Search Results

Search found 66233 results on 2650 pages for 'class method'.

Page 341/2650 | < Previous Page | 337 338 339 340 341 342 343 344 345 346 347 348  | Next Page >

  • What is the time taken by java to call a method in another package?

    - by satish
    I have an assignment where i need to do feasibility study on two of my approaches and find optimized one. There are two packages A and B User input is gathered in A and then sent to B for execution. Now my approaches are 1. Call B methods from package A one by one. 2. Create a common method in B and send all the input as parameters. Which is feasible and best one, I know 2 is good in terms of code optimization and less number of calls. But, I want to understand what will be the execution time difference in above approaches How much time does java take to call method in another package? Is there any default value like x nano seconds or y milli seconds. Based on the time taken I can choose the appropiate one. Thanks

    Read the article

  • Any chances to imitate times() Ruby method in C#?

    - by Alexander Prokofyev
    Every time I need to do something N times inside an algorithm using C# I write this code for (int i = 0; i < N; i++) { ... } Studying Ruby I have learned about method times() which can be used with the same semantics like this N.times do ... end Code fragment in C# looks more complex and we should declare useless variable i. I tried to write extension method which returns IEnumerable, but I am not satisfied with the result because again I have to declare a cycle variable i. public static class IntExtender { public static IEnumerable Times(this int times) { for (int i = 0; i < times; i++) yield return true; } } ... foreach (var i in 5.Times()) { ... } Is it possible using some new C# 3.0 language features to make N times cycle more elegant?

    Read the article

  • android dev: how to implement for add method in menu interface?

    - by wsgjj
    who knows about the method "add" in menu interface? how to implement exactly? which is it's imlementation class? i couldn't find detail implementation in source code of sdk. i want to check how to implement in android source code for "add" method. e.g. public boolean onCreateOptionsMenu(Menu menu) { menu.***add***(0, MENU_NEW_GAME, 0, "New Game"); menu.add(0, MENU_QUIT, 0, "Quit"); return true; } thanks in advance!

    Read the article

  • very simple delegate musing

    - by Ted
    Sometimes the simplest questions make me love C/C++ and C# more and more. Today sitting on the bus musing aout delegates I remembered reading somwhere you don't need to use the new keyword when instaniating a new delegate. For example: public static void SomeMethod(string message) { ... } ... public delegate void TestDelgate(string message); //Define a delegate ........... //create a new instance ..METHOD 1 TestDelgate t = new TestDelgate(SomeMethod); //OR another way to create a new instance ..METHOD 2 TestDelgate t = SomeMethod; //create a new instance ..METHOD 2 So todays questions are What happens under the hood in method 2. Does the compiler expand method 2 into method 1, hence writing TestDelgate t = SomeMethod; is just a shortcut for TestDelgate t = new TestDelgate(SomeMethod);, or is there another reason for the exsitence of method 2 Do you guys think method 1 or method 2 is better for readability (this is a subjective question, but I'd just like to get a unscientific feel of general opinion of stackoverflow :-))

    Read the article

  • How can I find out how much memory an object (rather the instance of an object) of a C++ class consu

    - by Shadow
    Hi, I am developing a Graph-class, based on boost-graph-library. A Graph-object contains a boost-graph, so to say an adjacency_list, and a map. When monitoring the total memory usage of my program, it consumes quite a lot (checked with pmap). Now, I would like to know, how much of the memory is exactly consumed by a filled object of this Graph-class? With filled I mean when the adjacency_list is full of vertices and edges. I found out, that using sizeof() doesn't bring me far. Using valgrind is also not an alternative as there is quite some memory allocation done previously and this makes the usage of valgrind impractical for this purpose. I'm also not interested in what other parts of the program cost in memory, I want to focus on one single object. Thank you.

    Read the article

  • Visual Studio IDE - how do you quickly find the implementation(s) of an interface's method?

    - by Jess
    Is there a quick way to find all of the implementations of, not references to, an interface's method/property/etc? Here's some sample code: public class SomeClass : IBaseClass { public Int32 GetInt() { return 1; } } public interface IBaseClass { public Int32 GetInt(); } public class SomeOtherClass { ISomeClass _someClass; private TestMethod() { _someClass = new SomeClass(); _someClass.GetInt(); } } I want to quickly get to SomeClass.GetInt() while reviewing SomeOtherClass.TestMethod(). If I right click on _someClass.GetInt() and click 'Go To Definition', it takes me to the interface. If I click 'Find All References', I could potentially see a list of all uses ... not just the classes that implement the GetInt() method. Is there a faster way to find this? Any tips from other developers? We are using D.I. for most of our dependencies, which means that tracing through deeply nested code takes forever.

