Search Results

Search found 6945 results on 278 pages for 'azure use cases'.

Page 169/278 | < Previous Page | 165 166 167 168 169 170 171 172 173 174 175 176  | Next Page >

  • Unicode data from NSData to NSString

    - by Jeff
    So if I have NSData from an HTTP request, then I do something like this: NSString *test = [[NSString alloc] initWithData:data encoding:NSUTF8StringEncoding]; This will result in null if the data contains weird unicode data (title is from reddit): {"title":"click..¦¦me..and..then¦¦________ ¦¦check¦¦_.your...¦¦.__...¦¦____ ¦¦....¦¦¦¦¦¦¦¦¦¦¦¦¦¦....¦¦____ ¦¦¦¦¦¦....¦¦¦¦¦¦....¦¦¦¦¦¦____ ¦¦¦¦¦¦¦¦¦¦¦¦¦¦¦¦¦¦¦¦¦¦____ ....¦¦¦¦¦¦¦¦¦¦¦¦¦¦¦¦¦¦______ ........¦¦..._recently....¦¦________ ....¦¦....viewed....links....¦¦_____"}, How would I convert the data to a string? Ideally, it would best if the string wasn't null so I could parse it as JSON, but even a lossy conversion is fine with me in these cases. I'm not familiar with unicode (naive American I am), so any enlightenment about that would be a nice bonus :)

    Read the article

  • Is a class that is hard to unit test badly designed?

    - by Extrakun
    I am now doing unit testing on an application which was written over the year, before I started to do unit-testing diligently. I realized that the classes I wrote are hard to unit test, for the following reasons: Relies on loading data from database. Which means I have to setup a row in the table just to run the unit test (and I am not testing database capabilities). Requires a lot of other external classes just to get the class I am testing to its initial state. On the whole, there don't seem to be anything wrong with the design except that it is too tightly coupled (which by itself is a bad thing). I figure that if I have written automated test cases with each of the class, hence ensuring that I don't heap extra dependencies or coupling for the class to work, the class might be better designed. Does this reason holds water? What are your experiences?

    Read the article

  • ASP.NET MVC: run code after view has rendered (close db transaction)

    - by thermal7
    Hi, I am using ASP.NET MVC2 with NHibernate, but am facing an issue. All calls to the database via NHibernate should be inside a transaction, however code inside the view kicks off database calls in some instances. Thus there is a need to be able to commit the transaction after the view has rendered. For example displaying a list of users and their user roles you might show the user role using this code: <%: Model.UserRole.Name % This will cause a hit on the database as the UserRole is loaded using a NHibernate proxy. You can fetch the UserRole eagerly which circumvents the issue in this case, but there are cases where it is much faster to use lazy loading. Anyway, is there a way to run code after a view has rendered?

    Read the article

  • facebook graph api..get friends of friends

    - by Krishnamurthy
    I am using the FB graph api to get a list of my friends' friends using: https://graph.facebook.com/<id>/friends?access_token=<token> Now the weird thing is that this works for some of my friends and does not work for some others. Could somebody tell me why this is happening and a possible fix? In cases where it does not work, I get the following message: "Can't lookup all friends of YYYYYY. Can only lookup for the logged in user (XXXXX), or friends of the logged in user with the appropriate permission" And I repeat, I AM ABLE to get friend lists of some of my friends using the above URL. What permissions are being mentioned here?

    Read the article

  • How to force restart a Windows box using VBScript?

    - by tloach
    I'm trying to find a way to force Windows to reboot, and I am running into issues. I've tried Set OpSysSet = GetObject("winmgmts:{authenticationlevel=Pkt," _ & "(Shutdown)}").ExecQuery("select * from Win32_OperatingSystem where "_ & "Primary=true") for each OpSys in OpSysSet retVal = OpSys.Reboot() next I've also tried using the shutdown -f -r command, and in both cases I sometimes get no response, and if I try again I get an error saying "Action could not complete because the system is shutting down" even though no matter how long I leave it it doesn't shut down, it still allows me to start new programs, and doing a shutdown -a gives the same error. How can a script be used to force Windows to reboot?

