Search Results

Search found 15634 results on 626 pages for 'foreach loop container'.

Page 124/626 | < Previous Page | 120 121 122 123 124 125 126 127 128 129 130 131  | Next Page >

• Shuffle Two NSMutableArray independently

  - by Superman

  I'm creating two NSMutableArray in my viewDidLoad, I add it in a NSMutableDictionary. When I tried shuffling it with one array, It is okay. But the problem is when Im shuffling two arrays independently its not working,somehow the indexes got mixed up. Here is my code for my array (I have two of these): self.items1 = [NSMutableArray new]; for(int i = 0; i <= 100; i++) { NSArray *paths = NSSearchPathForDirectoriesInDomains(NSCachesDirectory, NSUserDomainMask, YES); NSString *documentsDir = [paths objectAtIndex:0]; NSString *savedImagePath = [documentsDir stringByAppendingPathComponent:[NSString stringWithFormat:@"[Images%d.png", i]]; if([[NSFileManager defaultManager] fileExistsAtPath:savedImagePath]){ self.container = [[NSMutableDictionary alloc] init]; [container setObject:[UIImage imageWithContentsOfFile:savedImagePath] forKey:@"items1"]; [container setObject:[NSNumber numberWithInt:i] forKey:@"index1"]; [items1 addObject:container]; } } NSLog(@"Count : %d", [items1 count]); [items1 enumerateObjectsUsingBlock:^(id object, NSUInteger index, BOOL *stop) { NSLog(@"%@ images at index %d", object, index); }]; then my shuffle code(Which I tried duplicating for the other array,but not working also): srandom(time(NULL)); NSUInteger count = [items1 count]; for (NSUInteger i = 0; i < count; ++i) { int nElements = count - i; int n = (random() % nElements) + i; [items1 exchangeObjectAtIndex:i withObjectAtIndex:n]; } How am I going to shuffle it using above code (or if you have other suggestions) with two arrays? Thanks My other problem is when I tries subclassing the class for the shuffle method or either use the above code, their index mixed. For example: Object: apple, ball, carrots, dog Indexes: 1 2 3 4 but in my View when shuffled: Object: carrots, apple, dog, balle Indexes: 2 4 1 3

  Read the article

• Visitor Pattern can be replaced with Callback functions?

  - by getit

  Is there any significant benefit to using either technique? In case there are variations, the Visitor Pattern I mean is this: http://en.wikipedia.org/wiki/Visitor_pattern And below is an example of using a delegate to achieve the same effect (at least I think it is the same) Say there is a collection of nested elements: Schools contain Departments which contain Students Instead of using the Visitor pattern to perform something on each collection item, why not use a simple callback (Action delegate in C#) Say something like this class Department { List Students; } class School { List Departments; VisitStudents(Action<Student> actionDelegate) { foreach(var dep in this.Departments) { foreach(var stu in dep.Students) { actionDelegate(stu); } } } } School A = new School(); ...//populate collections A.Visit((student)=> { ...Do Something with student... }); *EDIT Example with delegate accepting multiple params Say I wanted to pass both the student and department, I could modify the Action definition like so: Action class School { List Departments; VisitStudents(Action<Student, Department> actionDelegate, Action<Department> d2) { foreach(var dep in this.Departments) { d2(dep); //This performs a different process. //Using Visitor pattern would avoid having to keep adding new delegates. //This looks like the main benefit so far foreach(var stu in dep.Students) { actionDelegate(stu, dep); } } } }

  Read the article

• does entity framework or mysql provider swallows timeout exceptions on enumeration of result?!

  - by Freddy Rios

  I'm trying to make sense of a situation I have using entity framework on .net 3.5 sp1 + MySQL 6.1.2.0 as the provider. It involves the following code: Response.Write("Products: " + plist.Count() + "<br />"); var total = 0; foreach (var p in plist) { //... some actions total++; //... other actions } Response.Write("Total Products Checked: " + total + "<br />"); Basically the total products is varying on each run, and it isn't matching the full total in plist. Its varies widely, from ~ 1/5th to half. There isn't any control flow code inside the foreach i.e. no break, continue, try/catch, conditions around total++, anything that could affect the count. As confirmation, there are other totals captured inside the loop related to the actions, and those match the lower and higher total runs. I don't find any reason to the above, other than something in entity framework or the mysql provider that causes it to end the foreach when retrieving an item. The body of the foreach can have some good variation in time, as the actions involve file & network access, my best shot at the time is that when it takes beyond certain threshold there is some type of timeout in the underlying framework/provider and instead of causing an exception it is silently reporting no more items for enumeration. Can anyone give some light in the above scenario and/or confirm if the entity framework/mysql provider has the above behavior?

  Read the article

• Display two array's in the same table

  - by Naeem Ahmed

  $row = $query->fetchAll(PDO::FETCH_ASSOC); $num_rows = count($row); for ($i = 0; $i < $num_rows; $i++) { $title = htmlspecialchars($row[$i]['title']); $author =htmlspecialchars($row[$i]['author']); $school =htmlspecialchars($row[$i]['school']); $solution = $row[$i]['solution']; $notes = $row[$i]['notes']; $ad = array($title, $price, $author, $school, $contact, $content, $date); $inlcude = array($solutions, $notes); $field = 0; echo "<table border='1'>"; // foreach($inlcude as $in) This failled miserably foreach ($ad as $post) { if ($field < 3) //The first three values are placed in the first row { echo "<td>$post</td>"; } if ($field >= 3) { echo "<tr><td>$post</td><td>$in</td></tr>"; } $field++; } echo '</table>'; } I have two arrays and I would like to display them in different columns in my table. $ad displays perfectly fine but I'm having trouble displaying the contents in $inlcude in the second column. I've tried putting another foreach loop to iterate through contents of the second array but that really screws up my table by placing random values in different places on the table. Besides the foreach loop, I don't know of any other way to iterate through the array. Any suggestions would be appreciated.Thanks!

  Read the article

• How to create a generic method in C# that's all applicable to many types - ints, strings, doubles et

  - by satyajit

  Let's I have a method to remove duplicates in an integer Array public int[] RemoveDuplicates(int[] elems) { HashSet<int> uniques = new HashSet<int>(); foreach (int item in elems) uniques.Add(item); elems = new int[uniques.Count]; int cnt = 0; foreach (var item in uniques) elems[cnt++] = item; return elems; } How can I make this generic such that now it accepts a string array and remove duplicates in it? How about a double array? I know I am probably mixing things here in between primitive and value types. For your reference the following code won't compile public List<T> RemoveDuplicates(List<T> elems) { HashSet<T> uniques = new HashSet<T>(); foreach (var item in elems) uniques.Add(item); elems = new List<T>(); int cnt = 0; foreach (var item in uniques) elems[cnt++] = item; return elems; } The reason is that all generic types should be closed at run time. Thanks for you comments

  Read the article

• HTML5 Video Element on iPad doesn't fire onclick?

  - by bhups

  I am using the video element in my HTML as following:<div id="container" style="background:black; overflow:hidden;width:320px;height:240px" <video style="background:black;display:block" id="vdo" height="240px" width="320px" src="http://mydomain/vid.mp4"</video</div And in javascript I am doing this:var video=document.getElementById('vdo'); var container=document.getElementById('container'); video.addEventListener('click', function(e) { e.preventDefault(); console.log("clicked"); }, false); container.addEventListener('click', function(e) { e.preventDefault(); console.log("clicked"); }, false); On desktop safari/chrome everything is working fine. I can see two "clicked" in the console. But on ipad there is nothing. First I tried with iOS versin 3.2, then I updated it to the latest one 4.2.1 without any success.I found a similar question HTML5 Video Element on iPad doesn't fire onclick or touchstart events? where it suggests not to use controls in video tag and I am not using it.

  Read the article

• Enlist a table's columns in other component

  - by bungrudi

  The main goal is to have a dropdown menu where each of its menuItems represents one column of a <rich:extendedDataTable />. Above the table I have this: <rich:dropDownMenu value="Column visibility" submitMode="none" direction="bottom-right"> <c:forEach var="columnConfigVO" items="#{gridConfigurationManager.getColumnConfigs(listId)}"> <rich:menuItem value="columnConfigVO.columnId" /> </c:forEach> </rich:dropDownMenu> And then bellow that I have the usual <rich:extendedDataTable /> with its columns. I register the table columns to gridConfigurationManager component by overriding beforeRenderResponse() in ExtendedDataTable class. The problem is that <c:forEach /> is executing before renderResponse phase, thus gridConfigurationManager.getColumnConfigs(listId) return empty. The question is, how do I register the columns in gridConfigurationManager component before <c:forEach /> start executing? Or, anyone know a different approach to accomplish this? Thanks.

  Read the article

• Recursive Iterators

  - by soandos

  I am having some trouble making an iterator that can traverse the following type of data structure. I have a class called Expression, which has one data member, a List<object>. This list can have any number of children, and some of those children might be other Expression objects. I want to traverse this structure, and print out every non-list object (but I do want to print out the elements of the list of course), but before entering a list, I want to return "begin nest" and after I just exited a list, I want to return "end nest". I was able to do this if I ignored the class wherever possible, and just had List<object> objects with List<object> items if I wanted a subExpression, but I would rather do away with this, and instead have an Expressions as the sublists (it would make it easier to do operations on the object. I am aware that I could use extension methods on the List<object> but it would not be appropriate (who wants an Evaluate method on their list that takes no arguments?). The code that I used to generate the origonal iterator (that works) is: public IEnumerator GetEnumerator(){ return theIterator(expr).GetEnumerator(); } private IEnumerable theIterator(object root) { if ((root is List<object>)){ yield return " begin nest "; foreach (var item in (List<object>)root){ foreach (var item2 in theIterator(item)){ yield return item2; } } yield return " end nest "; } else yield return root; } A type swap of List<object> for expression did not work, and lead to a stackOverflow error. How should the iterator be implemented? Update: Here is the swapped code: public IEnumerator GetEnumerator() { return this.GetEnumerator(); } private IEnumerable theIterator(object root) { if ((root is Expression)) { yield return " begin nest "; foreach (var item in (Expression)root) { foreach (var item2 in theIterator(item)) yield return item2; } yield return " end nest "; } else yield return root; }

  Read the article

• Align child element to bottom with CSS

  - by alex

  I have a form input, and the label spans multiple lines, and I want the corresponding checkbox to appear at the bottom (last line of the label element). Here is what I was playing with CSS .standard-form { width: 500px; border: 1px solid red; } .standard-form .input-row { overflow: hidden; margin-bottom: 0.8em; } .standard-form label { width: 25%; float: left; } .standard-form .input-container { width: 70%; float: right; } .standard-form .checkbox .input-container { display: table-cell; height: 100%; vertical-align: text-bottom; } HTML <form class="standard-form"> <div class="input-row checkbox" id="permission"> <label for="input-permission"> Do I hereby grant you permission to do whatever tasks are neccessary to achieve an ideal outcome? </label> <div class="input-container"> <input type="checkbox" id="input-permission" name="permission" value="true" /> </div> </div> </form> It is also online at JSbin. Is there any way to do this? I notice that div.input-container isn't expanding, which is the old multi column problem with CSS. I thought I could get this going with display: table-cell and vertical-align: bottom but I haven't been able to do it yet. I don't mind that IE6/7 won't render it correctly.

