Search Results

Search found 3987 results on 160 pages for 'captain obvious'.

Page 120/160 | < Previous Page | 116 117 118 119 120 121 122 123 124 125 126 127  | Next Page >

  • Converting non-delimited text into name/value pairs in Delphi

    - by robsoft
    I've got a text file that arrives at my application as many lines of the following form: <row amount="192.00" store="10" transaction_date="2009-10-22T12:08:49.640" comp_name="blah " comp_ref="C65551253E7A4589A54D7CCD468D8AFA" name="Accrington "/> and I'd like to turn this 'row' into a series of name/value pairs in a given TStringList (there could be dozens of these <row>s in the file, so eventually I will want to iterate through the file breaking each row into name/value pairs in turn). The problem I've got is that the data isn't obviously delimited (technically, I suppose it's space delimited). Now if it wasn't for the fact that some of the values contain leading or trailing spaces, I could probably make a few reasonable assumptions and code something to break a row up based on spaces. But as the values themselves may or may not contain spaces, I don't see an obvious way to do this. Delphi' TStringList.CommaText doesn't help, and I've tried playing around with Delimiter but I get caught-out by the spaces inside the values each time. Does anyone have a clever Delphi technique for turning the sample above into something resembling this? ; amount="192.00" store="10" transaction_date="2009-10-22T12:08:49.640" comp_name="blah " comp_ref="C65551253E7A4589A54D7CCD468D8AFA" name="Accrington " Unfortunately, as is usually the case with this kind of thing, I don't have any control over the format of the data to begin with - I can't go back and 'make' it comma delimited at source, for instance. Although I guess I could probably write some code to turn it into comma delimited - would rather find a nice way to work with what I have though. This would be in Delphi 2007, if it makes any difference.

    Read the article

  • TeamCity + HG. Only pull (push?) passing builds

    - by ColoradoMatt
    Feels like with the popularity of continuous integration this one should be a piece of cake but I am stumped. I am setting up TeamCity with HG. I want to be able to push changesets up to a repository that TeamCity watches and runs builds on changes. That's easy. Next, if a build passes, I want that changeset to be pulled into a "clean" repository... one that contains only passing changesets. Should be easy but... TeamCity 6 supports multiple build steps and if any step fails, the rest don't run. My thought was to put a build step at the end that does a pull (or optionally a push?) to get the passing changeset into the clean repository. I am trying to use PsExec to run hg on the box with the repositories. If I try to run just a plain 'hg pull' it can't find the hg.exe even though it is set in the path and I have used the -w flag. I have tried putting a .bat file in the clean repository that takes a revision parameter and it works fine... locally. When I try to run the .bat file remotely (using PsExec) it runs everything fine but it tries to run it on the build agent. Even if I set the -w argument it runs the .bat file there but tries to run the contents on the build agent box. Am I just WAY off in my approach? Seems like this is a pretty obvious thing to do so either my Google skills are waning or no one thinks this is worthy of writing about. Either way, I am stuck in SVN land trying to get out so I would appreciate some help!

    Read the article

  • What do these errors mean? ISOC++ forbids assignment of arrays...

    - by xunlinkx
    I'm trying to compile some code on one of our systems for our DBA...I've edited the makefiles to include the pertinent libraries listed in the documentation, but I keep getting these errors... Can you discern any obvious problems from my command lines in reference to the errors listed? Thank you! make -f /u01/app/banner/ban8/TEST3/links/Makefile_tm_linux64_redhat5_ban8.mk gcc -m64 -D_NOFIXARGPTR -fpic -shared -DTMCILIB_EXPORTS -D_TMUNICODE -I/usr/local/ban_icu -I/usr/local/src/icu/source/i18n/ -I/usr/local/src/icu/source/common/ -I/usr/local/src/icu/source/extra/ustdio/ -I/usr/local/src/icu/source/io -L/usr/lib64 -L/usr/lib -L/usr/local/src/icu/source/data/ -L/usr/local/src/icu/source/data/out/ -L/usr/local/src/icu/source/tools/toolutil/ -L/usr/lib/im/icuconv/ -L/usr/local/lib/ -L. -licui18n -licudata -licuuc -licu-toolutil -licuio msgfmttm.cpp umsgtm.cpp tmcilib.cpp -o /u01/app/banner/ban8/TEST3/general/exe/libtmciuc.so umsgtm.cpp: In function ‘void fixArgPtr(const UChar*, __va_list_tag (*)[1])’: umsgtm.cpp:158: error: array must be initialized with a brace-enclosed initializer umsgtm.cpp:194: error: ISO C++ forbids assignment of arrays umsgtm.cpp: In function ‘int32_t tmumsg_vformat(void*, UChar, int32_t, __va_list_tag*, UErrorCode*)’: umsgtm.cpp:305: error: cannot convert ‘__va_list_tag**’ to ‘__va_list_tag ()[1]’ for argument ‘2’ to ‘void fixArgPtr(const UChar, __va_list_tag (*)[1])’ tmcilib.cpp: In function ‘int tmprintf(TMBundle*, const UChar*, ...)’: tmcilib.cpp:743: error: array must be initialized with a brace-enclosed initializer tmcilib.cpp: In function ‘int tmfprintf(TMBundle*, UFILE*, const UChar*, ...)’: tmcilib.cpp:757: error: array must be initialized with a brace-enclosed initializer tmcilib.cpp: In function ‘int tmsprintf(TMBundle*, UChar*, const UChar*, ...)’: tmcilib.cpp:808: error: array must be initialized with a brace-enclosed initializer