    Read the article

  • Appending Integer array elements in Java

    - by Neville
    I have an array, say int a[]={2,0,1,0,1,1,0,2,1,1,1,0,1,0,1}; I need to append each of the 5 neighboring elements and assign them to a new array b with length=(a.length/5); and i want to append the 5 neighboring elements so that I have: int b[]={20101,10211,10101}; I need to do this for various length arrays, in most cases with length of a being greater than 15. Any help would be greatly appreciated, I'm programming in Java. Thanks in advance.

    Read the article

  • How do I use master page container in partial view

    - by user200295
    I have several partial views with Javascript that I am trying to move to the bottom of the page. To do this I am trying to use a container in the master page Master Page - <asp:ContentPlaceHolder ID="Foot" runat="server"></asp:ContentPlaceHolder> Partial view(ascx) <asp:Content ID="header" ContentPlaceHolderID="head" runat="server"> ... </asp:Content> But I get this error Parser Error Message: Content controls have to be top-level controls in a content page or a nested master page that references a master page. So how do I ensure that the Javascript for the partial view is at the bottom of the page? Especially in cases where the html layout needs to be at the top of the page?

    Read the article

  • How can I join on a CSV varchar?

    - by mgroves
    I have a varchar field that contains a string like "10,11,12,13". How can I use that CSV string to join to another table with those IDs? Here's the approach I'm taking now: select * from SomeTable a WHERE (',' + @csvString + ',') LIKE '%,' + CONVERT(varchar(25), a.ID) + ',%' Where @csvString is "10,11,12,...". I intend to use this method as a join condition as well. That method works, but it's rather slow (using CAST doesn't improve the speed). I understand that having CSVs in the database like that is usually a very silly idea in most cases, but there's nothing I can do about that.

    Read the article

  • How can I load a sql "dump" file into sql alchemy

    - by JudoWill
    I have a large sql dump file ... with multiple CREATE TABLE and INSERT INTO statements. Is there any way to load these all into a SQLAlchemy sqlite database at once. I plan to use the introspected ORM from sqlsoup after I've created the tables. However, when I use the engine.execute() method it complains: sqlite3.Warning: You can only execute one statement at a time. Is there a way to work around this issue. Perhaps splitting the file with a regexp or some kind of parser, but I don't know enough SQL to get all of the cases for the regexp. Any help would be greatly appreciated. Will EDIT: Since this seems important ... The dump file was created with a MySQL database and so it has quite a few commands/syntax that sqlite3 does not understand correctly.

    Read the article

  • latin bases language segmentation gramatical rules

    - by pravin
    Hi folks, I am working on one feature i.e. to apply language segmentation rules ( grammatical ) for Latin based language ( English currently ). Currently I am in phase of breaking sentences of user input. e.g.: "I am working in language translation". "I have used Google MT API for this" In above example i will break above sentence by full stop (.) This is normal cases where I am breaking sentence on dot, but there are n number of characters for breaking sentence like ( . ! ? etc ). I have following SRX rules for segmentation. Here my question are :- 1) Is there any reference ? which I can use for resolving my language segmentation rules. 2) Or Is there any forums on language segmentation ? , so that i can discuss efficiently Please let me know if anybody know about this ? Thanks a lot.

    Read the article

  • java string detection of ip

    - by user384706
    Hi, Assume a java string that contains an IP (v4 or v6) or a hostname. What is the best way to detect among these cases? I am not interested so much on whether the IP is in valid range (e.g. 999.999.999.999 for IPv4). I am interested in just a way to detect if a String is a hostname or an IP (v4 or v6). Initially I though this: if(theString.indexOf("@")!=-1){ //Not an Ip } but I am not sure if I should always expect a format containing @ always. Thanks

    Read the article

  • Evenly distribute items on the screen

    - by abolotnov
    I am trying to solve this little puzzle (the algorithm): I have N image icons and I want to distribute them evenly on users screen. Say, I put them in a table. If there is one image, there will be one cell in a table. If two - one row with two columns, if three - one row and three columns, if four - two rows, two columns... and so on until row space is gone and since then the table should only grow in columns without adding extra rows. I'm trying to figure an algorithm for this and perhaps this is something that has a solution already somewhere? My attempt is so far something like this: obtain_max_rows() obtain_visible_columns() if (number_of_pictures > max_rows*max_columns) { columns = roundup(number_of_pictures/max_rows) for(max_rows){generate row;for columns{generate column}} } else { **here comes to trouble...** } This logic is bit silly though - it somehow needs to think cases where there are 12 pictures on first screen and 2 on the other trying to balance it say 8/6 or somehow like that.