  Read the article

• WPF app startup problems

  - by Dave

  My brain is all over the map trying to fully understand Unity right now. So I decided to just dive in and start adding it in a branch to see where it takes me. Surprisingly enough (or maybe not), I am stuck just getting my darn Application to load properly. It seems like the right way to do this is to override OnStartup in App.cs. I've removed my StartupUri from App.xaml so it doesn't create my GUI XAML. My App.cs now looks something like this: public partial class App : Application { private IUnityContainer container { get; set; } protected override void OnStartup(StartupEventArgs e) { container = new UnityContainer(); GUI gui = new GUI(); gui.Show(); } protected override void OnExit(ExitEventArgs e) { container.Dispose(); base.OnExit(e); } } The problem is that nothing happens when I start the app! I put a breakpoint at the container assignment, and it never gets hit. What am I missing? App.xaml is currently set to ApplicationDefinition, but I'd expect this to work because some sample Unity + WPF code I'm looking at (from Codeplex) does the exact same thing, except that it works! I've also started the app by single-stepping, and it eventually hits the first line in App.xaml. When I step into this line, that's when the app just starts "running", but I don't see anything (and my breakpoint isn't hit). If I do the exact same thing in the sample application, stepping into App.xaml puts me right into OnStartup, which is what I'd expect to happen. Argh! Is it a Bad Thing to just put the Unity construction in my GUI's Window_Loaded event handler? Does it really need to be at the App level?

  Read the article

• "text-overflow: ellipsis" not working well in firefox with floating element around

  - by Freedom

  see jsfiddle: http://jsfiddle.net/9v8faLeh/1/ I have two elements .text and .badge in a .container with a limit width: <div class="container"> <span class="badge">(*)</span> <span class="text">this is a long long long long text.</span> </div> the .badge element may not exist in a .container according to the data. if a .badge exist, I want the .badge element to float to right. and if the .text is too long, the text should ellipsis. .container { width: 150px; border: 1px solid; padding: 5px; white-space: nowrap; text-overflow: ellipsis; overflow: hidden; } .badge { float: right; margin-left: 5px; } if you open the jsfiddle link in Chrome or IE, it displays correctly as my expectation. but if open in Firefox, the .text and .badge are overlay if the text is so long. I don't want to use any JavaScript. how can I achieve the same result in FireFox?

  Read the article

• html/js: Refresh 'Select' options

  - by framara

  Hi, There's a class 'Car' with brand and model as properties. I have a list of items of this class List<Car> myCars. I need to represent in a JSP website 2 ComboBox, one for brand and another for model, that when you select the brand, in the model list only appear the ones from that brand. I don't know how to do this in a dynamic way. Any suggestion where to start? Thanks Update Ok, what I do now is send in the request a list with all the brand names, and a list of the items. The JSP code is like: <select name="manufacturer" id="id_manufacturer" onchange="return getManufacturer();"> <option value=""></option> <c:forEach items="${manufacturers}" var="man"> <option value="${man}" >${man}</option> </c:forEach> </select> <select name="model" id="id_model"> <c:forEach items="${mycars}" var="car"> <c:if test="${car.manufacturer eq man_selected}"> <option value="${car.id}">${car.model}</option> </c:if> </c:forEach> </select> <script> function getManufacturer() { man_selected = document.getElementById('id_manufacturer').value; } </script> How do I do to refresh the 'model' select options according to the selected 'man_selected' ?

  Read the article

• HTML columns or rows for form layout?

  - by Valera

  I'm building a bunch of forms that have labels and corresponding fields (input element or more complex elements). Labels go on the left, fields go on the right. Labels in a given form should all be a specific width so that the fields all line up vertically. There are two ways (maybe more?) of achieving this: Rows: Float each label and each field left. Put each label and field in a field-row div/container. Set label width to some specific number. With this approach labels on different forms will have different widths, because they'll depend on the width of the text in the longest label. Columns: Put all labels in one div/container that's floated left, put all fields in another floated left container with padding-left set. This way the labels and even the label container don't need to have their widths set, because the column layout and the padding-left will uniformly take care of vertically lining up all the fields. So approach #2 seems to be easier to implement (because the widths don't need to be set all the time), but I think it's also less object oriented, because a label and a field that goes with that label are not grouped together, as they are in approach #1. Also, if building forms dynamically, approach #2 doesn't work as well with functions like addRow(label, field), since it would have to know about the label and the field containers, instead of just creating/adding one field-row element. Which approach do you think is better? Is there another, better approach than these two?

  Read the article

• Make an abstract class or use a processor?

  - by Tim Murphy

  I'm developing a class to compare two directories and run an action when a directory/file in the source directory does not exist in destination directory. Here is a prototype of the class: public abstract class IdenticalDirectories { private DirectoryInfo _sourceDirectory; private DirectoryInfo _destinationDirectory; protected abstract void DirectoryExists(DirectoryInfo sourceDirectory, DirectoryInfo destinationDirectory); protected abstract void DirectoryDoesNotExist(DirectoryInfo sourceDirectory, DirectoryInfo destinationDirectory); protected abstract void FileExists(DirectoryInfo sourceDirectory, DirectoryInfo destinationDirectory); protected abstract void FileDoesNotExist(DirectoryInfo sourceDirectory, DirectoryInfo destinationDirectory); public IdenticalDirectories(DirectoryInfo sourceDirectory, DirectoryInfo destinationDirectory) { ... } public void Run() { foreach (DirectoryInfo sourceSubDirectory in _sourceDirectory.GetDirectories()) { DirectoryInfo destinationSubDirectory = this.GetDestinationDirectoryInfo(subDirectory); if (destinationSubDirectory.Exists()) { this.DirectoryExists(sourceSubDirectory, destinationSubDirectory); } else { this.DirectoryDoesNotExist(sourceSubDirectory, destinationSubDirectory); } foreach (FileInfo sourceFile in sourceSubDirectory.GetFiles()) { FileInfo destinationFile = this.GetDestinationFileInfo(sourceFile); if (destinationFile.Exists()) { this.FileExists(sourceFile, destinationFile); } else { this.FileDoesNotExist(sourceFile, destinationFile); } } } } } The above prototype is an abstract class. I'm wondering if it would be better to make the class non-abstract and have the Run method receiver a processor? eg. public void Run(IIdenticalDirectoriesProcessor processor) { foreach (DirectoryInfo sourceSubDirectory in _sourceDirectory.GetDirectories()) { DirectoryInfo destinationSubDirectory = this.GetDestinationDirectoryInfo(subDirectory); if (destinationSubDirectory.Exists()) { processor.DirectoryExists(sourceSubDirectory, destinationSubDirectory); } else { processor.DirectoryDoesNotExist(sourceSubDirectory, destinationSubDirectory); } foreach (FileInfo sourceFile in sourceSubDirectory.GetFiles()) { FileInfo destinationFile = this.GetDestinationFileInfo(sourceFile); if (destinationFile.Exists()) { processor.FileExists(sourceFile, destinationFile); } else { processor.FileDoesNotExist(sourceFile, destinationFile); } } } } What do you see as the pros and cons of each implementation?

  Read the article

• Need help configuring Castle-Windsor

  - by Jonathas Costa

  I have these base interfaces and providers in one assembly (Assembly1): public interface IEntity { } public interface IDao { } public interface IReadDao<T> : IDao where T : IEntity { IEnumerable<T> GetAll(); } public class NHibernate<T> : IReadDao<T> where T : IEntity { public IEnumerable<T> GetAll() { return new List<T>(); } } And I have this implementation inside another assembly (Assembly2): public class Product : IEntity { public string Code { get; set; } } public interface IProductDao : IReadDao<Product> { IEnumerable<Product> GetByCode(string code); } public class ProductDao : NHibernate<Product>, IProductDao { public IEnumerable<Product> GetByCode(string code) { return new List<Product>(); } } I want to be able to get IRead<Product> and IProductDao from the container. I am using this registration: container.Register( AllTypes.FromAssemblyNamed("Assembly2") .BasedOn(typeof(IReadDao<>)).WithService.FromInterface(), AllTypes.FromAssemblyNamed("Assembly1") .BasedOn(typeof(IReadDao<>)).WithService.Base()); The IReadDao<Product> works great. The container gives me ProductDao. But if I try to get IProductDao, the container throws ComponentNotFoundException. How can I correctly configure the registration?

  Read the article

• Can anyone help with this (Javascript arrays)?

  - by Rich

  Hi I am new to Netui and Javascript so go easy on me please. I have a form that is populated with container.item data retuned from a database. I am adding a checkbox beside each repeater item returned and I want to add the container item data to an array when one of the checkboxes is checked for future processing. The old code used Anchor tag to capture the data but that does not work for me. <!--netui:parameter name="lineupNo" value="{container.item.lineupIdent.lineupNo}" /> here is my checkbox that is a repeater. <netui:checkBox dataSource="{pageFlow.checkIsSelected}" onClick="checkBoxClicked()" tagId="pceChecked"/> this is my Javascript function so far but I want to a way to store the container.item.lineupIdent.lineupNo in the array. function checkBoxClicked() { var checkedPce = []; var elem = document.getElementById("PceList").elements; for (var i = 0; i < elem.length; i ++) { if (elem[i].name == netui_names.pceChecked) { if (elem[i].checked == true) { //do some code. } } } } I hope this is enough info for someone to help me. I have searched the web but could not find any examples. Thanks.