    Read the article

  • Deploying ASP.NET MVC to IIS6: pages are just blank

    - by BryanGrimes
    I have an MVC app that is actually on a couple other servers but I didn't do the deploy. For this deploy I have added the wildcard to aspnet_isapi.dll which has gotten rid of the 404 error. But the pages are not pulling up, rather everything is just blank. I can't seem to find any IIS configuration differences. The Global asax.cs file does have routing defined, but as I've seen on a working server, that file isn't just hanging out in the root or anything so obvious. What could I be missing here? All of the servers are running IIS6 and I have compared the setups and they look the same to me at this point. Thanks... Bryan EDIT for the comments thus far: I've looked in the event logs with no luck, and scoured various IIS logs per David Wang: blogs.msdn.com. Below is the Global.asax.cs file... public class MvcApplication : System.Web.HttpApplication { public static void RegisterRoutes(RouteCollection routes) { routes.IgnoreRoute("{resource}.axd/{*pathInfo}"); routes.IgnoreRoute("error.axd"); // for Elmah // For deployment to IIS6 routes.Add(new Route ( "{controller}.mvc/{action}/{id}", new RouteValueDictionary(new { action = "Index", id = (string)null }), new MvcRouteHandler() )); routes.MapRoute( "WeeklyTimeSave", "Time/Save", new { controller = "Time", action = "Save" } ); routes.MapRoute( "WeeklyTimeAdd", "Time/Add", new { controller = "Time", action = "Add" } ); routes.MapRoute( "WeeklyTimeEdit", "Time/Edit/{id}", new { controller = "Time", action = "Edit", id = "" } ); routes.MapRoute( "FromSalesforce", "Home/{id}", new { controller = "Home", action = "Index", id = "" }); routes.MapRoute( "Default2", "{controller}/{id}", new { controller = "Home", action = "Index", id = "" } ); routes.MapRoute( "Default", "{controller}/{action}/{id}", new { controller = "Home", action = "Index", id = "" } ); } protected void Application_Start() { RegisterRoutes(RouteTable.Routes); } } Maybe this is as stupid as the asax file not being somewhere it needs to be, but heck if I know at this point.

    Read the article

  • Why does the OnDeserialization not fire for XML Deserialization?

    - by Jonathan
    I have a problem which I have been bashing my head against for the better part of three hours. I am almost certain that I've missed something blindingly obvious... I have a simple XML file: <?xml version="1.0" encoding="utf-8"?> <WeightStore xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"> <Records> <Record actual="150" date="2010-05-01T00:00:00" /> <Record actual="155" date="2010-05-02T00:00:00" /> </Records> </WeightStore> I have a simple class structure: [Serializable] public class Record { [XmlAttribute("actual")] public double weight { get; set; } [XmlAttribute("date")] public DateTime date { get; set; } [XmlIgnore] public double trend { get; set; } } [Serializable] [XmlRoot("WeightStore")] public class SimpleWeightStore { [XmlArrayAttribute("Records")] private List<Record> records = new List<Record>(); public List<Record> Records { get { return records; } } [OnDeserialized()] public void OnDeserialized_Method(StreamingContext context) { // This code never gets called Console.WriteLine("OnDeserialized"); } } I am using these in both calling code and in the class files: using System.Xml.Serialization; using System.Runtime.Serialization; I have some calling code: SimpleWeightStore weight_store_reload = new SimpleWeightStore(); TextReader reader = new StringReader(xml); XmlSerializer deserializer = new XmlSerializer(weight_store.GetType()); weight_store_reload = (SimpleWeightStore)deserializer.Deserialize(reader); The problem is that I am expecting OnDeserialized_Method to get called, and it isn't. I suspect it might have something to do with the fact that it's XML deserialization rather than Runtime deserialization, and perhaps I am using the wrong attribute name, but I can't find out what it might be. Any ideas, folks?

