender,
GridViewRowEventArgs e) {
int rowIndex = 0;
DataView sorted = null;
string strSortedHeader = string.Empty;
// Header
if (e.Row.RowType == DataControlRowType.Header) {
foreach (TableCell tc in e.Row.Cells) {
if (tc.Controls.Count > 0 && tc.Controls[0].GetType().ToString() ==
"System.Web.UI.WebControls.DataControlLinkButton") {
strSortedHeader = ((LinkButton)tc.Controls[0]).Text;
// Sort indicator (Webdings font does the job for us)
if (((LinkButton)tc.Controls[0]).CommandArgument ==
SortExpression) {
if (GridViewSortDirection == SortDirection.Descending)
((LinkButton)tc.Controls[0]).Text =
strSortedHeader.Replace(strSortedHeader,
strSortedHeader + "<font face='Webdings'>5</font>");
else
((LinkButton)tc.Controls[0]).Text =
strSortedHeader.Replace(strSortedHeader,
strSortedHeader + "<font face='Webdings'>6</font>");
}
}
}
} else if (e.Row.RowType == DataControlRowType.DataRow) { // DataRow
if (e.Row.Cells[e.Row.Cells.Count - 1].Controls.Count > 0 &&
SearchResults != null && SearchResults.Rows.Count > 0) {
sorted = new DataView(SearchResults);
sorted.Sort = GridViewSortCondition;
rowIndex = e.Row.DataItemIndex;
((CheckBox)e.Row.Cells[e.Row.Cells.Count - 1].Controls[0]).ID +=
sorted[e.Row.DataItemIndex]["accountNo"].ToString();
((CheckBox)e.Row.Cells[e.Row.Cells.Count - 1].Controls[0]).
Checked = (bool) sorted[e.Row.DataItemIndex]["isSelected"];
// Is it checked? (Non user interaction)
if (((CheckBox)e.Row.Cells[e.Row.Cells.Count - 1].
Controls[0]).Checked)
SelectedCheckBoxes += string.Format("{0};",
((CheckBox)e.Row.Cells[e.Row.Cells.Count - 1].
Controls[0]).ClientID);
}
}
}
Por favor, nótese la diferencia cuando se está creando/dibujando las filas basadas en su tipo, es decir, el encabezado va a mostrar una flecha que indica la dirección de ordenamiento, de lo contrario, tomamos los valores recuperados de la base de datos (CRM) y cambiamos el ID de las casillas de verificación en tiempo de ejecución, al agregar el número de cuenta a éste (esto es requerido para hacer seguimiento de la selección del usuario)
En este momento se estarán preguntando, ¿en donde se toman los valores almacenados del lado del cliente y se llama a CRM? Estoy seguro que muchos saben la respuesta… El manejador del evento click del botón.
private void btnUpdate_Click(object sender, EventArgs e) {
int rowIndex = 0;
int selectedCell = 0;
DataView sorted = null;
string[] splitData = null;
string selectedControl = string.Empty;
HiddenField clientSideSelection = null;
StringBuilder userSelection = new StringBuilder();
GridView grdSearchResults = FindControl("grdSearchResults") as GridView;
HiddenField checkboxes_cleared = FindControl("checkboxes_cleared") as
HiddenField;
// Is there any data to continue?
if ((SearchResults != null && SearchResults.Rows.Count > 0) &&
(grdSearchResults != null && grdSearchResults.Rows.Count > 0)) {
// Let's create a view to deal with the data as it is displayed
sorted = new DataView(SearchResults);
sorted.Sort = GridViewSortCondition;
// Let's sync both server and client checkboxes selection
ManageDeselectedItemsOnClientSide();
// Should we combine client-side selection with data from the DB?
if ((clientSideSelection = FindControl("selected_checkboxes") as HiddenField) != null
&& !string.IsNullOrEmpty(clientSideSelection.Value)) {
splitData = clientSideSelection.Value.Split(';');
var includeQuery = from checkBoxName in splitData.ToList()
.Where(controlName => !string.IsNullOrEmpty(controlName) &&
SelectedCheckBoxes.IndexOf(controlName) == -1 &&
controlName.IndexOf("_REMOVED") == -1)
select checkBoxName;
// Is there any item we must include?
if (includeQuery.ToList().Count > 0)
includeQuery.ToList().ForEach(controlToInclude =>
userSelection.Append(string.Format("{0};",
controlToInclude.Substring(controlToInclude.IndexOf("chkItem")))));
// Let's combine it with the information coming from the DB
userSelection.Append(SelectedCheckBoxes);
} else if (checkboxes_cleared != null && // Has the user interacted with the UI?