  Read the article

• Event-based interaction between two custom classes

  - by Antenka

  Hello everybody. I have such problem: I have 2 custom components, which have their own nesting hierarchy ... One is container for another. I have to "familiarize them" with each other. The way I'm trying to achieve that is using global events (one side is firing and the other one is catching): Application.application.addEventListener("Hello", function (data:Event):void{ // .. some actions }); //and Application.application.dispatchEvent(new Event(Hello)); Everything is pretty good, but there's one thingy .. when I'm trying to catch the event, I can't access the class, who is catching it. E.g.: Container fires the event. Child caughts it. Then should be created the connection between container and it's child. BUT, the only thing I could acheive is passing a reference to the Container in the DynamicEvent. Is there any chance that I could access the child at the event-handler function. Or maybe there's more elegant way to solve this problem ... Any help would be greately appreciated :)

  Read the article

• Basic CSS question regarding background images for divs

  - by Mike

  I'm a programmer trying to learn some css and I've already run into a stumbling block. I have the following HTML: <div class="container"> <div class="span-24 last"> Header </div> <div class="span-4"> Left sidebar </div> <div class="span-16"> <div class="span-8"> Box1 </div> <div class="span-4"> Box2 </div> <div class="span-4 last"> Box3 </div> <div class="span-16 last"> Main content </div> </div> <div class="span-4 last"> Right sidebar </div> <div class="span-24 last"> Footer </div> </div> In my css I have the following: body { background-color:#FFFFFF; } div.container { background:url(/images/bck.jpg); } I just want to display an image for the background area for the container div but nothing shows up. If I remove the background section from the css and add background-color:#000000; then I see a black background for the container div. What am I overlooking?

  Read the article

• Opera bug with JS autoselecting text (if more than 1 div)

  - by E L

  Here is HTML code. It supposed to select all text in "Container" div <B onclick="SelectText(document.getElementById('Container'));">select all text</B> <Div id="Container"> <Div>123456</Div> <Div>123456</Div> <Div onclick="SelectText();">123456</Div> </Div> here is JS code of the SelectText() function function SelectText(target){ if(target==null){ var e = window.event || e; if (!e) var e = window.event; var target=e.target || e.srcElement; } var rng, sel; if ( document.createRange ) { rng = document.createRange(); rng.selectNode( target ); sel = window.getSelection(); sel.removeAllRanges(); sel.addRange( rng ); } else { var rng = document.body.createTextRange(); rng.moveToElementText( target ); rng.select(); } } Problem is that in Opera 12.02 when "select all text" is clicked, all text seems like selected, but it's not selected (I can't rightclick it and copy). (terrific, but IE works fine with it) Why not in Opera?!!! And what can I do to make Opera 12.02 believe that all text in "Container" is selected?

  Read the article

• Embedding ADF UI Components into OAF regions

  - by Juan Camilo Ruiz

  Having finished the 2 Webcast on ADF integration with Oracle E-Business Suite, Sara Woodhull, Principal Product Manager on the Oracle E-Business Suite Applications Technology team and I are going to continue adding entries to the series on this topic, trying to cover as many use cases as possible. In this entry, Sara created an overview on how Oracle ADF pages can be embedded into an Oracle Application Framework region. This is a very interesting approach that will enable those of you who are exploring ADF as a technology stack to enhanced some of the Oracle E-Business Suite flows and leverage your skill on Oracle Applications Framework (OAF). In upcoming entries we will start unveiling the internals needed to achieve session sharing between the regions. Stay tuned for more entries and enjoy this new post.   Document Scope This document only covers information that is specific to embedding an Oracle ADF page in an Oracle Application Framework–based page. It assumes knowledge of Oracle ADF and Oracle Application Framework development. It also assumes knowledge of the material in My Oracle Support Note 974949.1, “Oracle E-Business Suite SDK for Java” and My Oracle Support Note 1296491.1, "FAQ for Integration of Oracle E-Business Suite and Oracle Application Development Framework (ADF) Applications". Prerequisite Patch Download Patch 12726556:R12.FND.B from My Oracle Support and install it. The implementation described below requires Patch 12726556:R12.FND.B to provide the accessors for the ADF page. This patch is required in addition to the Oracle E-Business Suite SDK for Java patch described in My Oracle Support Note 974949.1. Development Environments You need two different JDeveloper environments: Oracle ADF and OA Framework. Oracle ADF Development Environment You build your Oracle ADF page using JDeveloper 11g. You should use JDeveloper 11g R1 (the latest is 11.1.1.6.0) if you need to use other products in the Oracle Fusion Middleware Stack, such as Oracle WebCenter, Oracle SOA Suite, or BI. You should use JDeveloper 11g R2 (the latest is 11.1.2.3.0) if you do not need other Oracle Fusion Middleware products. JDeveloper 11g R2 is an Oracle ADF-specific release that supports the latest Java EE standards and has various core improvements. Oracle Application Framework Development Environment Build your OA Framework page using a development environment corresponding to your Oracle E-Business Suite version. You must use Release 12.1.2 or later because the rich content container was introduced in Release 12.1.2. See “OA Framework - How to find the correct version of JDeveloper to use with eBusiness Suite 11i or Release 12.x” (My Oracle Support Doc ID 416708.1). Building your Oracle ADF Page Typically you build your ADF page using the session management feature of the Oracle E-Business Suite SDK for Java as described in My Oracle Support Note 974949.1. Also see My Oracle Support Note 1296491.1, "FAQ for Integration of Oracle E-Business Suite and Oracle Application Development Framework (ADF) Applications". Building an ADF Page with the Hierarchy Viewer If you are using the ADF hierarchy viewer, you should set up the structure and settings of the ADF page as follows or the hierarchy viewer may not fill the entire area it is supposed to fill (especially a problem in Firefox). Create a stretchable component as the parent component for the hierarchy viewer, such as af:panelStretchLayout (underneath the af:form component in the structure). Use af:panelStretchLayout for Oracle ADF 11.1.1.6 and earlier. For later versions of Oracle ADF, use af:panelGridLayout. Create your hierarchy viewer component inside the stretchable component. Create Function in Oracle E-Business Suite Instance In your Oracle E-Business Suite instance, create a function for your ADF page with the following parameters. You can use either the Functions window in the System Administrator responsibility or the Functions page in the Functional Administrator responsibility. Function Function Name Type=External ADF Function (ADFX) HTML Call=GWY.jsp?targetPage=faces/<your ADF page> ">You must also add your function to an Oracle E-Business Suite menu or permission set and set up function security or role-based access control (RBAC) so that the user has authorization to access the function. If you do not want the function to appear on the navigation menu, add the function without a menu prompt. See the Oracle E-Business Suite System Administrator's Guide Documentation Set for more information. Testing the Function from the Oracle E-Business Suite Home Page It’s a good idea to test launching your ADF page from the Oracle E-Business Suite Home Page. Add your function to the navigation menu for your responsibility with a prompt and try launching it. If your ADF page expects parameters from the surrounding page, those might not be available, however. Setting up the Oracle Application Framework Rich Container Once you have built your Oracle ADF 11g page, you need to embed it in your Oracle Application Framework page. Create Rich Content Container in your OA Framework JDeveloper environment In the OA Extension Structure pane for your OAF page, select the region where you want to add the rich content, and add a richContainer item to the region. Set the following properties on the richContainer item: id Content Type=Others (for Release 12.1.3. This property value may change in a future release.) Destination Function=[function code] Width (in pixels or percent, such as 100%) Height (in pixels) Parameters=[any parameters your Oracle ADF page is expecting to receive from the Oracle Application Framework page] Parameters In the Parameters property, specify parameters that will be passed to the embedded content as a list of comma-separated, name-value pairs. Dynamic parameters may be specified as paramName={@viewAttr}. Dynamic Rich Content Container Properties If you want your rich content container to display a different Oracle ADF page depending on other information, you would set up a different function for each different Oracle ADF page. You would then set the Destination Function and Parameters properties programmatically, instead of setting them in the Property Inspector. In the processRequest() method of your Oracle Application Framework page controller, where OAFRichContentPage is the ID of your richContainer item and the parameters are whatever parameters your ADF page expects, your code might look similar to this code fragment: OARichContainerBean richBean = (OARichContainerBean) webBean.findChildRecursive("OAFRichContentPage"); if(richBean != null){ if(isFirstCondition){ richBean.setFunctionName("ADF_EXAMPLE_EMBEDDED"); richBean.setParameters("ParamLoginPersonId="+loginPersonId +"&ParamPersonId="+personId+"&ParamUserId="+userId +"&ParamRespId="+respId+"&ParamRespApplId="+respApplId +"&ParamFromOA=Y"+"&ParamSecurityGroupId="+securityGroupId); } else if(isSecondCondition){ richBean.setFunctionName("ADF_EXAMPLE_OTHER_FUNCTION"); richBean.setParameters("ParamLoginPersonId=" +loginPersonId+"&ParamPersonId="+personId +"&ParamUserId="+userId+"&ParamRespId="+respId +"&ParamRespApplId="+respApplId +"&ParamFromOA=Y" +"&ParamSecurityGroupId="+securityGroupId); } }