    Read the article

  • In asp.Net, writing code in the control tag generates compile error

    - by Nour Sabouny
    Hi this is really strange !! But look at the following asp code: <div runat="server" id="MainDiv"> <%foreach (string str in new string[]{"First#", "Second#"}) { %> <div id="<%=str.Replace("#","div") %>"> </div> <%} %> </div> now if you put this code inside any web page (and don't worry about the moral of this code, I made it just to show the idea) you'll get this error : Compiler Error Message: CS1518: Expected class, delegate, enum, interface, or struct Of course the error has nothing to do with the real problem, I searched for the code that was generated by asp.net and figured out the following : private void @__RenderMainDiv(System.Web.UI.HtmlTextWriter @__w, System.Web.UI.Control parameterContainer) { @__w.Write("\r\n "); #line 20 "blabla\blabla\Default.aspx" foreach (string str in new string[] { "First#", "Second#" }) { #line default #line hidden @__w.Write("\r\n <div id=\""); #line 22 "blabla\blabla\Default.aspx" @__w.Write(str.Replace("#", "div")); #line default #line hidden @__w.Write("\">\r\n "); } This is the code that was generated from the asp page and this is the method that is meant to render our div (MainDiv), I found out that there is a missing bracket "}" that closes the method or the (for loop). now the problem has three parts: 1- first you should have a server control (in our situation is the MainDiv) and I'm not sure if it is only the div tag. 2- HTML control inside the server control and a code inside it using the double quotation mark ( for example <div id="<%=str instead of <div id='<%=str. 3-Any keyword which has block brackets e.g.:for{},while{},using{}...etc. now removing any part, will solve the problem !!! how is this happening ?? any ideas ? BTW: please help me to make the question more obvious, because I couldn't find the best words to describe the problem.

    Read the article

  • Debug NAudio MP3 reading difference?

    - by Conrad Albrecht
    My code using NAudio to read one particular MP3 gets different results than several other commercial apps. Specifically: My NAudio-based code finds ~1.4 sec of silence at the beginning of this MP3 before "audible audio" (a drum pickup) starts, whereas other apps (Windows Media Player, RealPlayer, WavePad) show ~2.5 sec of silence before that same drum pickup. The particular MP3 is "Like A Rolling Stone" downloaded from Amazon.com. Tested several other MP3s and none show any similar difference between my code and other apps. Most MP3s don't start with such a long silence so I suspect that's the source of the difference. Debugging problems: I can't actually find a way to even prove that the other apps are right and NAudio/me is wrong, i.e. to compare block-by-block my code's results to a "known good reference implementation"; therefore I can't even precisely define the "error" I need to debug. Since my code reads thousands of samples during those 1.4 sec with no obvious errors, I can't think how to narrow down where/when in the input stream to look for a bug. The heart of the NAudio code is a P/Invoke call to acmStreamConvert(), which is a Windows "black box" call which I can't think how to error-check. Can anyone think of any tricks/techniques to debug this?

    Read the article

  • Usability for content editors: Drupal or PHP framework?

    - by Jim
    Greetings: I am going to develop a basic Web site that supports some custom content types, is multilingual, and has content moderation workflow for a restricted group of content editors. This seems like an obvious choice for Drupal, except... the content editors will have little computer experience. In my opinion, that is a show-stopper for Drupal. For example, placing arbitrary inline images in content is a task that WordPress does well. I find Drupal's alternatives (IMCE, img_assist, etc.) clunky and not well integrated, and that will be a problem for this group of content editors. Also, none of the Drupal content workflow modules I tried seemed well-integrated; they all had a "tacked on" feel to them. As an admin I can understand why an "Accessible content" menu item (via Module Grants module) is necessary to view draft content (fixed in D7 but I can't wait for all the modules to be ported), but I'm pretty sure it'll confuse the content editors. An alternative is to use a PHP framework. I've read a few threads suggesting that it will take roughly the same amount of time using a good framework as it will to bend Drupal to my willing... maybe wishful thinking? I'm looking at Symfony, which gives me a basic auto-generated back-end, but which I believe I can customize to my heart's content. How do you make Drupal accessible to non-savvy content editors? If you recommend a PHP framework, which one? TIA!

    Read the article

  • Add new row in a databound form with a Oracle Sequence as the primary key

    - by Ranhiru
    I am connecting C# with Oracle 11g. I have a DataTable which i fill using an Oracle Data Adapter. OracleDataAdapter da; DataTable dt = new DataTable(); da = new OracleDataAdapter("SELECT * FROM Author", con); da.Fill(dt); I have few text boxes that I have databound to various rows in the data table. txtAuthorID.DataBindings.Add("Text", dt, "AUTHORID"); txtFirstName.DataBindings.Add("Text", dt, "FIRSTNAME"); txtLastName.DataBindings.Add("Text", dt, "LASTNAME"); txtAddress.DataBindings.Add("Text", dt, "ADDRESS"); txtTelephone.DataBindings.Add("Text", dt, "TELEPHONE"); txtEmailAddress.DataBindings.Add("Text", dt, "EMAIL"); I also have a DataGridView below the Text Boxes, showing the contents of the DataTable. dgvAuthor.DataSource = dt; Now when I want to add a new row, i do bm.AddNew(); where bm is defined in Form_Load as BindingManagerBase bm; bm = this.BindingContext[dt]; And when the save button is clicked after all the information is entered and validated, i do this.BindingContext[dt].EndCurrentEdit(); try { da.Update(dt); } catch (Exception ex) { MessageBox.Show(ex.Message); } However the problem comes where when I usually enter a row to the database (using SQL Plus) , I use a my_pk_sequence.nextval for the primary key. But how do i specify that when i add a new row in this method? I catch this exception ORA-01400: cannot insert NULL into ("SYSMAN".AUTHOR.AUTHORID") which is obvious because nothing was specified for the primary key. How do get around this? Thanx a lot in advance :)