    Read the article

  • Does JavaScript have an equivalent to Perl's DESTROY method?

    - by Eric Strom
    Is there any method that is called or event that is dispatched right before an Element is cleaned up by the JavaScript garbage collector? In Perl I would write: package MyObj; sub new {bless {}} sub DESTROY {print "cleaning up @_\n"} and then later: { my $obj = MyObj->new; # do something with obj } # scope ends, and assuming there are no external references to $obj, # the DESTROY method will be called before the object's memory is freed My target platform is Firefox (and I have no need to support other browsers), so if there is only a Firefox specific way of doing this, that is fine. And a little background: I am writing the Perl module XUL::Gui which serves as a bridge between Perl and Firefox, and I am currently working on plugging a few memory leaks related to DOM Elements sticking around forever, even after they are gone and no more references remain on the Perl side. So I am looking for ways to either figure out when JavaScript Elements will be destroyed, or a way to force JavaScript to cleanup an object.

    Read the article

  • Is there a way of providing a final transform method when chaining operations (like map reduce) in underscore.js?

    - by latentflip
    (Really strugging to title this question, so if anyone has suggestions feel free.) Say I wanted to do an operation like: take an array [1,2,3] multiply each element by 2 (map): [2,4,6] add the elements together (reduce): 12 multiply the result by 10: 120 I can do this pretty cleanly in underscore using chaining, like so: arr = [1,2,3] map = (el) -> 2*el reduce = (s,n) -> s+n out = (r) -> 10*r reduced = _.chain(arr).map(map).reduce(reduce).value() result = out(reduced) However, it would be even nicer if I could chain the 'out' method too, like this: result = _.chain(arr).map(map).reduce(reduce).out(out).value() Now this would be a fairly simple addition to a library like underscore. But my questions are: Does this 'out' method have a name in functional programming? Does this already exist in underscore (tap comes close, but not quite).

    Read the article

  • What if I made an explicit reference to 'this' for use inside an inner class?

    - by badp
    So far, I've used this approach to access this from the scope of an inner class: class FooManagementWindow extends JFrame { JButton rejectFoo; //... void getFooAcceptingPanel(){ //... final FooManagementWindow referenceToThis = this; rejectFoo = new JButton("Reject"); rejectFoo.addEventListener(new EventListener() { @Override public void actionPerformed(ActionEvent arg) { referenceToThis.setEnabled(false); //this requires a network call //... referenceToThis.setEnabled(true); //the user may resume his work } }); //... } } However, I just learned that instead of declaring referenceToThis, a direct reference is kept for me as: FooManagementWindow.this I have no reason to think my less standard approach may lead to errors or weird corner cases. Or are there?

    Read the article

  • How to change the view if I am not present in the viewController class?

    - by wolverine
    I am doing an app with about 6-7 views and their corresponding controllers. I am in a view with a toolbar and button. When I click this button a popup view comes and my control is being moved to the popup view's Controller. But my view is still the same. From this popupview Controller's class, I want to change my original view. But since I am not in my original viewControllers class, I am not able to change it. How can I overcome this? Can anyone plz help me?

    Read the article

  • SEO/PHP: How to Convert Form-Submit URL (Get-Method) without Javascript SEO-Friendly?

    - by elmas
    hello, i have this code <form action="index.php" method="get" class="search-form"><input type="text" size="35" name="search" class="searchBox" value="" /><input type="submit" value="Start Searching!" /></form> and actually i convert the url with javascript <script type="text/javascript"> $(document).ready(function() { $('.search-form').submit(function() { var value = $('.search-form input:text').val(); value = value = value.replace(/\W/,''); // replace window.location.href = value + "-keyword" + ".html"; return false; }); }); </script> is there a method to convert the url seo-friendly without javascript? maybe with php?

    Read the article

  • How do I dynamically point an interface to a specific class at runtime?

    - by Jon
    As a simple example, I have an xml file with a list of names of classes which actually carry out the work and all implement interface IDoWork with a method Process(). I loop through the items in the xml file. How do I actually dynamically assign the class to the interface from a string name? e.g. var IDoWork = new "DoWorkType1"(); IDoWork.Process(); <work> <item id="DoWorkType1"> </item> <item id="DoWorkType2"> </item> </work> I want to achieve a plugin type architecture, except the plugin isn't at an assembly level only a class level within my program.