string.IsNullOrEmpty(checkboxes_cleared.Value))
userSelection.Append(SelectedCheckBoxes);
// Accounts to update
splitData = userSelection.ToString().Split(';');
// We loop through the control names collection
// (by default, we unblock those unselected checkboxes)
foreach (string checkBoxId in splitData) {
rowIndex = 0;
foreach (GridViewRow selectedRow in grdSearchResults.Rows) {
selectedCell = selectedRow.Cells.Count - 1;
// Is it the right one?
if (selectedRow.Cells[selectedCell].Controls.Count > 0 &&
selectedRow.Cells[selectedCell].Controls[0] is CheckBox) {
selectedControl = ((CheckBox)selectedRow.Cells[selectedCell].Controls[0]).ID;
if (selectedControl.Equals(checkBoxId)) {
// Account to block
sorted[rowIndex]["isSelected"] = true;
break;
} else if (!splitData.Contains(selectedControl)) {
// Account to unblock (default behaviour)
sorted[rowIndex]["isSelected"] = false;
}
}
rowIndex++;
}
}
// Block/Unblock accounts based on the user's selection
var accountsToProcess = from currentSelection in sorted.Table
.AsEnumerable()
select new {
AccountId = currentSelection["accountId"].ToString(),
IsSelected = (bool)currentSelection["isSelected"]
};
accountsToProcess.ToList().ForEach(account =>
OrderBlockManagement.updateAccount(account.AccountId,
account.IsSelected));
// Let's clear DB selection information
SelectedCheckBoxes = string.Empty;
// Let's reflect the changes on the UI
GridBindingHelper();
}
}
El resultado final es mostrado abajo
Nuestro webpart integrando diferentes sistemas como mencioné al principio del post además de proveer una interfaz rápida al usuario.
Saludos,
Angel
Hi Community,
Please feel free to download the slide deck and demos for my break out session at Tech-Ed
Regards,
Angel
Hola Comunidad,
Por favor siéntanse libres de descargar las láminas y demos de mi sesión en Tech-Ed
Saludos,
Angel
Hi Community,
Please feel free to watch my Tech Talk interview with Catherine Eibner
Regards,
Angel
Hola Comunidad,
Por favor siéntanse libres de ver mi charla técnica con Catherine Eibner
Saludos,
Angel
Hi Community,
Please feel free to download the attached Slide Deck and demo used for today's "Blue Theater" Session and Yesterday's Tech Talk with Catherine Eibner for TechEd Online.
Regards,
Angel
Hola Comunidad,
Por favor siéntanse libres de descargar la presentación y demo utlizados en la sesión en "Blue Theater" de hoy y la charla técnica de ayer con Catherine Eibner para TechEd Online.
Saludos,
Angel
These are my upcoming speaking engagements:
Regards,
Angel
Estas son mis próximas ponencias y/o charlas:
Saludos,
Angel
This post is about a question that many developers ask themselves, ¿how can I get the SharePoint version programmatically?
There are at least three possible ways to do this (that I'm aware of):
In my case, I'm always using WMI and Windows provides us with the right tool (WBEMTEST.EXE) to try our WQL queries
Regards,
Angel
Este post trata de una pregunta que muchos desarrolladores se hacen, ¿cómo determino la versión de SharePoint que tengo instalada por medio de código?