    Read the article

  • Caching sitemaps in django

    - by michuk
    I implemented a simple sitemap class using django's default sitemap app. As it was taking a long time to execute, I added manual caching: class ShortReviewsSitemap(Sitemap): changefreq = "hourly" priority = 0.7 def items(self): # try to retrieve from cache result = get_cache(CACHE_SITEMAP_SHORT_REVIEWS, "sitemap_short_reviews") if result!=None: return result result = ShortReview.objects.all().order_by("-created_at") # store in cache set_cache(CACHE_SITEMAP_SHORT_REVIEWS, "sitemap_short_reviews", result) return result def lastmod(self, obj): return obj.updated_at The problem is that memcache allows only max 1MB object. This one was bigger that 1MB, so storing into cache failed: >7 SERVER_ERROR object too large for cache The problem is that django has an automated way of deciding when it should divide the sitemap file into smalled ones. According to the docs (http://docs.djangoproject.com/en/dev/ref/contrib/sitemaps/): You should create an index file if one of your sitemaps has more than 50,000 URLs. In this case, Django will automatically paginate the sitemap, and the index will reflect that. What do you think would be the best way to enable caching sitemaps? - Hacking into django sitemaps framework to restrict a single sitemap size to, let's say, 10,000 records seems like the best idea. Why was 50,000 chosen in the first place? Google advice? random number? - Or maybe there is a way to allow memcached store bigger files? - Or perhaps onces saved, the sitemaps should be made available as static files? This would mean that instead of caching with memcached I'd have to manually store the results in the filesystem and retrieve them from there next time when the sitemap is requested (perhaps cleaning the directory daily in a cron job). All those seem very low level and I'm wondering if an obvious solution exists...

    Read the article

  • Combo box in a scrollable panel causing problems

    - by Dennis
    I have a panel with AutoScroll set to true. In it, I am programmatically adding ComboBox controls. If I add enough controls to exceed the viewable size of the panel a scroll bar appears (so far so good). However, if I open one of the combo boxes near the bottom of the viewable area the combo list isn't properly displayed and the scrollable area seems to be expanded. This results in all of the controls being "pulled" to the new bottom of the panel with some new blank space at the top. If I continue to tap on the drop down at the bottom of the panel the scrollable area will continue to expand indefinitely. I'm anchoring the controls to the left, right and top so I don't think anchoring is involved. Is there something obvious that could be causing this? Update: It looks like the problem lies with anchoring the controls to the right. If I don't anchor to the right then I don't get the strange behavior. However, without right anchoring the control gets cut off by the scroll bar. Here's a simplified test case I built that shows the issue: public Form1() { InitializeComponent(); Panel panel = new Panel(); panel.Size = new Size(80, 200); panel.AutoScroll = true; for (int i = 0; i < 10; ++i) { ComboBox cb = new ComboBox(); cb.Anchor = AnchorStyles.Left | AnchorStyles.Right | AnchorStyles.Top; cb.Items.Add("Option 1"); cb.Items.Add("Option 2"); cb.Items.Add("Option 3"); cb.Items.Add("Option 4"); cb.Location = new Point(0, i * 24); panel.Controls.Add(cb); } Controls.Add(panel); } If you scroll the bottom of the panel and tap on the combo boxes near the bottom you'll notice the strange behavior.

    Read the article

  • ASP.NET MVC - Binding a Child Entity to the Model

    - by Nathan Taylor
    This one seems painfully obvious to me, but for some reason I can't get it working the way I want it to. Perhaps it isn't possible the way I am doing it, but that seems unlikely. This question may be somewhat related: http://stackoverflow.com/questions/1274855/asp-net-mvc-model-binding-related-entities-on-same-page. I have an EditorTemplate to edit an entity with multiple related entity references. When the editor is rendered the user is given a drop down list to select related entities from, with the drop down list returning an ID as its value. <%=Html.DropDownListFor(m => m.Entity.ID)%> When the request is sent the form value is named as expected: "Entity.ID", however my strongly typed Model defined as an action parameter doesn't have Entity.ID populated with the value passed in the request. public ActionResult AddEntity(EntityWithChildEntities entityWithChildEntities) { } I tried fiddling around with the Bind() attribute and specified Bind(Include = "Entity.ID") on the entityWithChildEntities, but that doesn't seem to work. I also tried Bind(Include = "Entity"), but that resulted in the ModelBinder attempting to bind a full "Entity" definition (not surprisingly). Is there any way to get the default model binder to fill the child entity ID or will I need to add action parameters for each child entity's ID and then manually copy the values into the model definition?