    Read the article

  • What method of UIView gets called when instantiated from a NIB?

    - by retailevolved
    I have a simple custom view that is connected via outlet to a NIB. For this particular view, there are actions that I would like to perform on the view when it is initialized, no matter what NIB it is on. Trouble is, neither the (id)init or the (id)initWithFrame:(CGRect)frame methods are getting called on the custom view. Which method gets called on a UIView when it is instantiated from a NIB? I would just use the view controller and viewDidLoad method except that this particular view appears on a lot of different NIBs.

    Read the article

  • Using Where method in Linq 2 Entities with OR clause.

    - by Dani
    I want to use Where method in Linq 2 entities that will be equal to this userRepository.Users.Where(u=>u.RoleID == 1 || u=>u.RoldID == 2).Select(o => new SelectListItem { Text = o.Role.RoleName, Value = o.RoleID.ToString() }).ToList(); The problem of course is in Where(u=u.RoleID == 1 || u=u.RoldID == 2) The problem is that I don't know how to use WHERE method with OR inside the WHERE clause. any ideas (the code above will not compile of-course b/c of the lambda expression. userRepository.Users returns an list of Users entities. I guess that and can be done using concatenation of .Where().Where() but I need an OR.

    Read the article

  • CSS: Centering a floated block level element in IE6 (It almost works)

    - by Louis W
    I have a block level element which I am centering on the page. I have gotten it to work for all other browsers except IE6 where it ALMOST works. http://tinyurl.com/28sh9eq If I view the page in IE6 the red box is slightly off center of the pink one in IE. If I then resize the browser window it snaps into place where I want it. Uhhhhh.... yea.... what gives? How come resizing the window makes it work? I have also tried setting an explicit width on the wrapper with no avail. <?xml version="1.0" encoding="UTF-8"?> <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"> <html> <head> <title></title> <meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /> <meta http-equiv="X-UA-Compatible" content="IE=7" /> <style type="text/css"> BODY { text-align: center; font-family: Arial; } .row_wrap { height: 100px; margin-bottom: 30px; background-color: pink; } .row { float: right; position: relative; left: -50%; text-align: left; clear: both; } .button1 { color: #FFF; height: 36px; text-decoration: none; position: relative; padding: 0 30px; background: url('button.gif') no-repeat 0 0; display: block; float: left; left: 50%; } .button1 .end { width: 20px; height: 37px; position: absolute; right: -2px; top: 0; background: url('button.gif') no-repeat right 0; } .button1 .text { font-size: 16px; font-weight: bold; white-space: nowrap; height: 36px; padding-top: 7px; display: block; float: left; } .button1 .text .arrow { vertical-align: 1px; } </style> </head> <body> <h2>RTL: Button 1</h2> <div class="row_wrap"> <div class="row" dir="rtl"> <a href="#" class="button1"> <span class="end"></span> <span class="text"><span class="arrow">»</span> Hello 1.</span> </a> </div> </div> <h2>RTL: Button 1-2</h2> <div class="row_wrap" style="width: 400px;"> <div class="row" dir="rtl"> <a href="#" class="button1"> <span class="end"></span> <span class="text"><span class="arrow">»</span> Hello 1.</span> </a> </div> </div> <br/><br/> <h2>Normal: Button 1</h2> <div class="row_wrap"> <div class="row"> <a href="#" class="button1"> <span class="end"></span> <span class="text"><span class="arrow">»</span> Hello.</span> </a> </div> </div> </body> Thanks for your help.

    Read the article

  • Will an optimizing compiler remove calls to a method whose result will be multiplied by zero?

    - by Tim R.
    Suppose you have a computationally expensive method, Compute(p), which returns some float, and another method, Falloff(p), which returns another float from zero to one. If you compute Falloff(p) * Compute(p), will Compute(p) still run when Falloff(p) returns zero? Or would you need to write a special case to prevent Compute(p) from running unnecessarily? Theoretically, an optimizing compiler could determine that omitting Compute when Falloff returns zero would have no effect on the program. However, this is kind of hard to test, since if you have Compute output some debug data to determine whether it is running, the compiler would know not to omit it because of that debug info, resulting in sort of a Schrodinger's cat situation. I know the safe solution to this problem is just to add the special case, but I'm just curious.

    Read the article

  • Which method of adding items to the ASP.NET Dictionary class is more efficient?