  Read the article

• WatiN screenshot saver

  - by Brian Schroer

  In addition to my automated unit, system and integration tests for ASP.NET projects, I like to give my customers something pretty that they can look at and visually see that the web site is behaving properly. I use the Gallio test runner to produce a pretty HTML report, and WatiN (Web Application Testing In .NET) to test the UI and create screenshots. I have a couple of issues with WatiN’s “CaptureWebPageToFile” method, though: It blew up the first (and only) time I tried it, possibly because… It scrolls down to capture the entire web page (I tried it on a very long page), and I usually don’t need that Also, sometimes I don’t need a picture of the whole browser window - I just want a picture of the element that I'm testing (for example, proving that a button has the correct caption). I wrote a WatiN screenshot saver helper class with these methods: SaveBrowserWindowScreenshot(Watin.Core.IE ie)  / SaveBrowserWindowScreenshot(Watin.Core.Element element) saves a screenshot of the browser window SaveBrowserWindowScreenshotWithHighlight(Watin.Core.Element element) saves a screenshot of the browser window, with the specified element scrolled into view and highlighted SaveElementScreenshot(Watin.Core.Element element) saves a picture of only the specified element The element highlighting improves on the built-in WatiN method (which just gives the element a yellow background, and makes the element pretty much unreadable when you have a light foreground color) by adding the ability to specify a HighlightCssClassName that points to a style in your site’s stylesheet. This code is specifically for testing with Internet Explorer (‘cause that’s what I have to test with at work), but you’re welcome to take it and do with it what you want… using System; using System.Drawing; using System.Drawing.Imaging; using System.IO; using System.Reflection; using System.Runtime.InteropServices; using System.Text; using System.Threading; using SHDocVw; using WatiN.Core; using mshtml; namespace BrianSchroer.TestHelpers { public static class WatinScreenshotSaver { public static void SaveBrowserWindowScreenshotWithHighlight (Element element, string screenshotName) { HighlightElement(element, true); SaveBrowserWindowScreenshot(element, screenshotName); HighlightElement(element, false); } public static void SaveBrowserWindowScreenshotWithHighlight(Element element) { HighlightElement(element, true); SaveBrowserWindowScreenshot(element); HighlightElement(element, false); } public static void SaveBrowserWindowScreenshot(Element element, string screenshotName) { SaveScreenshot(GetIe(element), screenshotName, SaveBitmapForCallbackArgs); } public static void SaveBrowserWindowScreenshot(Element element) { SaveScreenshot(GetIe(element), null, SaveBitmapForCallbackArgs); } public static void SaveBrowserWindowScreenshot(IE ie, string screenshotName) { SaveScreenshot(ie, screenshotName, SaveBitmapForCallbackArgs); } public static void SaveBrowserWindowScreenshot(IE ie) { SaveScreenshot(ie, null, SaveBitmapForCallbackArgs); } public static void SaveElementScreenshot(Element element, string screenshotName) { // TODO: Figure out how to get browser window "chrome" size and not have to go to full screen: var iex = (InternetExplorerClass) GetIe(element).InternetExplorer; bool fullScreen = iex.FullScreen; if (!fullScreen) iex.FullScreen = true; ScrollIntoView(element); SaveScreenshot(GetIe(element), screenshotName, args => SaveElementBitmapForCallbackArgs(element, args)); iex.FullScreen = fullScreen; } public static void SaveElementScreenshot(Element element) { SaveElementScreenshot(element, null); } private static void SaveScreenshot(IE browser, string screenshotName, Action<ScreenshotCallbackArgs> screenshotCallback) { string fileName = string.Format("{0:000}{1}{2}.jpg", ++_screenshotCount, (string.IsNullOrEmpty(screenshotName)) ? "" : " ", screenshotName); string path = Path.Combine(ScreenshotDirectoryName, fileName); Console.WriteLine(); // Gallio HTML-encodes the following display, but I have a utility program to // remove the "HTML===" and "===HTML" and un-encode the rest to show images in the Gallio report: Console.WriteLine("HTML===<div><b>{0}:</br></b><img src=\"{1}\" /></div>===HTML", screenshotName, new Uri(path).AbsoluteUri); MakeBrowserWindowTopmost(browser); try { var args = new ScreenshotCallbackArgs { InternetExplorerClass = (InternetExplorerClass)browser.InternetExplorer, ScreenshotPath = path }; Thread.Sleep(100); screenshotCallback(args); } catch (Exception ex) { Console.WriteLine(ex.Message); } } public static void HighlightElement(Element element, bool doHighlight) { if (!element.Exists) return; if (string.IsNullOrEmpty(HighlightCssClassName)) { element.Highlight(doHighlight); return; } string jsRef = element.GetJavascriptElementReference(); if (string.IsNullOrEmpty(jsRef)) return; var sb = new StringBuilder("try { "); sb.AppendFormat(" {0}.scrollIntoView(false);", jsRef); string format = (doHighlight) ? "{0}.className += ' {1}'" : "{0}.className = {0}.className.replace(' {1}', '')"; sb.AppendFormat(" " + format + ";", jsRef, HighlightCssClassName); sb.Append("} catch(e) {}"); string script = sb.ToString(); GetIe(element).RunScript(script); } public static void ScrollIntoView(Element element) { string jsRef = element.GetJavascriptElementReference(); if (string.IsNullOrEmpty(jsRef)) return; var sb = new StringBuilder("try { "); sb.AppendFormat(" {0}.scrollIntoView(false);", jsRef); sb.Append("} catch(e) {}"); string script = sb.ToString(); GetIe(element).RunScript(script); } public static void MakeBrowserWindowTopmost(IE ie) { ie.BringToFront(); SetWindowPos(ie.hWnd, HWND_TOPMOST, 0, 0, 0, 0, TOPMOST_FLAGS); } public static string HighlightCssClassName { get; set; } private static int _screenshotCount; private static string _screenshotDirectoryName; public static string ScreenshotDirectoryName { get { if (_screenshotDirectoryName == null) { var asm = Assembly.GetAssembly(typeof(WatinScreenshotSaver)); var uri = new Uri(asm.CodeBase); var fileInfo = new FileInfo(uri.LocalPath); string directoryName = fileInfo.DirectoryName; _screenshotDirectoryName = Path.Combine( directoryName, string.Format("Screenshots_{0:yyyyMMddHHmm}", DateTime.Now)); Console.WriteLine("Screenshot folder: {0}", _screenshotDirectoryName); Directory.CreateDirectory(_screenshotDirectoryName); } return _screenshotDirectoryName; } set { _screenshotDirectoryName = value; _screenshotCount = 0; } } [DllImport("user32.dll")] [return: MarshalAs(UnmanagedType.Bool)] private static extern bool SetWindowPos(IntPtr hWnd, IntPtr hWndInsertAfter, int X, int Y, int cx, int cy, uint uFlags); private static readonly IntPtr HWND_TOPMOST = new IntPtr(-1); private const UInt32 SWP_NOSIZE = 0x0001; private const UInt32 SWP_NOMOVE = 0x0002; private const UInt32 TOPMOST_FLAGS = SWP_NOMOVE | SWP_NOSIZE; private static IE GetIe(Element element) { if (element == null) return null; var container = element.DomContainer; while (container as IE == null) container = container.DomContainer; return (IE)container; } private static void SaveBitmapForCallbackArgs(ScreenshotCallbackArgs args) { InternetExplorerClass iex = args.InternetExplorerClass; SaveBitmap(args.ScreenshotPath, iex.Left, iex.Top, iex.Width, iex.Height); } private static void SaveElementBitmapForCallbackArgs(Element element, ScreenshotCallbackArgs args) { InternetExplorerClass iex = args.InternetExplorerClass; Rectangle bounds = GetElementBounds(element); SaveBitmap(args.ScreenshotPath, iex.Left + bounds.Left, iex.Top + bounds.Top, bounds.Width, bounds.Height); } /// <summary> /// This method is used instead of element.NativeElement.GetElementBounds because that /// method has a bug (http://sourceforge.net/tracker/?func=detail&aid=2994660&group_id=167632&atid=843727). /// </summary> private static Rectangle GetElementBounds(Element element) { var ieElem = element.NativeElement as WatiN.Core.Native.InternetExplorer.IEElement; IHTMLElement elem = ieElem.AsHtmlElement; int left = elem.offsetLeft; int top = elem.offsetTop; for (IHTMLElement parent = elem.offsetParent; parent != null; parent = parent.offsetParent) { left += parent.offsetLeft; top += parent.offsetTop; } return new Rectangle(left, top, elem.offsetWidth, elem.offsetHeight); } private static void SaveBitmap(string path, int left, int top, int width, int height) { using (var bitmap = new Bitmap(width, height)) { using (Graphics g = Graphics.FromImage(bitmap)) { g.CopyFromScreen( new Point(left, top), Point.Empty, new Size(width, height) ); } bitmap.Save(path, ImageFormat.Jpeg); } } private class ScreenshotCallbackArgs { public InternetExplorerClass InternetExplorerClass { get; set; } public string ScreenshotPath { get; set; } } } }