    Read the article

  • MySQL Cursor Issue

    - by James Inman
    I've got the following code - this is the first time I've really attempted using cursors. DELIMITER $$ DROP PROCEDURE IF EXISTS demo$$ DROP TABLE IF EXISTS temp$$ CREATE TEMPORARY TABLE temp( id INTEGER NOT NULL AUTO_INCREMENT, start DATETIME NOT NULL, end DATETIME NOT NULL, PRIMARY KEY(id) ) $$ CREATE PROCEDURE demo() BEGIN DECLARE done INT DEFAULT 0; DECLARE a, b DATETIME; DECLARE cur1 CURSOR FOR SELECT MAX(end) AS end FROM ( SELECT id, start, end, @r := @r + (start > @edate) AS num, @edate := GREATEST(@edate, end) FROM ( SELECT @r := 0, @edate := '0001-01-01' ) vars, student_lectures WHERE ( student_id = 1 AND start >= '2010-04-26 00:00:00' AND end <= '2010-04-30 23:59:59' ) ORDER BY start ) q GROUP BY num; DECLARE cur2 CURSOR FOR SELECT MIN(start) AS start FROM ( SELECT id, start, end, @r := @r + (start > @edate) AS num, @edate := GREATEST(@edate, end) FROM ( SELECT @r := 0, @edate := '0001-01-01' ) vars, student_lectures WHERE ( student_id = 1 AND start >= '2010-04-26 00:00:00' AND end <= '2010-04-30 23:59:59' ) ORDER BY start ) q GROUP BY num LIMIT 1, 18446744073709551615; DECLARE CONTINUE HANDLER FOR NOT FOUND SET done = 1; OPEN cur1; OPEN cur2; REPEAT FETCH cur1 INTO a; FETCH cur2 INTO b; IF NOT done THEN INSERT INTO temp(start, end) VALUES(a,b); END IF; UNTIL done END REPEAT; CLOSE cur1; CLOSE cur2; END $$ SELECT * FROM temp; I'm not getting anything outputted into the temp table. Running the following query gives me output, so I know there's rows it should be matching - but I imagine I've made some obvious mistake. SELECT MAX(end) AS end FROM ( SELECT id, start, end, @r := @r + (start > @edate) AS num, @edate := GREATEST(@edate, end) FROM ( SELECT @r := 0, @edate := '0001-01-01' ) vars, student_lectures WHERE ( student_id = 1 AND start >= '2010-04-26 00:00:00' AND end <= '2010-04-30 23:59:59' ) ORDER BY start ) q GROUP BY num; The output this query returns: +---------------------+ | end | +---------------------+ | 2010-04-26 13:00:00 | | 2010-04-26 18:15:00 | | 2010-04-27 11:00:00 | | 2010-04-27 13:00:00 | | 2010-04-27 18:15:00 | | 2010-04-28 13:00:00 | | 2010-04-29 13:00:00 | | 2010-04-29 18:15:00 | | 2010-04-30 13:00:00 | | 2010-04-30 15:15:00 | | 2010-04-30 17:15:00 | +---------------------+ 11 rows in set (0.02 sec)

    Read the article

  • Good mobile oriented GWT widget library alternatives

    - by Michael Donohue
    I've been developing a travel planning site - tripgrep.com - which is built on appengine, GWT and smartgwt, among other technologies. It is still early days, and the site is now working well on my development environment, which is either a windows or mac computer. However, I am frequently talking up the website to my friends when we are at a bar or other venue, so I am standing there while they try to access the site via an iPhone, Android or Blackberry - I've witnessed all three. It has been painfully obvious that the browser based frontend takes a long time to download on a mobile device. I am pretty sure this is because of the javascript download for SmartGWT. So, I would like to look at alternatives to SmartGWT. What I like about SmartGWT is that it has a reasonable look and feel out of the box - I don't need to learn any design or css and it has an office application look. This is considerably better than the GWT built-in widgets, which just get a blue border. The better look-and-feel is why I went with SmartGWT early on. However, the slow load times are killing me on these mobile demos. So now I want a fast loading widget alternative that has good look-and-feel out of the box. The features I care about are: tabs, good form layout, Google maps API integration, grid data viewing. If those are all available in a library that loads quickly on a mobile device, then that's the library I want.