A continuación enumero las tres posibles maneras (que yo conozco para hacerlo):
En mi caso particular, siempre hago uso de WMI y Windows nos ofrece WBEMTEST.EXE para probar nuestras consultas WQL
Saludos,
Angel
Writing code nowadays can be considered as a "cheerful experience", we have a complete set of tools and development environments that make us more productive, helping us in the delivery of high-quality software. 15 years ago it was a hard task to find IDEs with Intellisense support, we used to store our application settings in .INI files and we didn't even have any virtualization environment in order to perform tests before sending out our solution to the customers, that's why it's a new adventure each time we got engaged on a new development project and coding hasn't ever been as enjoyable as it is today. A few years ago I read a book written by Jeffrey Richter called "Advanced Windows" and I learned heaps from this book, as a matter of fact, last year I bought his latest book called "Windows via C/C++" which is a classic and a must have in every Windows developer toolbox, same thing occurs with Charles Petzold's book, however I learned a lot about Windows internal mechanisms by reading Richter's, for instance, how can processes exchange information between them? What's the difference between a Mutex and a Semaphore for threads synchronization? just to mention a few. Windows is a product that has evolved and grown throughout the years, despite of new APIs and functionality that have been included into it we will always need access to the PC's memory, for example we can mention Inter-Process Communication (IPC) and virtual memory access, this post is about it pretty much, how can I get access to the virtual memory so I can address and use a given amount of bytes? I think you know already what I'm talking about. Memory Mapped Files can be defined as "a segment of virtual memory which has been assigned a direct byte-for-byte correlation with some portion of a file or file-like resource. This resource is typically a file that is physically present on-disk, but can also be a device, shared memory object or other resource that the operating system can reference through a file descriptor". They have three main purposes which are:
-
Load and execute .exe files and DLLs
-
Access files on-disk very quickly without using buffers. Windows handles this for us
-
Allow many processes to share information in memory
The following diagram depicts some implementations for memory mapped files
Once the file is mapped in memory, a view based on that file needs to be created, I mean, a region or space within the file which allows a process to read and write onto it, when creating the view we specify bytes to be mapped, file offset and desired access, this process is shown in the image below
One of the new features bundled in .NET Framework 4.0 besides DLR (Dynamic Language Runtime), parallel processing and support for memory mapped files. The code that accompanies this post demonstrates how to do it using the new MemoryMappedFile class plus my own implementation with .NET 3.5, C# and a little of Interop.
Further reading about this topic can be found at Virtual Memory and Paging
Regards,
Angel
Format: swf
Duration: 22 min
Hoy en día es ameno “echar código”, pues tenemos herramientas de desarrollo que nos ayudan a ser productivos y aseguramos la calidad del software que hacemos. Hace 15 años atrás era difícil encontrar entornos de desarrollo con Intellisense, guardábamos nuestra configuración en archivos .INI y no disponíamos de virtualización para probar nuestra aplicación antes de enviarla al cliente, por eso en la actualidad cada día es una aventura y programar es más divertido. Hace unos años atrás leí un libro escrito por Jeffrey Richter titulado “Programación avanzada para Windows 95/NT”, aprendí mucho de ese libro de hecho el año pasado me compré la última edición del mismo titulado “Windows via C/C++”, es un clásico como el libro de Charles Petzold, sin embargo con el libro de Richter aprendí cosas sobre el funcionamiento interno del Sistema Operativo, por ejemplo, ¿cómo los procesos pueden intercambiar información entre sí?, ¿cuál es la diferencia entre un Mutex y un semáforo para sincronizar hilos? y muchas otras cosas más. Windows es un producto que ha madurado y crecido con el paso de los años, a pesar de nuevas APIs y funcionalidad siempre tendremos que tener acceso a la memoria del PC, un ejemplo de esto es la comunicación entre procesos (IPC) y el acceso a la memoria virtual, el siguiente post trata de eso, ¿cómo utilizar la memoria virtual para así direccionar y/o utilizar una cantidad de memoria que yo desee? Me imagino, ya saben a lo que me refiero y es Archivos mapeados en memoria (Memory Mapped Files) que lo podríamos definir como “un segmento de la memoria virtual la cual se le ha asignado una correlación directa byte a byte con alguna porción de un archivo ó algún otro recurso, el cuál típicamente es un archivo en disco pero también puede ser un dispositivo, un objeto compartido de memoria u otro recurso que el sistema operativo pueda referenciar a través de un descriptor de archivo”. Los archivos mapeados en memoria tienen tres propósitos los cuales son:
-
Cargar y ejecutar archivos .exe y bibliotecas de enlace dinámico (DLL)
-
Accesar archivos en disco rápidamente sin necesidad de usar búferes, Windows se encarga de esto por nosotros
-
Permitir a varios procesos compartir la misma información en memoria
El siguiente diagrama muestra alguna de las implementaciones para los archivos mapeados en memoria
Una vez que el archivo se encuentra mapeado en memoria, es necesario crear una vista basada en el mismo, es decir, una región ó espacio dentro del archivo a la cuál un proceso puede leer y escribir, al momento de crear la vista se especifica desde donde se va a comenzar a manipular el archivo así como el número de bytes a utilizar, este proceso lo podemos ver en la imagen mostrada a continuación
Una de las características nuevas que trae consigo el .NET Framework 4.0 además de DLR (Dynamic Language Runtime), procesamiento paralelo y muchas otras más, es el soporte para archivos mapeados en memoria, el código de este posting demuestra como hacerlo con la nueva clase MemoryMappedFile así como mi propia implementación con .NET 3.5, C# y un poco de Interop.