  Read the article

• Scripting with the Sun ZFS Storage 7000 Appliance

  - by Geoff Ongley

  The Sun ZFS Storage 7000 appliance has a user friendly and easy to understand graphical web based interface we call the "BUI" or "Browser User Interface".This interface is very useful for many tasks, but in some cases a script (or workflow) may be more appropriate, such as:Repetitive tasksTasks which work on (or obtain information about) a large number of shares or usersTasks which are triggered by an alert threshold (workflows)Tasks where you want a only very basic input, but a consistent output (workflows)The appliance scripting language is based on ECMAscript 3 (close to javascript). I'm not going to cover ECMAscript 3 in great depth (I'm far from an expert here), but I would like to show you some neat things you can do with the appliance, to get you started based on what I have found from my own playing around.I'm making the assumption you have some sort of programming background, and understand variables, arrays, functions to some extent - but of course if something is not clear, please let me know so I can fix it up or clarify it.Variable Declarations and ArraysVariablesECMAScript is a dynamically and weakly typed language. If you don't know what that means, google is your friend - but at a high level it means we can just declare variables with no specific type and on the fly.For example, I can declare a variable and use it straight away in the middle of my code, for example:projects=list();Which makes projects an array of values that are returned from the list(); function (which is usable in most contexts). With this kind of variable, I can do things like:projects.length (this property on array tells you how many objects are in it, good for for loops etc). Alternatively, I could say:projects=3;and now projects is just a simple number.Should we declare variables like this so loosely? In my opinion, the answer is no - I feel it is a better practice to declare variables you are going to use, before you use them - and given them an initial value. You can do so as follows:var myVariable=0;To demonstrate the ability to just randomly assign and change the type of variables, you can create a simple script at the cli as follows (bold for input):fishy10:> script("." to run)> run("cd /");("." to run)> run ("shares");("." to run)> var projects;("." to run)> projects=list();("." to run)> printf("Number of projects is: %d\n",projects.length);("." to run)> projects=152;("." to run)> printf("Value of the projects variable as an integer is now: %d\n",projects);("." to run)> .Number of projects is: 7Value of the projects variable as an integer is now: 152You can also confirm this behaviour by checking the typeof variable we are dealing with:fishy10:> script("." to run)> run("cd /");("." to run)> run ("shares");("." to run)> var projects;("." to run)> projects=list();("." to run)> printf("var projects is of type %s\n",typeof(projects));("." to run)> projects=152;("." to run)> printf("var projects is of type %s\n",typeof(projects));("." to run)> .var projects is of type objectvar projects is of type numberArraysSo you likely noticed that we have already touched on arrays, as the list(); (in the shares context) stored an array into the 'projects' variable.But what if you want to declare your own array? Easy! This is very similar to Java and other languages, we just instantiate a brand new "Array" object using the keyword new:var myArray = new Array();will create an array called "myArray".A quick example:fishy10:> script("." to run)> testArray = new Array();("." to run)> testArray[0]="This";("." to run)> testArray[1]="is";("." to run)> testArray[2]="just";("." to run)> testArray[3]="a";("." to run)> testArray[4]="test";("." to run)> for (i=0; i < testArray.length; i++)("." to run)> {("." to run)>    printf("Array element %d is %s\n",i,testArray[i]);("." to run)> }("." to run)> .Array element 0 is ThisArray element 1 is isArray element 2 is justArray element 3 is aArray element 4 is testWorking With LoopsFor LoopFor loops are very similar to those you will see in C, java and several other languages. One of the key differences here is, as you were made aware earlier, we can be a bit more sloppy with our variable declarations.The general way you would likely use a for loop is as follows:for (variable; test-case; modifier for variable){}For example, you may wish to declare a variable i as 0; and a MAX_ITERATIONS variable to determine how many times this loop should repeat:var i=0;var MAX_ITERATIONS=10;And then, use this variable to be tested against some case existing (has i reached MAX_ITERATIONS? - if not, increment i using i++);for (i=0; i < MAX_ITERATIONS; i++){ // some work to do}So lets run something like this on the appliance:fishy10:> script("." to run)> var i=0;("." to run)> var MAX_ITERATIONS=10;("." to run)> for (i=0; i < MAX_ITERATIONS; i++)("." to run)> {("." to run)>    printf("The number is %d\n",i);("." to run)> }("." to run)> .The number is 0The number is 1The number is 2The number is 3The number is 4The number is 5The number is 6The number is 7The number is 8The number is 9While LoopWhile loops again are very similar to other languages, we loop "while" a condition is met. For example:fishy10:> script("." to run)> var isTen=false;("." to run)> var counter=0;("." to run)> while(isTen==false)("." to run)> {("." to run)>    if (counter==10) ("." to run)>    { ("." to run)>            isTen=true;   ("." to run)>    } ("." to run)>    printf("Counter is %d\n",counter);("." to run)>    counter++;    ("." to run)> }("." to run)> printf("Loop has ended and Counter is %d\n",counter);("." to run)> .Counter is 0Counter is 1Counter is 2Counter is 3Counter is 4Counter is 5Counter is 6Counter is 7Counter is 8Counter is 9Counter is 10Loop has ended and Counter is 11So what do we notice here? Something has actually gone wrong - counter will technically be 11 once the loop completes... Why is this?Well, if we have a loop like this, where the 'while' condition that will end the loop may be set based on some other condition(s) existing (such as the counter has reached 10) - we must ensure that we  terminate this iteration of the loop when the condition is met - otherwise the rest of the code will be followed which may not be desirable. In other words, like in other languages, we will only ever check the loop condition once we are ready to perform the next iteration, so any other code after we set "isTen" to be true, will still be executed as we can see it was above.We can avoid this by adding a break into our loop once we know we have set the condition - this will stop the rest of the logic being processed in this iteration (and as such, counter will not be incremented). So lets try that again:fishy10:> script("." to run)> var isTen=false;("." to run)> var counter=0;("." to run)> while(isTen==false)("." to run)> {("." to run)>    if (counter==10) ("." to run)>    { ("." to run)>            isTen=true;   ("." to run)>            break;("." to run)>    } ("." to run)>    printf("Counter is %d\n",counter);("." to run)>    counter++;    ("." to run)> }("." to run)> printf("Loop has ended and Counter is %d\n", counter);("." to run)> .Counter is 0Counter is 1Counter is 2Counter is 3Counter is 4Counter is 5Counter is 6Counter is 7Counter is 8Counter is 9Loop has ended and Counter is 10Much better!Methods to Obtain and Manipulate DataGet MethodThe get method allows you to get simple properties from an object, for example a quota from a user. The syntax is fairly simple:var myVariable=get('property');An example of where you may wish to use this, is when you are getting a bunch of information about a user (such as quota information when in a shares context):var users=list();for(k=0; k < users.length; k++){     user=users[k];     run('select ' + user);     var username=get('name');     var usage=get('usage');     var quota=get('quota');...Which you can then use to your advantage - to print or manipulate infomation (you could change a user's information with a set method, based on the information returned from the get method). The set method is explained next.Set MethodThe set method can be used in a simple manner, similar to get. The syntax for set is:set('property','value'); // where value is a string, if it was a number, you don't need quotesFor example, we could set the quota on a share as follows (first observing the initial value):fishy10:shares default/test-geoff> script("." to run)> var currentQuota=get('quota');("." to run)> printf("Current Quota is: %s\n",currentQuota);("." to run)> set('quota','30G');("." to run)> run('commit');("." to run)> currentQuota=get('quota');("." to run)> printf("Current Quota is: %s\n",currentQuota);("." to run)> .Current Quota is: 0Current Quota is: 32212254720This shows us using both the get and set methods as can be used in scripts, of course when only setting an individual share, the above is overkill - it would be much easier to set it manually at the cli using 'set quota=3G' and then 'commit'.List MethodThe list method can be very powerful, especially in more complex scripts which iterate over large amounts of data and manipulate it if so desired. The general way you will use list is as follows:var myVar=list();Which will make "myVar" an array, containing all the objects in the relevant context (this could be a list of users, shares, projects, etc). You can then gather or manipulate data very easily.We could list all the shares and mountpoints in a given project for example:fishy10:shares another-project> script("." to run)> var shares=list();("." to run)> for (i=0; i < shares.length; i++)("." to run)> {("." to run)>    run('select ' + shares[i]);("." to run)>    var mountpoint=get('mountpoint');("." to run)>    printf("Share %s discovered, has mountpoint %s\n",shares[i],mountpoint);("." to run)>    run('done');("." to run)> }("." to run)> .Share and-another discovered, has mountpoint /export/another-project/and-anotherShare another-share discovered, has mountpoint /export/another-project/another-shareShare bob discovered, has mountpoint /export/another-projectShare more-shares-for-all discovered, has mountpoint /export/another-project/more-shares-for-allShare yep discovered, has mountpoint /export/another-project/yepWriting More Complex and Re-Usable CodeFunctionsThe best way to be able to write more complex code is to use functions to split up repeatable or reusable sections of your code. This also makes your more complex code easier to read and understand for other programmers.We write functions as follows:function functionName(variable1,variable2,...,variableN){}For example, we could have a function that takes a project name as input, and lists shares for that project (assuming we're already in the 'project' context - context is important!):function getShares(proj){        run('select ' + proj);        shares=list();        printf("Project: %s\n", proj);        for(j=0; j < shares.length; j++)        {                printf("Discovered share: %s\n",shares[i]);        }        run('done'); // exit selected project}Commenting your CodeLike any other language, a large part of making it readable and understandable is to comment it. You can use the same comment style as in C and Java amongst other languages.In other words, sngle line comments use://at the beginning of the comment.Multi line comments use:/*at the beginning, and:*/ at the end.For example, here we will use both:fishy10:> script("." to run)> // This is a test comment("." to run)> printf("doing some work...\n");("." to run)> /* This is a multi-line("." to run)> comment which I will span across("." to run)> three lines in total */("." to run)> printf("doing some more work...\n");("." to run)> .doing some work...doing some more work...Your comments do not have to be on their own, they can begin (particularly with single line comments this is handy) at the end of a statement, for examplevar projects=list(); // The variable projects is an array containing all projects on the system.Try and Catch StatementsYou may be used to using try and catch statements in other languages, and they can (and should) be utilised in your code to catch expected or unexpected error conditions, that you do NOT wish to stop your code from executing (if you do not catch these errors, your script will exit!):try{  // do some work}catch(err) // Catch any error that could occur{ // do something here under the error condition}For example, you may wish to only execute some code if a context can be reached. If you can't perform certain actions under certain circumstances, that may be perfectly acceptable.For example if you want to test a condition that only makes sense when looking at a SMB/NFS share, but does not make sense when you hit an iscsi or FC LUN, you don't want to stop all processing of other shares you may not have covered yet.For example we may wish to obtain quota information on all shares for all users on a share (but this makes no sense for a LUN):function getShareQuota(shar) // Get quota for each user of this share{        run('select ' + shar);        printf("  SHARE: %s\n", shar);        try        {                run('users');                printf("    %20s        %11s    %11s    %3s\n","Username","Usage(G)","Quota(G)","Quota(%)");                printf("    %20s        %11s    %11s    %4s\n","--------","--------","--------","----");                                users=list();                for(k=0; k < users.length; k++)                {                        user=users[k];                        getUserQuota(user);                }                run('done'); // exit user context        }        catch(err)        {                printf("    SKIPPING %s - This is NOT a NFS or CIFs share, not looking for users\n", shar);        }        run('done'); // done with this share}Running Scripts Remotely over SSHAs you have likely noticed, writing and running scripts for all but the simplest jobs directly on the appliance is not going to be a lot of fun.There's a couple of choices on what you can do here:Create scripts on a remote system and run them over sshCreate scripts, wrapping them in workflow code, so they are stored on the appliance and can be triggered under certain circumstances (like a threshold being reached)We'll cover the first one here, and then cover workflows later on (as these are for the most part just scripts with some wrapper information around them).Creating a SSH Public/Private SSH Key PairLog on to your handy Solaris box (You wouldn't be using any other OS, right? :P) and use ssh-keygen to create a pair of ssh keys. I'm storing this separate to my normal key:[geoff@lightning ~] ssh-keygen -t rsa -b 1024Generating public/private rsa key pair.Enter file in which to save the key (/export/home/geoff/.ssh/id_rsa): /export/home/geoff/.ssh/nas_key_rsaEnter passphrase (empty for no passphrase): Enter same passphrase again: Your identification has been saved in /export/home/geoff/.ssh/nas_key_rsa.Your public key has been saved in /export/home/geoff/.ssh/nas_key_rsa.pub.The key fingerprint is:7f:3d:53:f0:2a:5e:8b:2d:94:2a:55:77:66:5c:9b:14 geoff@lightningInstalling the Public Key on the ApplianceOn your Solaris host, observe the public key:[geoff@lightning ~] cat .ssh/nas_key_rsa.pub ssh-rsa AAAAB3NzaC1yc2EAAAABIwAAAIEAvYfK3RIaAYmMHBOvyhKM41NaSmcgUMC3igPN5gUKJQvSnYmjuWG6CBr1CkF5UcDji7v19jG3qAD5lAMFn+L0CxgRr8TNaAU+hA4/tpAGkjm+dKYSyJgEdMIURweyyfUFXoerweR8AWW5xlovGKEWZTAfvJX9Zqvh8oMQ5UJLUUc= geoff@lightningNow, copy and paste everything after "ssh-rsa" and before "user@hostname" - in this case, geoff@lightning. That is, this bit:AAAAB3NzaC1yc2EAAAABIwAAAIEAvYfK3RIaAYmMHBOvyhKM41NaSmcgUMC3igPN5gUKJQvSnYmjuWG6CBr1CkF5UcDji7v19jG3qAD5lAMFn+L0CxgRr8TNaAU+hA4/tpAGkjm+dKYSyJgEdMIURweyyfUFXoerweR8AWW5xlovGKEWZTAfvJX9Zqvh8oMQ5UJLUUc=Logon to your appliance and get into the preferences -> keys area for this user (root):[geoff@lightning ~] ssh [email protected]: Last login: Mon Dec  6 17:13:28 2010 from 192.168.0.2fishy10:> configuration usersfishy10:configuration users> select rootfishy10:configuration users root> preferences fishy10:configuration users root preferences> keysOR do it all in one hit:fishy10:> configuration users select root preferences keysNow, we create a new public key that will be accepted for this user and set the type to RSA:fishy10:configuration users root preferences keys> createfishy10:configuration users root preferences key (uncommitted)> set type=RSASet the key itself using the string copied previously (between ssh-rsa and user@host), and set the key ensuring you put double quotes around it (eg. set key="<key>"):fishy10:configuration users root preferences key (uncommitted)> set key="AAAAB3NzaC1yc2EAAAABIwAAAIEAvYfK3RIaAYmMHBOvyhKM41NaSmcgUMC3igPN5gUKJQvSnYmjuWG6CBr1CkF5UcDji7v19jG3qAD5lAMFn+L0CxgRr8TNaAU+hA4/tpAGkjm+dKYSyJgEdMIURweyyfUFXoerweR8AWW5xlovGKEWZTAfvJX9Zqvh8oMQ5UJLUUc="Now set the comment for this key (do not use spaces):fishy10:configuration users root preferences key (uncommitted)> set comment="LightningRSAKey" Commit the new key:fishy10:configuration users root preferences key (uncommitted)> commitVerify the key is there:fishy10:configuration users root preferences keys> lsKeys:NAME     MODIFIED              TYPE   COMMENT                                  key-000  2010-10-25 20:56:42   RSA    cycloneRSAKey                           key-001  2010-12-6 17:44:53    RSA    LightningRSAKey                         As you can see, we now have my new key, and a previous key I have created on this appliance.Running your Script over SSH from a Remote SystemHere I have created a basic test script, and saved it as test.ecma3:[geoff@lightning ~] cat test.ecma3 script// This is a test script, By Geoff Ongley 2010.printf("Testing script remotely over ssh\n");.Now, we can run this script remotely with our keyless login:[geoff@lightning ~] ssh -i .ssh/nas_key_rsa root@fishy10 < test.ecma3Pseudo-terminal will not be allocated because stdin is not a terminal.Testing script remotely over sshPutting it Together - An Example Completed Quota Gathering ScriptSo now we have a lot of the basics to creating a script, let us do something useful, like, find out how much every user is using, on every share on the system (you will recognise some of the code from my previous examples): script/************************************** Quick and Dirty Quota Check script ** Written By Geoff Ongley            ** 25 October 2010                    **************************************/function getUserQuota(usr){        run('select ' + usr);        var username=get('name');        var usage=get('usage');        var quota=get('quota');        var usage_g=usage / 1073741824; // convert bytes to gigabytes        var quota_g=quota / 1073741824; // as above        var quota_percent=0        if (quota > 0)        {                quota_percent=(usage / quota)*(100/1);        }        printf("    %20s        %8.2f           %8.2f           %d%%\n",username,usage_g,quota_g,quota_percent);        run('done'); // done with this selected user}function getShareQuota(shar){        //printf("DEBUG: selecting share %s\n", shar);        run('select ' + shar);        printf("  SHARE: %s\n", shar);        try        {                run('users');                printf("    %20s        %11s    %11s    %3s\n","Username","Usage(G)","Quota(G)","Quota(%)");                printf("    %20s        %11s    %11s    %4s\n","--------","--------","--------","--------");                                users=list();                for(k=0; k < users.length; k++)                {                        user=users[k];                        getUserQuota(user);                }                run('done'); // exit user context        }        catch(err)        {                printf("    SKIPPING %s - This is NOT a NFS or CIFs share, not looking for users\n", shar);        }        run('done'); // done with this share}function getShares(proj){        //printf("DEBUG: selecting project %s\n",proj);        run('select ' + proj);        shares=list();        printf("Project: %s\n", proj);        for(j=0; j < shares.length; j++)        {                share=shares[j];                getShareQuota(share);        }        run('done'); // exit selected project}function getProjects(){        run('cd /');        run('shares');        projects=list();                for (i=0; i < projects.length; i++)        {                var project=projects[i];                getShares(project);        }        run('done'); // exit context for all projects}getProjects();.Which can be run as follows, and will print information like this:[geoff@lightning ~/FISHWORKS_SCRIPTS] ssh -i ~/.ssh/nas_key_rsa root@fishy10 < get_quota_utilisation.ecma3Pseudo-terminal will not be allocated because stdin is not a terminal.Project: another-project  SHARE: and-another                Username           Usage(G)       Quota(G)    Quota(%)                --------           --------       --------    --------                  nobody            0.00            0.00        0%                 geoffro            0.05            0.00        0%                   Billy            0.10            0.00        0%                    root            0.00            0.00        0%            testing-user            0.05            0.00        0%  SHARE: another-share                Username           Usage(G)       Quota(G)    Quota(%)                --------           --------       --------    --------                    root            0.00            0.00        0%                  nobody            0.00            0.00        0%                 geoffro            0.05            0.49        9%            testing-user            0.05            0.02        249%                   Billy            0.10            0.29        33%  SHARE: bob                Username           Usage(G)       Quota(G)    Quota(%)                --------           --------       --------    --------                  nobody            0.00            0.00        0%                    root            0.00            0.00        0%  SHARE: more-shares-for-all                Username           Usage(G)       Quota(G)    Quota(%)                --------           --------       --------    --------                   Billy            0.10            0.00        0%            testing-user            0.05            0.00        0%                  nobody            0.00            0.00        0%                    root            0.00            0.00        0%                 geoffro            0.05            0.00        0%  SHARE: yep                Username           Usage(G)       Quota(G)    Quota(%)                --------           --------       --------    --------                    root            0.00            0.00        0%                  nobody            0.00            0.00        0%                   Billy            0.10            0.01        999%            testing-user            0.05            0.49        9%                 geoffro            0.05            0.00        0%Project: default  SHARE: Test-LUN    SKIPPING Test-LUN - This is NOT a NFS or CIFs share, not looking for users  SHARE: test-geoff                Username           Usage(G)       Quota(G)    Quota(%)                --------           --------       --------    --------                 geoffro            0.05            0.00        0%                    root            3.18           10.00        31%                    uucp            0.00            0.00        0%                  nobody            0.59            0.49        119%^CKilled by signal 2.Creating a WorkflowWorkflows are scripts that we store on the appliance, and can have the script execute either on request (even from the BUI), or on an event such as a threshold being met.Workflow BasicsA workflow allows you to create a simple process that can be executed either via the BUI interface interactively, or by an alert being raised (for some threshold being reached, for example).The basics parameters you will have to set for your "workflow object" (notice you're creating a variable, that embodies ECMAScript) are as follows (parameters is optional):name: A name for this workflowdescription: A Description for the workflowparameters: A set of input parameters (useful when you need user input to execute the workflow)execute: The code, the script itself to execute, which will be function (parameters)With parameters, you can specify things like this (slightly modified sample taken from the System Administration Guide):          ...parameters:        variableParam1:         {                             label: 'Name of Share',                             type: 'String'                  },                  variableParam2                  {                             label: 'Share Size',                             type: 'size'                  },execute: ....};  Note the commas separating the sections of name, parameters, execute, and so on. This is important!Also - there is plenty of properties you can set on the parameters for your workflow, these are described in the Sun ZFS Storage System Administration Guide.Creating a Basic Workflow from a Basic ScriptTo make a basic script into a basic workflow, you need to wrap the following around your script to create a 'workflow' object:var workflow = {name: 'Get User Quotas',description: 'Displays Quota Utilisation for each user on each share',execute: function() {// (basic script goes here, minus the "script" at the beginning, and "." at the end)}};However, it appears (at least in my experience to date) that the workflow object may only be happy with one function in the execute parameter - either that or I'm doing something wrong. As far as I can tell, after execute: you should only have a basic one function context like so:execute: function(){}To deal with this, and to give an example similar to our script earlier, I have created another simple quota check, to show the same basic functionality, but in a workflow format:var workflow = {name: 'Get User Quotas',description: 'Displays Quota Utilisation for each user on each share',execute: function () {        run('cd /');        run('shares');        projects=list();                for (i=0; i < projects.length; i++)        {                run('select ' + projects[i]);                shares=list('filesystem');                printf("Project: %s\n", projects[i]);                for(j=0; j < shares.length; j++)                {                        run('select ' +shares[j]);                        try                        {                                run('users');                                printf("  SHARE: %s\n", shares[j]);                                printf("    %20s        %11s    %11s    %3s\n","Username","Usage(G)","Quota(G)","Quota(%)");                                printf("    %20s        %11s    %11s    %4s\n","--------","--------","--------","-------");                                users=list();                                for(k=0; k < users.length; k++)                                {                                        run('select ' + users[k]);                                        username=get('name');                                        usage=get('usage');                                        quota=get('quota');                                        usage_g=usage / 1073741824; // convert bytes to gigabytes                                        quota_g=quota / 1073741824; // as above                                        quota_percent=0                                        if (quota > 0)                                        {                                                quota_percent=(usage / quota)*(100/1);                                        }                                        printf("    %20s        %8.2f   %8.2f   %d%%\n",username,usage_g,quota_g,quota_percent);                                        run('done');                                }                                run('done'); // exit user context                        }                        catch(err)                        {                        //      printf("    %s is a LUN, Not looking for users\n", shares[j]);                        }                        run('done'); // exit selected share context                }                run('done'); // exit project context        }        }};SummaryThe Sun ZFS Storage 7000 Appliance offers lots of different and interesting features to Sun/Oracle customers, including the world renowned Analytics. Hopefully the above will help you to think of new creative things you could be doing by taking advantage of one of the other neat features, the internal scripting engine!Some references are below to help you continue learning more, I'll update this post as I do the same! Enjoy...More information on ECMAScript 3A complete reference to ECMAScript 3 which will help you learn more of the details you may be interested in, can be found here:http://www.ecma-international.org/publications/files/ECMA-ST-ARCH/ECMA-262,%203rd%20edition,%20December%201999.pdfMore Information on Administering the Sun ZFS Storage 7000The Sun ZFS Storage 7000 System Administration guide can be a useful reference point, and can be found here:http://wikis.sun.com/download/attachments/186238602/2010_Q3_2_ADMIN.pdf