    Read the article

  • Detecting what changed in an HTML Textfield

    - by teehoo
    For a major school project I am implementing a real-time collaborative editor. For a little background, basically what this means is that two(or more) users can type into a document at the same time, and their changes are automatically propagated to one another (similar to Etherpad). Now my problem is as follows: I want to be able to detect what changes a user carried out onto an HTML textfield. They could: Insert a character Delete a character Paste a string of characters Cut a string of characters I want to be able to detect which of these changes happened and then notify other clients similar to "insert character 'c' at position 2" etc. Anyway I was hoping to get some advice on how I would go about implementing the detection of these changes? My first attempt was to consider the carot position before and after a change occurred, but this failed miserably. For my second attempt I was thinking about doing a diff on the entire contents of the textfields old and new value. Am I missing anything obvious with this solution? Is there something simpler?

    Read the article

  • Deleting a non-owned dynamic array through a pointer

    - by ayanzo
    Hello all, I'm relatively novice when it comes to C++ as I was weened on Java for much of my undergraduate curriculum (tis a shame). Memory management has been a hassle, but I've purchased a number books on ansi C and C++. I've poked around the related questions, but couldn't find one that matched this particular criteria. Maybe it's so obvious nobody mentions it? This question has been bugging me, but I feel as if there's a conceptual point i'm not utilizing. Suppose: char original[56]; cstr[0] = 'a'; cstr[1] = 'b'; cstr[2] = 'c'; cstr[3] = 'd'; cstr[4] = 'e'; cstr[5] = '\0'; char *shaved = shavecstr(cstr); delete[] cstrn; where char* shavecstr(char* cstr) { size_t len = strlen(cstr); char* ncstr = new char[len]; strcpy(ncstr,cstr); return ncstr; } In that the whole point is to have 'original' be a buffer that fills with characters and routinely has its copy shaved and used elsewhere. To prevent leaks, I want to free up the memory held by 'shaved' to be used again after it passes through some arguments. There is probably a good reason for why this is restricted, but there should be some way to free the memory as by this configuration, there is no way to access the original owner (pointer) of the data.

    Read the article

  • Getting started with massive data

    - by Max
    I'm a math guy and occasionally do some statistics/machine learning analysis consulting projects on the side. The data I have access to are usually on the smaller side, at most a couple hundred of megabytes (and almost always far less), but I want to learn more about handling and analyzing data on the gigabyte/terabyte scale. What do I need to know and what are some good resources to learn from? Hadoop/MapReduce is one obvious start. Is there a particular programming language I should pick up? (I primarily work now in Python, Ruby, R, and occasionally Java, but it seems like C and Clojure are often used for large-scale data analysis?) I'm not really familiar with the whole NoSQL movement, except that it's associated with big data. What's a good place to learn about it, and is there a particular implementation (Cassandra, CouchDB, etc.) I should get familiar with? Where can I learn about applying machine learning algorithms to huge amounts of data? My math background is mostly on the theory side, definitely not on the numerical or approximation side, and I'm guessing most of the standard ML algorithms don't really scale. Any other suggestions on things to learn would be great!

    Read the article

  • KD-Trees and missing values (vector comparison)

    - by labratmatt
    I have a system that stores vectors and allows a user to find the n most similar vectors to the user's query vector. That is, a user submits a vector (I call it a query vector) and my system spits out "here are the n most similar vectors." I generate the similar vectors using a KD-Tree and everything works well, but I want to do more. I want to present a list of the n most similar vectors even if the user doesn't submit a complete vector (a vector with missing values). That is, if a user submits a vector with three dimensions, I still want to find the n nearest vectors (stored vectors are of 11 dimensions) I have stored. I have a couple of obvious solutions, but I'm not sure either one seem very good: Create multiple KD-Trees each built using the most popular subset of dimensions a user will search for. That is, if a user submits a query vector of thee dimensions, x, y, z, I match that query to my already built KD-Tree which only contains vectors of three dimensions, x, y, z. Ignore KD-Trees when a user submits a query vector with missing values and compare the query vector to the vectors (stored in a table in a DB) one by one using something like a dot product. This has to be a common problem, any suggestions? Thanks for the help.

    Read the article

  • How can I get the mapi system stub dll to pass extended mapi calls to my dll?