    Read the article

  • Best way to test a batch email script

    - by John
    I have a php mail script sitting on a LAMP vps server. The script grabs about 1000 emails and sends them each an html email. I tested the script with about half a dozen of my own test email accounts and things worked fine. But I am concerned something may go wrong when I actually use this script for 1000 emails. Some things I would like to test for are 1) Confirm all 1000 emails were sent and received 2) Test to make sure emails did not end up in people's spam folders 3) Detect any other general failures Does anyone have suggestions on how I can test for the above cases? I would like to read about your experiences building batch email scripts.

    Read the article

  • How to get the LAN IP of a client using Java?

    - by Henrik
    How can i get the LAN IP-address of a computer using Java? I want the IP-address which is connected to the router and the rest of the network. I've tried something like this: Socket s = new Socket("www.google.com", 80); String ip = s.getLocalAddress().getHostAddress(); s.close(); This seem to work on some cases, but sometimes it returns the loopback-address or something completely different. Also, it requires internet connection. Does anyone got a more accurate method of doing this?

    Read the article

  • Remote Development Workflow with Tomcat and Eclipse

    - by Smithers
    Currently, I have tomcat installed on the production server to serve my java webapps. I develop in eclipse at my personal workstation and then I use an ant script to build the project into a war file and deploy that on the server. This setup works well when I am on the same network as the server because deploying is almost instantaneous. However, now that I am working remotely uploading the war file to the server is slow and in most cases very redundant (there are about .5 GB of static media included in the war file). Is there a better way to update my webapp on tomcat from eclipse and if so what are the best options for implementing such a solution with minimal effort?

    Read the article

  • my first shader in WebGL

    - by Diego
    Hello, I am writing my first shader in WebGL. I was wondering if the GLSL language has any way to evaluate if an attribute or a uniform is null. According to the specs it does not support to do something like if (attributeX) { dothis(); } else{ dothat(): } And I think it would be a waste to write a bool attribute for each of these cases would be a waste. Another question: what happen during rendering when you don't pass along the uniforms or attribs to the shader? Thanks!

    Read the article

  • Is ActiveMQ unreliable?

    - by user122991
    Hello, We have been using ActiveMQ 5.2 in our distributed enterprise application for about 3 months. During that time, we have experienced debilitating failures at least twice weekly. In particular, we see: 1) Topic publisher has its connection arbitrarily closed and experiences EOF on attempt to publish. Note well that this issue is not a function of some timeout. It does not correlate reliably with any inactivity. 2) Queue listeners never receive message. Message simply sits on Queue. 2) is much rarer (hardly ever) than 1). In both cases, the failures are highly intermittent-- they cannot be reliably reproduced through any testing usage pattern. Also, there are no errors or warning in the AMQ logs. Have others experienced similar problems? Is there an opinion that some other JMS provider is more reliable? thanks, Joe

    Read the article

  • C++ Testing framework integrated with Eclipse

    - by Mike
    I'm writing a C++ unit testing framework and I would like it if it could be integrated with Eclipse CDT. In other testing suites that work with Eclipse, JUnit for example, the user is provided a graphical list of all test cases and their results. Something like this would be the ideal. I'm just getting into this, so I need some advice on getting started. There are two approaches I see Use an existing Eclipse testing plugin (such as JUnit) and make the framework return output in the same format as the plugin's input. Write a plugin from scratch that can work with my framework (seems like it would take a lot of time) Thoughts appreciated

    Read the article

  • How Do I See The Final Text Of A Query Resulting From A Call To mysqli->prepare?

    - by Joshua
    After code like this: $stmt = $mysqli->prepare("SELECT District FROM City WHERE Name=?")) { $stmt->bind_param("s", $city); $stmt->execute(); $stmt->bind_result($district); $stmt->fetch(); printf("%s is in district %s\n", $city, $district); How Do I See The Actual SQL Statement That Was Executed? (It Should Look Something Like "SELECT District FROM City WHERE Name='Simi Valley';") I already realize that in this simplistic case it would be very easy to simply reconstruct the query... but how can I access it in a general way that will work for very complicated prepared statements, and cases where I don't necessarily already understand the intended structure of the query, etc. Isn't there some function or method that can be called on the statement object that will return the actual text of the SQL query, after binding?