Si deseas leer un poco más sobre este tema puedes revisar estos links en Wikipedia: Memoria Virtual y Paginación de Memoria
Saludos,
Angel
Format: swf
Duration: 38 Min
Structured Exception Handling (SEH), it’s one of the top used features by developers, it’s not required any more to be dealing with On Error labels or any other less elegant mechanism to handle error conditions, however .NET Framework is an existing layer between our application and the operating system, even when the Exception class describes an error condition it doesn’t provide the error code and message associated to the operating system, that’s why it’s frustrating sometimes to interpret some “less-descriptive” exceptions and hence recovering any Windows error information can be of great help. Throughout the years Windows API has evolved and grown in size and error messages as well, many of these messages can be found in the SDK and WDK, so I had this idea about extending the Exception class by implementing an extension method. Extension methods enable us to “add” methods to existing types without creating a new derived type, recompiling, or otherwise modifying the original type. They are a special kind of static method, but they are called as if they were instance methods on the extended type. The most common extension methods are the LINQ standard query operators that add query functionality to the existing IEnumerable and IEnumerable<T>.
To retrieve the description for a given error code we use the FormatMessage function, as it’s shown below
[DllImport("kernel32.dll", CharSet = CharSet.Auto, SetLastError = true,
CallingConvention = CallingConvention.Winapi)]
public static extern int FormatMessage(int dwFlags, int lpSource, int dwMessageId,
int dwLanguageId, out StringBuilder lpBuffer,
int nSize, IntPtr va_list);
The extension method used is the following
public static class ExceptionExtension {
/// <summary>
/// Gets the win32 error description.
/// </summary>
/// <param name="ex">The ex.</param>
/// <returns></returns>
public static Win32Error GetWin32ErrorDescription(this Exception ex) {
Win32Error retval = Win32Error.Empty;
StringBuilder pMsgBuf = new StringBuilder();
int lastError = FormatErrorHelper.GetLastError();
FormatErrorHelper.FormatMessage(
FormatErrorHelper.FORMAT_MESSAGE_ALLOCATE_BUFFER |
FormatErrorHelper.FORMAT_MESSAGE_FROM_SYSTEM,
FormatErrorHelper.NULL, lastError,
FormatErrorHelper.MakeLangID(FormatErrorHelper.LANG_NEUTRAL,
FormatErrorHelper.SUBLANG_DEFAULT), out pMsgBuf, 0, IntPtr.Zero);
retval.ErrorCode = lastError;
retval.Description = pMsgBuf != null ? pMsgBuf.ToString() : string.Empty;
return retval;
}
}
So if we catch any exception, besides obtaining the exception object itself we get the error and message code from the operating system
Win32Error osError = Win32Error.Empty;
// Some funky exception here...
osError = ex.GetWin32ErrorDescription();
Console.WriteLine(string.Format("Code: {0} - Description: {1}",
new object[] {osError.ErrorCode, osError.Description}));
If we implement everything we’ve previously mentioned, see the difference between the exception message and the message provided by the operating system
![image_thumb[2]](http://msmvps.com/cfs-file.ashx/__key/CommunityServer.Blogs.Components.WeblogFiles/angelhernandez.metablogapi/5810.image_5F00_thumb2_5F00_6E1B5EFB.png "image_thumb[2]")
Hope this helps,
Regards,
Angel
El manejo estructurado de excepciones, es una de las características que como desarrolladores utilizamos más a menudo, ya no es necesario estar implementando etiquetas On Error u otro mecanismo poco elegante para manejar condiciones de error, sin embargo el .NET Framework es una capa entre el sistema operativo y nuestra aplicación, aunque la clase Exception describe una condición de error, carece del código de error y descripción asociado al sistema operativo, en muchas ocasiones el interpretar algunas excepciones puede llegar en cierto punto ser algo frustrante porque no es muy descriptivo, por lo que no sería mala idea recuperar la información de error emitida por Windows. Con el paso de los años, el API de Windows ha crecido y así mismo la cantidad de mensajes de error subyacentes, en el SDK de la plataforma se consigue información sobre estos mensajes de error y lo mismo sucede con el WDK, por lo que se me ocurrió la idea de extender la clase Exception a través de un método de extensión. Los métodos de extensión, nos permiten “agregar “métodos a tipos de datos existentes sin la necesidad de crear un tipo derivado, recompilar ó modificar el tipo original. Estos métodos son un tipo especial de método estático que son llamados como si fueran métodos de instancia en el tipo extendido. Su uso más común es con LINQ para así conseguir funcionalidad adicional de los tipos IEnumerable e IEnumerable<T>.