    - by Bogatyr
    For various reasons (questioning the reasons is not helpful to me), I'd like to implement my own extended mapi dll for windows xp. I have a skeleton dll now, just a few entrypoints exist for testing, but the system mapi stub (c:\windows\system32\mapi32.dll, I've checked that it's identical to mapistub.dll) will not pass through calls to my dll, while it happily passes the same calls through to MS Outlook's msmapi32.dll, (MAPIInitialize, MAPILoginEx are two such calls). There's some secret handshake between the stub and the extended mapi dll wherein the stub checks that "yup, it's an extended mapi dll": maybe it's the presence of some additional entrypoints I haven't implemented yet, maybe it's the return value from some function, I don't know. I've tried tracing a sample app I wrote that calls MAPIInitialize with STraceNT and ProcessMonitor but that didn't show anything obvious. Tracing has shown that indeed the stub loads my dll, but then finds the secret sauce is missing apparently, and returns an error code instead of calling my dll's function. What more could be needed for calling MAPIInitialize than the presence of MAPIInitialize in my dll's exports table? GetProcAddress says it's there. What I'd like to know is how to minimally extend my skeleton extended mapi dll so that the stub mapi dll will pass through extended mapi calls to my dll. What's the secret sauce? I'd rather not spend a painful week in msvc reverse engineering the stub behavior.

    Read the article

  • What are the benefits of the PHP the different PHP compression libraries?

    - by Christopher W. Allen-Poole
    I've been looking into ways to compress PHP libraries, and I've found several libraries which might be useful, but I really don't know much about them. I've specifically been reading about bcompiler and PHAR libraries. Is there any performance benefit in either of these? Are there any "gotchas" I need to watch out for? What are the relative benefits? Do either of them add to/detract from performance? I'm also interested in learning of other libs which might be out there which are not obvious in the documentation? As an aside, does anyone happen to know whether these work more like zip files which just happen to have the code in there, or if they operate more like Python's pre-compiling which actually runs a pseudo-compiler? ======================= EDIT ======================= I've been asked, "What are you trying to accomplish?" Well, I suppose the answer is that this is all hypothetical. It is a combination of these: What if my pet project becomes the most popular web project on earth and I want to distribute it quickly and easily? (hay, a man can dream, right?) It also seems if using PHAR can be done easily, it would be the best way to create a subversion snapshot. Python has this really cool pre-compiling policy, I wonder if PHP has something like that? These libraries seem to do something similar. Will they do that? Hey, these libraries seem pretty neat, but I'd like clarification on the differences as they seem to do the same thing

    Read the article

  • Strange Bubble sort behaviour.

    - by user271528
    Can anyone explain why this bubble sort function doesn't work and why I loose number in my output. I'm very new to C, so please forgive me if this is something very obvious I have missed. #include <stdio.h> #include int bubble(int array[],int length) { int i, j; int temp; for(i = 0; i < (length); ++i) { for(j = 0; j < (length - 1); ++j) { if(array[i] array[i+1]) { temp = array[i+1]; array[i+1] = array[i]; array[i] = temp; } } } return 0; } int main() { int array[] = {12,234,3452,5643,0}; int i; int length; length = (sizeof(array)/sizeof(int)); printf("Size of array = %d\n", length); bubble(array, length); for (i = 0; i < (length); ++i) { printf("%d\n", array[i]); } return 0; } Output Size of array = 5 12 234 3452 0 0

    Read the article

  • VBA compare and sort strings with quirky characters

    - by Smandoli
    I am comparing text values from two DAO recordsets in MS Access. I sort on the text field, then go through both recordsets comparing the values from each. The sets are substantially different and while they're mostly alpha-numeric, spaces and symbols like hyphens and periods are very common. My program depends on predictable sorting and fool-proof comparing. But unfortunately, the sort will rank two values differently than the comparison function. StrComp is the obvious first choice: varResult = StrComp(Val_1, Val_2) RFA-300 14.9044 RFA300 14-2044 But for the two pairs above, StrComp returns a different value than one would expect based on the sort. Including vbTextCompare or vbBinaryCompare affects StrComp's result, but not so as to solve the problem. Note the values must always be compared as strings. Of course I make sure that "14-2044" and "14.9044" aren't evaluated as -2030 and ~15. That's not the cause of my problem. I learned API-based functions are more reliable for quirky texts, so I tried these: varResult = CompareString(LOCALE_SYSTEM_DEFAULT, _ SORT_STRINGSORT, strVal_2, -1, strVal_1, -1) varResult = CompareString(LOCALE_SYSTEM_DEFAULT, _ NORM_IGNOREWIDTH, strVal_2, -1, strVal_1, -1) The first one returns the opposite of StrComp. The second one returns the same as StrComp. But neither yields a result that is consistent with the sort order. (NORM_IGNOREWIDTH is probably not relevant, but I needed a place-holder substitute and it looked as good as any.) UPDATE: This is a complete rewrite of the original post, deleting all the info about why I really need this -- just take my word for it and enjoy the brevity.