    Read the article

  • Int PK inner join Vs Guid PK inner Join on SQL Server. Execution plan.

    - by bigb
    I just did some testing for Int PK join Vs Guid PK. Tables structure and number of records looking like that: Performance of CRUD operations using EF4 are pretty similar in both cases. As we know Int PK has better performance rather than strings. So SQL server execution plan with INNER JOINS are pretty different Here is an execution plan. As i understand according with execution plan from attached image Int join has better performance because it is taking less resources for Clustered index scan and it is go in two ways, am i right? May be some one may explain this execution plan in more details?

    Read the article

  • Need help from CSS purist!

    - by Darcy
    I'm beginning to use CSS more and more but I'm not sure if I'm using it as intended in all cases. I have classes named things like 'box' that will wrap the content inside it with a border, which seems like good CSS to me. On the other hand I have classes like 'margin-right-5' and 'float-left' that sets the css to margin-right: 5px and float:left respectively. I'm wondering if this is good practice. Then in my markup I'd do something like: <div class="box float-left margin-right-5"> <!-- CONTENT HERE --> </div> Sometimes I may want to float right or not have a margin at all and still use the 'box' class, so I use several classes in order to make the css more flexible. Am I destroying the principles of CSS?

    Read the article

  • Linux - How do i know the block map of the given file and/or the free space map of the partition.

    - by Inso Reiges
    Hello, I am on Linux and need to know either of the two things: 1) If i have a regular file on some file system on a partition under Linux is there a way to know the set of the physical blocks that this file occupies on the drive from user space? Or at least the set of the file system's clusters? 2) Is there a way to get the same information about the whole free space of the given file system? In both cases i understand that if there is any possible way to extract this info it will probably be totally unsafe and racy (anything could happen to these set of blocks between the time i see them and act on them somehow). I also really don't want an implementation that will have to know a lot about every filesystem.

    Read the article

  • Difference in techniques for setting a stubbed method's return value with Rhino Mocks

    - by CRice
    What is the main difference between these following two ways to give a method some fake implementation? I was using the second way fine in one test but in another test the behaviour can not be achieved unless I go with the first way. These are set up via: IMembershipService service = test.Stub<IMembershipService>(); so (the first), using (test.Record()) //test is MockRepository instance { service.GetUser("dummyName"); LastCall.Return(new LoginUser()); } vs (the second). service.Stub(r => r.GetUser("dummyName")).Return(new LoginUser()); Edit The problem is that the second technique returns null in the test, when I expect it to return a new LoginUser. The first technique behaves as expected by returning a new LoginUser. All other test code used in both cases is identical.

    Read the article

  • Java: is Exception class thread-safe?

    - by Vilius Normantas
    As I understand, Java's Exception class is certainly not immutable (methods like initCause and setStackTrace give some clues about that). So is it at least thread-safe? Suppose one of my classes has a field like this: private final Exception myException; Can I safely expose this field to multiple threads? I'm not willing to discuss concrete cases where and why this situation could occur. My question is more about the principle: can I tell that a class which exposes field of Exception type is thread-safe? Another example: class CustomException extends Exception { ... } Is this class thread-safe?

    Read the article

  • Strategy for developing namespaced and non-namespaced versions of same PHP code

    - by porneL
    I'm maintaining library written for PHP 5.2 and I'd like to create PHP 5.3-namespaced version of it. However, I'd also keep non-namespaced version up to date until PHP 5.3 becomes so old, that even Debian stable ships it ;) I've got rather clean code, about 80 classes following Project_Directory_Filename naming scheme (I'd change them to \Project\Directory\Filename of course) and only few functions and constants (also prefixed with project name). Question is: what's the best way to develop namespaced and non-namespaced versions in parallel? Should I just create fork in repository and keep merging changes between branches? Are there cases where backslash-sprinkled code becomes hard to merge? Should I write script that converts 5.2 version to 5.3 or vice-versa? Should I use PHP tokenizer? sed? C preprocessor? Is there a better way to use namespaces where available and keep backwards compatibility with older PHP?

    Read the article

< Previous Page | 165 166 167 168 169 170 171 172 173 174 175 176  | Next Page >