Para recuperar la descripción de un código de error utilizamos la función FormatMessage, como mostramos a continuación
[DllImport("kernel32.dll", CharSet = CharSet.Auto, SetLastError = true,
CallingConvention = CallingConvention.Winapi)]
public static extern int FormatMessage(int dwFlags, int lpSource, int dwMessageId,
int dwLanguageId, out StringBuilder lpBuffer,
int nSize, IntPtr va_list);
El método de extensión utilizado con la clase Exception es el siguiente
public static class ExceptionExtension {
/// <summary>
/// Gets the win32 error description.
/// </summary>
/// <param name="ex">The ex.</param>
/// <returns></returns>
public static Win32Error GetWin32ErrorDescription(this Exception ex) {
Win32Error retval = Win32Error.Empty;
StringBuilder pMsgBuf = new StringBuilder();
int lastError = FormatErrorHelper.GetLastError();
FormatErrorHelper.FormatMessage(
FormatErrorHelper.FORMAT_MESSAGE_ALLOCATE_BUFFER |
FormatErrorHelper.FORMAT_MESSAGE_FROM_SYSTEM,
FormatErrorHelper.NULL, lastError,
FormatErrorHelper.MakeLangID(FormatErrorHelper.LANG_NEUTRAL,
FormatErrorHelper.SUBLANG_DEFAULT), out pMsgBuf, 0, IntPtr.Zero);
retval.ErrorCode = lastError;
retval.Description = pMsgBuf != null ? pMsgBuf.ToString() : string.Empty;
return retval;
}
}
Por lo que al atrapar alguna excepción, además de obtener el objeto Exception como tal también obtengo el código y mensaje de error del sistema operativo
Win32Error osError = Win32Error.Empty;
// Some funky exception here...
osError = ex.GetWin32ErrorDescription();
Console.WriteLine(string.Format("Code: {0} - Description: {1}",
new object[] {osError.ErrorCode, osError.Description}));
Si ponemos en práctica todo lo mencionado anteriormente, vean lo diferente que es el mensaje de proporcionado por el .NET Framework al del sistema operativo
Espero sea de utilidad,
Saludos,
Angel
I usually read a lot about technology stuff from Microsoft and other vendors as well. One of my favourite sites to gather information about ongoing projects and things to come is Microsoft Research so today after having lunch I started to browse for some projects and I found Pex which in my humble opinion seems to be really interesting, why? Because it’s a new tool that helps in understanding the behavior of .NET code, debugging issues, and in creating a test suite that covers all corner cases – fully automatically. Through a context menu in the code editor, the user can invoke Pex to analyze an entire class or a single method. For any method, Pex computes and displays interesting input-output pairs. Pex systematically hunts for bugs – exceptions or assertion failures. As Pex discovers boundary conditions in code, Pex generates new tests that target these conditions. The result is a small test suite with high code coverage. Pex enables Parameterized Unit Testing, an extension of traditional unit testing that reduces test maintenance costs. Pex has been used in Microsoft to test core .NET components. Pex is developed at Microsoft Research and is integrated into Microsoft Visual Studio, so if you’re like me that loves to deliver high quality code then Pex is a must in your toolbox.
Regards,
Angel
Usualmente leo bastante sobre tecnologías Microsoft y otros fabricantes también. Uno de mis sitios favoritos para conseguir información acerca de proyectos en ejecución y cosas por venir es Microsoft Research así que hoy después de almorzar comencé a navegar por algunos proyectos y encontré Pex, el cual en mi humilde opinión parece ser realmente interesante, ¿por qué? Pues porque es una nueva herramienta que ayuda a entender el comportamiento de código .NET, problemas de depuración y en la creación de pruebas para la mayoría de los casos – todo esto de manera automática. A través de un menú contextual en el editor de código, el usuario puede invocar Pex para analizar una clase entera ó un sólo método. Para cualquier método, Pex calcula y muestra las entradas y las salidas. Pex sistemáticamente busca bugs – excepciones ó fallas de aserciones. Mientras Pex descubre las condiciones de límites en el código, a su vez genera nuevas pruebas para dichas condiciones. El resultado es una pequeña suite con alta cobertura de código. Pex permite realizar prueba unitarias parametrizadas, que son una extensión de las pruebas unitarias tradicionales que reducen los costos en el mantenimiento de las pruebas. Pex ha sido utilizado por Microsoft para probar componentes importantes de .NET. Pex es desarrollado en Microsoft Research y está integrado con Microsoft Visual Studio, así que si eres como yo que le encanta entregar código de alta calidad entonces Pex es indispensable en tu caja de herramientas.