    Read the article

  • Detect IE version in Javascript

    - by Chad Decker
    I want to bounce users of our web site to an error page if they're using a version of Internet Explorer prior to v9. It's just not worth our time and money to support IE pre-v9. Users of all other non-IE browsers are fine and shouldn't be bounced. Here's the proposed code: if(navigator.appName.indexOf("Internet Explorer")!=-1){ //yeah, he's using IE var badBrowser=( navigator.appVersion.indexOf("MSIE 9")==-1 && //v9 is ok navigator.appVersion.indexOf("MSIE 1")==-1 //v10, 11, 12, etc. is fine too ); if(badBrowser){ // navigate to error page } } Will this code do the trick? To head off a few comments that will probably be coming my way: [1] Yes, I know that users can forge their useragent string. I'm not concerned. [2] Yes, I know that programming pros prefer sniffing out feature-support instead of browser-type but I don't feel this approach makes sense in this case. I already know that all (relevant) non-IE browsers support the features that I need and that all pre-v9 IE browsers don't. Checking feature by feature throughout the site would be a waste. [3] Yes, I know that someone trying to access the site using IE v1 (or = 20) wouldn't get 'badBrowser' set to true and the warning page wouldn't be displayed properly. That's a risk we're willing to take. [4] Yes, I know that Microsoft has "conditional comments" that can be used for precise browser version detection. IE no longer supports conditional comments as of IE 10, rendering this approach absolutely useless. Any other obvious issues to be aware of? Thanks.

    Read the article

  • Binding ListBox ItemCount to IvalueConverter

    - by Ben
    Hi All, I am fairly new to WPF so forgive me if I am missing something obvious. I'm having a problem where I have a collection of AggregatedLabels and I am trying to bind the ItemCount of each AggregatedLabel to the FontSize in my DataTemplate so that if the ItemCount of an AggregatedLabel is large then a larger fontSize will be displayed in my listBox etc. The part that I am struggling with is the binding to the ValueConverter. Can anyone assist? Many thanks! XAML Snippet <DataTemplate x:Key="TagsTemplate"> <WrapPanel> <TextBlock Text="{Binding Name, Mode=Default}" TextWrapping="Wrap" FontSize="{Binding ItemCount, Converter={StaticResource CountToFontSizeConverter}, Mode=Default}" Foreground="#FF0D0AF7"/> </WrapPanel> </DataTemplate> <ListBox x:Name="tagsList" ItemsSource="{Binding AggregatedLabels, Mode=Default}" ItemTemplate="{StaticResource TagsTemplate}" Style="{StaticResource tagsStyle}" Margin="200,10,16.171,11.88" /> AggregatedLabel Collection using (DB2DataReader dr = command.ExecuteReader()) { while (dr.Read()) { AggregatedLabelModel aggLabel = new AggregatedLabelModel(); aggLabel.ID = Convert.ToInt32(dr["LABEL_ID"]); aggLabel.Name = dr["LABEL_NAME"].ToString(); LabelData.Add(aggLabel); } } Converter public class CountToFontSizeConverter : IValueConverter { #region IValueConverter Members public object Convert(object value, Type targetType, object parameter, System.Globalization.CultureInfo culture) { const int minFontSize = 6; const int maxFontSize = 38; const int increment = 3; int count = (int)value; if ((minFontSize + count + increment) < maxFontSize) { return minFontSize + count + increment; } return maxFontSize; } public object ConvertBack(object value, Type targetType, object parameter, System.Globalization.CultureInfo culture) { throw new NotImplementedException(); } #endregion } AggregatedLabel Class public class AggregatedLabelModel { public int ID { get; set; } public string Name { get; set; } } CollectionView ListCollectionView labelsView = new ListCollectionView(LabelData); labelsView.GroupDescriptions.Add(new PropertyGroupDescription("Name"));

    Read the article

  • How to validate presence of an uploaded file in rails?

    - by brad
    I'm playing around creating a rails file uploader and have struck a problem that should have an obvious solution. How do I check that a file has been selected in my form and uploaded? Here is my new.html.erb view <h2>Upload File</h2> <% form_for(@upload_file, :url => {:action => 'save'}, :html => {:multipart => true}) do |f| %> <%= f.error_messages %> <p> <%= f.label :file -%> <%= f.file_field :upload -%> </p> <p> <%= f.label :description %> <%= f.text_field :description %> </p> <p> <%= f.label :file_type %> <%= f.select :file_type, ["XML Data"] %> </p> <p><%= f.submit 'Upload File' %></p> <% end %> and here is my upload_file.rb model class UploadFile < ActiveRecord::Base validates_presence_of :description validates_presence_of :file_type validates_presence_of :upload def upload=(upload_file_field) self.name = "#{Time.now.strftime("%Y%m%d%H%M%S")}_#{upload_file_field.original_filename}" File.open("#{RAILS_ROOT}/public/upload/#{self.name}", "wb") { |f| f.write(upload_file_field.read) } end end If I use this as shown here, the validation validates_presence_of :upload always fails and I am returned to my form with an error message. I'd be very grateful if someone could explain how to do this validation correctly, and I'd be even more grateful if they could explain why it works. Thanks.

    Read the article

< Previous Page | 116 117 118 119 120 121 122 123 124 125 126 127  | Next Page >