  Read the article

• More Fun with C# Iterators and Generators

  - by James Michael Hare

  In my last post, I talked quite a bit about iterators and how they can be really powerful tools for filtering a list of items down to a subset of items.  This had both pros and cons over returning a full collection, which, in summary, were:   Pros: If traversal is only partial, does not have to visit rest of collection. If evaluation method is costly, only incurs that cost on elements visited. Adds little to no garbage collection pressure.    Cons: Very slight performance impact if you know caller will always consume all items in collection. And as we saw in the last post, that con for the cost was very, very small and only really became evident on very tight loops consuming very large lists completely.    One of the key items to note, though, is the garbage!  In the traditional (return a new collection) method, if you have a 1,000,000 element collection, and wish to transform or filter it in some way, you have to allocate space for that copy of the collection.  That is, say you have a collection of 1,000,000 items and you want to double every item in the collection.  Well, that means you have to allocate a collection to hold those 1,000,000 items to return, which is a lot especially if you are just going to use it once and toss it.   Iterators, though, don't have this problem.  Each time you visit the node, it would return the doubled value of the node (in this example) and not allocate a second collection of 1,000,000 doubled items.  Do you see the distinction?  In both cases, we're consuming 1,000,000 items.  But in one case we pass back each doubled item which is just an int (for example's sake) on the stack and in the other case, we allocate a list containing 1,000,000 items which then must be garbage collected.   So iterators in C# are pretty cool, eh?  Well, here's one more thing a C# iterator can do that a traditional "return a new collection" transformation can't!   It can return **unbounded** collections!   I know, I know, that smells a lot like an infinite loop, eh?  Yes and no.  Basically, you're relying on the caller to put the bounds on the list, and as long as the caller doesn't you keep going.  Consider this example:   public static class Fibonacci {     // returns the infinite fibonacci sequence     public static IEnumerable<int> Sequence()     {         int iteration = 0;         int first = 1;         int second = 1;         int current = 0;         while (true)         {             if (iteration++ < 2)             {                 current = 1;             }             else             {                 current = first + second;                 second = first;                 first = current;             }             yield return current;         }     } }   Whoa, you say!  Yes, that's an infinite loop!  What the heck is going on there?  Yes, that was intentional.  Would it be better to have a fibonacci sequence that returns only a specific number of items?  Perhaps, but that wouldn't give you the power to defer the execution to the caller.   The beauty of this function is it is as infinite as the sequence itself!  The fibonacci sequence is unbounded, and so is this method.  It will continue to return fibonacci numbers for as long as you ask for them.  Now that's not something you can do with a traditional method that would return a collection of ints representing each number.  In that case you would eventually run out of memory as you got to higher and higher numbers.  This method, though, never runs out of memory.   Now, that said, you do have to know when you use it that it is an infinite collection and bound it appropriately.  Fortunately, Linq provides a lot of these extension methods for you!   Let's say you only want the first 10 fibonacci numbers:       foreach(var fib in Fibonacci.Sequence().Take(10))     {         Console.WriteLine(fib);     }   Or let's say you only want the fibonacci numbers that are less than 100:       foreach(var fib in Fibonacci.Sequence().TakeWhile(f => f < 100))     {         Console.WriteLine(fib);     }   So, you see, one of the nice things about iterators is their power to work with virtually any size (even infinite) collections without adding the garbage collection overhead of making new collections.    You can also do fun things like this to make a more "fluent" interface for for loops:   // A set of integer generator extension methods public static class IntExtensions {     // Begins counting to inifity, use To() to range this.     public static IEnumerable<int> Every(this int start)     {         // deliberately avoiding condition because keeps going         // to infinity for as long as values are pulled.         for (var i = start; ; ++i)         {             yield return i;         }     }     // Begins counting to infinity by the given step value, use To() to     public static IEnumerable<int> Every(this int start, int byEvery)     {         // deliberately avoiding condition because keeps going         // to infinity for as long as values are pulled.         for (var i = start; ; i += byEvery)         {             yield return i;         }     }     // Begins counting to inifity, use To() to range this.     public static IEnumerable<int> To(this int start, int end)     {         for (var i = start; i <= end; ++i)         {             yield return i;         }     }     // Ranges the count by specifying the upper range of the count.     public static IEnumerable<int> To(this IEnumerable<int> collection, int end)     {         return collection.TakeWhile(item => item <= end);     } }   Note that there are two versions of each method.  One that starts with an int and one that starts with an IEnumerable<int>.  This is to allow more power in chaining from either an existing collection or from an int.  This lets you do things like:   // count from 1 to 30 foreach(var i in 1.To(30)) {     Console.WriteLine(i); }     // count from 1 to 10 by 2s foreach(var i in 0.Every(2).To(10)) {     Console.WriteLine(i); }     // or, if you want an infinite sequence counting by 5s until something inside breaks you out... foreach(var i in 0.Every(5)) {     if (someCondition)     {         break;     }     ... }     Yes, those are kinda play functions and not particularly useful, but they show some of the power of generators and extension methods to form a fluid interface.   So what do you think?  What are some of your favorite generators and iterators?