    - by ahmd0
    I'm converting a comma separated list of strings into a dictionary using C# in ASP.NET (by omitting any duplicates): string str = "1,2, 4, 2, 4, item 3,item2, item 3"; //Just a random string for the sake of this example and I was wondering which method is more efficient? 1 - Using try/catch block: Dictionary<string, string> dic = new Dictionary<string, string>(); string[] strs = str.Split(','); foreach (string s in strs) { if (!string.IsNullOrWhiteSpace(s)) { try { string s2 = s.Trim(); dic.Add(s2, s2); } catch { } } } 2 - Or using ContainsKey() method: string[] strs = str.Split(','); foreach (string s in strs) { if (!string.IsNullOrWhiteSpace(s)) { string s2 = s.Trim(); if (!dic.ContainsKey(s2)) dic.Add(s2, s2); } }

    Read the article

  • Where should the line between property and method be?

    - by Catskul
    For many situations it is obvious whether something should be a property or a method however there are items that might be considered ambiguous. Obvious Properties: "name" "length" Obvious Methods: "SendMessage" "Print" Ambiguous: "Valid" / "IsValid" / "Validate" "InBounds" / "IsInBounds" / "CheckBounds" "AverageChildValue" / "CalcAverageChildValue" "ColorSaturation" / "SetColorSaturation" I suppose I would lean towards methods for the ambiguous, but does anyone know of a rule or convention that helps decide this? E.g. should all properties be O(1)? Should a property not be able to change other data (ColorSaturation might change R,G,B values)? Should it not be a property if there is calculation or aggregation? Just from an academic perspective, (and not because I think it's a good idea) is there a reason not to go crazy with properties and just make everything that is an interrogation of the class without taking an argument, and everything that can be changed about the class with a single argument and cant fail, a property?

    Read the article

  • How to map inheritance with property returned other inheritance?

    - by dario-g
    Hi I have abstract class Vehicle and two classes that derive from: Car and ForkLift. public abstract class Vehicle { public EngineBase Engine { get; set; } } public class Car : Vehicle { public GasEngine Engine { get; set; } } public class ForkLift : Vehicle { public ElectricEngine Engine { get; set; } } and Engine clasess: public abstract class Engine { } public class GasEngine : Engine { } public class ElectricEngine : Engine { } Engines are mapped with "table per class hierarchy". With Vehicles I want to use the same pattern. How to map Engine class and derived with that Engine property?

    Read the article

  • Java Interfaces Methodology: Should every class implement an interface?

    - by Amir Rachum
    I've been programming in Java for a few courses in the University and I have the following question: Is it methodologically accepted that every class should implement an interface? Is it considered bad practice not to do so? Can you describe a situation where it's not a good idea to use interfaces? Edit: Personally, I like the notion of using Interfaces for everything as a methodology and habit, even if it's not clearly beneficial. Eclipse automatically created a class file with all the methods, so it doesn't waste any time anyway.

    Read the article

  • WCF: What happens if a channel is established but no method is called?

    - by mafutrct
    In my specific case: A WCF connection is established, but the only method with "IsInitiating=true" (the login method) is never called. What happens? In case the connection is closed due to inactivity after some time: Which setting configures this timeout? Is there still a way for a client to keep the connection alive? Reason for this question: I'm considering the above case as a possible security hole. Imagine many clients connecting to a server without logging in thus preventing other clients from connecting due to bandwidth problems or port shortage or lack of processing power or ... Am I dreaming, or is this an actual issue?

    Read the article

  • Why this friend function can't access a private member of the class?

    - by Alceu Costa
    I am getting the following error when I try to access bins private member of the GHistogram class from within the extractHistogram() implementation: error: 'QVector<double> MyNamespace::GHistogram::bins' is private error: within this context Where the 'within this context' error points to the extractHistogram() implementation. Does anyone knows what's wrong with my friend function declaration? Here's the code: namespace MyNamespace{ class GHistogram { public: GHistogram(qint32 numberOfBins); qint32 getNumberOfBins(); /** * Returns the frequency of the value i. */ double getValueAt(qint32 i); friend GHistogram * Gbdi::extractHistogram(GImage *image, qint32 numberOfBins); private: QVector<double> bins; }; GHistogram * extractHistogram(GImage * image, qint32 numberOfBins); }

    Read the article

< Previous Page | 337 338 339 340 341 342 343 344 345 346 347 348  | Next Page >