Saludos
Yesterday afternoon I was coming home on the train, when suddenly I had this idea of writing something about policy injection, function pointers and my preference of using .NET instead of Java. Firstly, allow me to get started with policy injection, it’s a definition commonly used in developer talks but, what does it mean? Or what is it about? In a few words we can define it as “one technique becoming increasingly popular by the adoption of Aspect Oriented Programming (AOP), which provides an array of mechanisms to change the behavior of business objects and other classes by applying policies thus making it easier to implement crosscutting concerns such as logging, validation, exception handling, caching and more”.
The terminology of AOP uses the word “concern” to mean a task or feature of an application. Core concerns are the features usually unique to a class or object, for example, a class to connect to a database or serialize an object to disk. Tasks common to more than one class or object are crosscutting concerns. Poor management of these can result in duplicated and hard-to-manage code, and unreliable applications.
Policy injection was introduced by the Microsoft patterns & practices group into the release (version 3.0) of their Enterprise Library which integrates with the other application blocks in the library. The latest version can be downloaded here
However, I would like to share with you the idea or principle behind “Policy injection”, if you’re like me that prefer designing and writing all the components for an application by yourself, this can be a useful post to you then.
In my humble opinion, we have at least two possible ways to “inject” either by using generics or through callbacks (delegates), let’s proceed to describe them:
1-.Using Generics: Generics was introduced in .NET 2.0 and it gives us the ability to write “generic” code plus enforcing type-safety, I mean, no more casting to object to achieve multiple reuse. Let’s suppose we have the following class
It implements the ISample interface which contains a single method PerformAction that throws an exception if the time’s current second is even (we do this to simulate that something goes wrong during the execution of this method), please note that we don’t have any exception handling in that method. Now, let’s suppose that I have heaps of classes implementing the same interface without any exception handling mechanism too, what could I possibly do in that case? If we take into consideration the definitions for concern and crosscutting concern we can then say: I have different concerns (even when the implemented interface is the same) but at the same time I have crosscutting concerns, I mean, all of my classes implement this interface without a proper exception handling mechanism.
Having said that, let’s have a look at the following code fragment
It’s a generic class and its usage is restricted to ISample objects plus requires the creation of a new instance, in its Ctor, through Reflection we collect information of the class used with the generic class, I mean, the object we’ll be applying our “policy” to. The generic class contains only a method called Execute which invokes the PerformAction method of our class implementing ISample hence exception handling is added which it didn’t exist in our original implementation.
The implementation on the client or application would be like this
Please note that I didn’t have to modify the original code but I’m using another mechanism instead, in this case “Policy injection” through a generic class that acts as a proxy which allows me to achieve what I want plus isolating concerns from each other but considering crosscutting concerns.
2-.Using Callbacks: Callbacks, function pointers or delegates refer to the same mechanism for passing executable code as an argument to another code. In old C/C++ school we can do it like this
However in .NET we do it like this
So the way we can achieve “policy injection” through callbacks differs a little from the one previously explained, which is cleaner, easier to maintain and suggested from my point of view, because we use a proxy that communicates with the class we’re applying the policy to; if we’re using callback it would be something like this
The greater difference between the two approaches, it’s that policy code is on the client side and not in the proxy class, so code is repeated and future maintainability is at risk. We can achieve what we want but it’s not the best approach to follow, however this method it’s pretty useful in situations such as, impersonating a given user to execute a piece of code.
Changing topics and almost wrapping up, one of the reasons I prefer .NET instead of Java it’s because of Callbacks support which have been with us for a long time ago and they’re a pretty useful as we can see, for instance, the qsort function (quicksort implementation in CRT) which expects to receive as last parameter a Callback to a comparison function. In Java we don’t have this functionality out-of-the-box but we need to “simulate it” by implementing interfaces (Callback pattern) and I start wondering, what the heck? I can do this in C/C++ using native code and in any other .NET language without implementing any “pattern” so I can focus on the requirement, for reasons like this and many other reasons I prefer .NET instead of Java because it just makes me more productive.
Regards,
Angel
More Posts
Next page »