  Read the article

• Neo4J and Azure and VS2012 and Windows 8

  - by Chris Skardon

  Now, I know that this has been written about, but both of the main places (http://www.richard-banks.org/2011/02/running-neo4j-on-azure.html and http://blog.neo4j.org/2011/02/announcing-neo4j-on-windows-azure.html) utilise VS2010, and well, I’m on VS2012 and Windows 8. Not that I think Win 8 had anything to do with it really, anyhews! I’m going to begin from the beginning, this is my first foray into running something on Azure, so it’s been a bit of a learning curve. But luckily the Neo4J guys have got us started, so let’s download the VS2010 solution: http://neo4j.org/get?file=Neo4j.Azure.Server.zip OK, the other thing we’ll need is the VS2012 Azure SDK, so let’s get that as well: http://www.windowsazure.com/en-us/develop/downloads/ (I just did the full install). Now, unzip the VS2010 solution and let’s open it in VS2012: <your location>\Neo4j.Azure.Server\Neo4j.Azure.Server.sln One-way-upgrade? Yer! Ignore the migration report – we don’t care! Let’s build that sucker… Ahhh 14 errors… WindowsAzure does not exist in the namespace ‘Microsoft’ Not a problem right? We’ve installed the SDK, just need to update the references: We can ignore the Test projects, they don’t use Azure, we’re interested in the other projects, so what we’ll do is remove the broken references, and add the correct ones, so expand the references bit of each project: hunt out those yellow exclamation marks, and delete them! You’ll need to add the right ones back in (listed below), when you go to the ‘Add Reference’ dialog make sure you have ‘Assemblies’ and ‘Framework’ selected before you seach (and search for ‘microsoft.win’ to narrow it down) So the references you need for each project are: CollectDiagnosticsData Microsoft.WindowsAzure.Diagnostics Microsoft.WindowsAzure.StorageClient Diversify.WindowsAzure.ServiceRuntime Microsoft.WindowsAzure.CloudDrive Microsoft.WindowsAzure.ServiceRuntime Microsoft.WindowsAzure.StorageClient Right, so let’s build again… Sweet! No errors.   Now we need to setup our Blobs, I’m assuming you are using the most up-to-date Java you happened to have downloaded :) in my case that’s JRE7, and that is located in: C:\Program Files (x86)\Java\jre7 So, zip up that folder into whatever you want to call it, I went with jre7.zip, and stuck it in a temp folder for now. In that same temp folder I also copied the neo4j zip I was using: neo4j-community-1.7.2-windows.zip OK, now, we need to get these into our Blob storage, this is where a lot of stuff becomes unstuck - I didn’t find any applications that helped me use the blob storage, one would crash (because my internet speed is so slow) and the other just didn’t work – sure it looked like it had worked, but when push came to shove it didn’t. So this is how I got my files into Blob (local first): 1. Run the ‘Storage Emulator’ (just search for that in the start menu) 2. That takes a little while to start up so fire up another instance of Visual Studio in the mean time, and create a new Console Application. 3. Manage Nuget Packages for that solution and add ‘Windows Azure Storage’ Now you’re set up to add the code: public static void Main() { CloudStorageAccount cloudStorageAccount = CloudStorageAccount.DevelopmentStorageAccount; CloudBlobClient client = cloudStorageAccount.CreateCloudBlobClient(); client.Timeout = TimeSpan.FromMinutes(30); CloudBlobContainer container = client.GetContainerReference("neo4j"); //This will create it as well   UploadBlob(container, "jre7.zip", "c:\\temp\\jre7.zip"); UploadBlob(container, "neo4j-community-1.7.2-windows.zip", "c:\\temp\\neo4j-community-1.7.2-windows.zip"); }   private static void UploadBlob(CloudBlobContainer container, string blobName, string filename) { CloudBlob blob = container.GetBlobReference(blobName);   using (FileStream fileStream = File.OpenRead(filename)) blob.UploadFromStream(fileStream); } This will upload the files to your local storage account (to switch to an Azure one, you’ll need to create a storage account, and use those credentials when you make your CloudStorageAccount above) To test you’ve got them uploaded correctly, go to: http://localhost:10000/devstoreaccount1/neo4j/jre7.zip and you will hopefully download the zip file you just uploaded. Now that those files are there, we are ready for some final configuration… Right click on the Neo4jServerHost role in the Neo4j.Azure.Server cloud project: Click on the ‘Settings’ tab and we’ll need to do some changes – by default, the 1.7.2 edition of neo4J unzips to: neo4j-community-1.7.2 So, we need to update all the ‘neo4j-1.3.M02’ directories to be ‘neo4j-community-1.7.2’, we also need to update the Java runtime location, so we start with this: and end with this: Now, I also changed the Endpoints settings, to be HTTP (from TCP) and to have a port of 7410 (mainly because that’s straight down on the numpad) The last ‘gotcha’ is some hard coded consts, which had me looking for ages, they are in the ‘ConfigSettings’ class of the ‘Neo4jServerHost’ project, and the ones we’re interested in are: Neo4jFileName JavaZipFileName Change those both to what that should be. OK Nearly there (I promise)! Run the ‘Compute Emulator’ (same deal with the Start menu), in your system tray you should have an Azure icon, when the compute emulator is up and running, right click on the icon and select ‘Show Compute Emulator UI’ The last steps! Make sure the ‘Neo4j.Azure.Server’ cloud project is set up as the start project and let’s hit F5 tension mounts, the build takes place (you need to accept the UAC warning) and VS does it’s stuff. If you look at the Compute Emulator UI you’ll see some log stuff (which you’ll need if this goes awry – but it won’t don’t worry!) In a bit, the console and a Java window will pop up: Then the console will bog off, leaving just the Java one, and if we switch back to the Compute Emulator UI and scroll up we should be able to see a line telling us the port number we’ve been assigned (in my case 7411): (If you can’t see it, don’t worry.. press CTRL+A on the emulator, then CTRL+C, copy all the text and paste it into something like Notepad, then just do a Find for ‘port’ you’ll soon see it) Go to your favourite browser, and head to: http://localhost:YOURPORT/ and you should see the WebAdmin! See you on the cloud side hopefully! Chris PS Other gotchas! OK, I’ve been caught out a couple of times: I had an instance of Neo4J running as a service on my machine, the Azure instance wanted to run the https version of the server on the same port as the Service was running on, and so Java would complain that the port was already in use.. The first time I converted the project, it didn’t update the version of the Azure library to load, in the App.Config of the Neo4jServerHost project, and VS would throw an exception saying it couldn’t find the Azure dll version 1.0.0.0.

  Read the article

• Why do we (really) program to interfaces?

  - by Kyle Burns

  One of the earliest lessons I was taught in Enterprise development was "always program against an interface".  This was back in the VB6 days and I quickly learned that no code would be allowed to move to the QA server unless my business objects and data access objects each are defined as an interface and have a matching implementation class.  Why?  "It's more reusable" was one answer.  "It doesn't tie you to a specific implementation" a slightly more knowing answer.  And let's not forget the discussion ending "it's a standard".  The problem with these responses was that senior people didn't really understand the reason we were doing the things we were doing and because of that, we were entirely unable to realize the intent behind the practice - we simply used interfaces and had a bunch of extra code to maintain to show for it. It wasn't until a few years later that I finally heard the term "Inversion of Control".  Simply put, "Inversion of Control" takes the creation of objects that used to be within the control (and therefore a responsibility of) of your component and moves it to some outside force.  For example, consider the following code which follows the old "always program against an interface" rule in the manner of many corporate development shops: 1: ICatalog catalog = new Catalog(); 2: Category[] categories = catalog.GetCategories(); In this example, I met the requirement of the rule by declaring the variable as ICatalog, but I didn't hit "it doesn't tie you to a specific implementation" because I explicitly created an instance of the concrete Catalog object.  If I want to test the functionality of the code I just wrote I have to have an environment in which Catalog can be created along with any of the resources upon which it depends (e.g. configuration files, database connections, etc) in order to test my functionality.  That's a lot of setup work and one of the things that I think ultimately discourages real buy-in of unit testing in many development shops. So how do I test my code without needing Catalog to work?  A very primitive approach I've seen is to change the line the instantiates catalog to read: 1: ICatalog catalog = new FakeCatalog();   once the test is run and passes, the code is switched back to the real thing.  This obviously poses a huge risk for introducing test code into production and in my opinion is worse than just keeping the dependency and its associated setup work.  Another popular approach is to make use of Factory methods which use an object whose "job" is to know how to obtain a valid instance of the object.  Using this approach, the code may look something like this: 1: ICatalog catalog = CatalogFactory.GetCatalog();   The code inside the factory is responsible for deciding "what kind" of catalog is needed.  This is a far better approach than the previous one, but it does make projects grow considerably because now in addition to the interface, the real implementation, and the fake implementation(s) for testing you have added a minimum of one factory (or at least a factory method) for each of your interfaces.  Once again, developers say "that's too complicated and has me writing a bunch of useless code" and quietly slip back into just creating a new Catalog and chalking any test failures up to "it will probably work on the server". This is where software intended specifically to facilitate Inversion of Control comes into play.  There are many libraries that take on the Inversion of Control responsibilities in .Net and most of them have many pros and cons.  From this point forward I'll discuss concepts from the standpoint of the Unity framework produced by Microsoft's Patterns and Practices team.  I'm primarily focusing on this library because it questions about it inspired this posting. At Unity's core and that of most any IoC framework is a catalog or registry of components.  This registry can be configured either through code or using the application's configuration file and in the most simple terms says "interface X maps to concrete implementation Y".  It can get much more complicated, but I want to keep things at the "what does it do" level instead of "how does it do it".  The object that exposes most of the Unity functionality is the UnityContainer.  This object exposes methods to configure the catalog as well as the Resolve<T> method which is used to obtain an instance of the type represented by T.  When using the Resolve<T> method, Unity does not necessarily have to just "new up" the requested object, but also can track dependencies of that object and ensure that the entire dependency chain is satisfied. There are three basic ways that I have seen Unity used within projects.  Those are through classes directly using the Unity container, classes requiring injection of dependencies, and classes making use of the Service Locator pattern. The first usage of Unity is when classes are aware of the Unity container and directly call its Resolve method whenever they need the services advertised by an interface.  The up side of this approach is that IoC is utilized, but the down side is that every class has to be aware that Unity is being used and tied directly to that implementation. Many developers don't like the idea of as close a tie to specific IoC implementation as is represented by using Unity within all of your classes and for the most part I agree that this isn't a good idea.  As an alternative, classes can be designed for Dependency Injection.  Dependency Injection is where a force outside the class itself manipulates the object to provide implementations of the interfaces that the class needs to interact with the outside world.  This is typically done either through constructor injection where the object has a constructor that accepts an instance of each interface it requires or through property setters accepting the service providers.  When using dependency, I lean toward the use of constructor injection because I view the constructor as being a much better way to "discover" what is required for the instance to be ready for use.  During resolution, Unity looks for an injection constructor and will attempt to resolve instances of each interface required by the constructor, throwing an exception of unable to meet the advertised needs of the class.  The up side of this approach is that the needs of the class are very clearly advertised and the class is unaware of which IoC container (if any) is being used.  The down side of this approach is that you're required to maintain the objects passed to the constructor as instance variables throughout the life of your object and that objects which coordinate with many external services require a lot of additional constructor arguments (this gets ugly and may indicate a need for refactoring). The final way that I've seen and used Unity is to make use of the ServiceLocator pattern, of which the Patterns and Practices team has also provided a Unity-compatible implementation.  When using the ServiceLocator, your class calls ServiceLocator.Retrieve in places where it would have called Resolve on the Unity container.  Like using Unity directly, it does tie you directly to the ServiceLocator implementation and makes your code aware that dependency injection is taking place, but it does have the up side of giving you the freedom to swap out the underlying IoC container if necessary.  I'm not hugely concerned with hiding IoC entirely from the class (I view this as a "nice to have"), so the single biggest problem that I see with the ServiceLocator approach is that it provides no way to proactively advertise needs in the way that constructor injection does, allowing more opportunity for difficult to track runtime errors. This blog entry has not been intended in any way to be a definitive work on IoC, but rather as something to spur thought about why we program to interfaces and some ways to reach the intended value of the practice instead of having it just complicate your code.  I hope that it helps somebody begin or continue a journey away from being a "Cargo Cult Programmer".

  Read the article

< Previous Page | 120 121 122 123 124 125 126 127 128 129 130 131  | Next Page >