Search Results

Search found 4835 results on 194 pages for 'practice'.

Page 19/194 | < Previous Page | 15 16 17 18 19 20 21 22 23 24 25 26  | Next Page >

  • best practice to persist classes model

    - by Yaron Naveh
    My application contains a set of model classes. e.g. Person, Department... The user changes values for instances of these classes in the UI and the classes are persisted to my "project" file. Next time the user can open and edit the project. Next version of my product may change the model classes drastically. It will still need to open existing projects files (I will know how to handle missing data). How is it best to persist my model classes to the project file? The easiest way to persist classes is Data contract serialization. However it will fail on breaking changes (I expect to have such). How to handle this? use some other persistence, e.g. name-value collection or db which is more tolerance ship a "project converter" application to migrate old projects. This requires to either ship with both old and new models or to manipulate xml, which is best?

    Read the article

  • Best practice - logging events (general) and changes (database)

    - by b0x0rz
    need help with logging all activities on a site as well as database changes. requirements: * should be in database * should be easily searchable by initiator (user name / session id), event (activity type) and event parameters i can think of a database design but either it involves a lot of tables (one per event) so i can log each of the parameters of an event in a separate field OR it involves one table with generic fields (7 int numeric and 7 text types) and log everything in one table with event type field determining what parameter got written where (and hoping that i don't need more than 7 fields of a certain type, or 8 or 9 or whatever number i choose)... example of entries (the usual things): [username] login failed @datetime [username] login successful @datetime [username] changed password @datetime, estimated security of password [low/ok/high/perfect] @datetime [username] clicked result [result number] [result id] after searching for [search string] and got [number of results] @datetime [username] clicked result [result number] [result id] after searching for [search string] and got [number of results] @datetime [username] changed profile name from [old name] to [new name] @datetime [username] verified name with [credit card type] credit card @datetime datbase table [table name] purged of old entries @datetime via automated process etc... so anyone dealt with this before? any best practices / links you can share? i've seen it done with the generic solution mentioned above, but somehow that goes against what i learned from database design, but as you can see the sheer number of events that need to be trackable (each user will be able to see this info) is giving me headaches, BUT i do LOVE the one event per table solution more than the generic one. any thoughts? edit: also, is there maybe an authoritative list of such (likely) events somewhere? thnx stack overflow says: the question you're asking appears subjective and is likely to be closed. my answer: probably is subjective, but it is directly related to my issue i have with designing a database / writing my code, so i'd welcome any help. also i tried narrowing down the ideas to 2 so hopefully one of these will prevail, unless there already is an established solution for these kinds of things.

    Read the article

  • Cocoa Touch best practice for capturing/logging/diagnosing crashes

    - by Dylan
    As I get closer to releasing my first public iPhone app I'm concerned about catching crashes in the field. I'm curious to hear how others have gone about this. I'm not sure what's possible outside of the debugger. Is all lost with an EXC_BAD_ACCESS or can I still catch it and get something useful into a log? Is the program main() the spot to put a @try/@catch?

    Read the article

  • best practice - loging events (general) and changes (database)

    - by b0x0rz
    need help with logging all activities on a site as well as database changes. requirements: * should be in database * should be easily searchable by initiator (user name / session id), event (activity type) and event parameters i can think of a database design but either it involves a lot of tables (one per event) so i can log each of the parameters of an event in a separate field OR it involves one table with generic fields (7 int numeric and 7 text types) and log everything in one table with event type field determining what parameter got written where (and hoping that i don't need more than 7 fields of a certain type, or 8 or 9 or whatever number i choose)... example of entries (the usual things): [username] login failed @datetime [username] login successful @datetime [username] changed password @datetime, estimated security of password [low/ok/high/perfect] @datetime [username] clicked result [result number] [result id] after searching for [search string] and got [number of results] @datetime [username] clicked result [result number] [result id] after searching for [search string] and got [number of results] @datetime [username] changed profile name from [old name] to [new name] @datetime [username] verified name with [credit card type] credit card @datetime datbase table [table name] purged of old entries @datetime etc... so anyone dealt with this before? any best practices / links you can share? i've seen it done with the generic solution mentioned above, but somehow that goes against what i learned from database design, but as you can see the sheer number of events that need to be trackable (each user will be able to see this info) is giving me headaches, BUT i do LOVE the one event per table solution more than the generic one. any thoughts? edit: also, is there maybe an authoritative list of such (likely) events somewhere? thnx stack overflow says: the question you're asking appears subjective and is likely to be closed. my answer: probably is subjective, but it is directly related to my issue i have with designing a database / writing my code, so i'd welcome any help. also i tried narrowing down the ideas to 2 so hopefully one of these will prevail, unless there already is an established solution for these kinds of things.

    Read the article

  • Secure database connection. DAL .net architecture best practice

    - by Andrew Florko
    We have several applications that are installed in several departments that interact with database via Intranet. Users tend to use weak passwords or store login/password written on a shits of paper where everybody can see them. I'm worried about login/password leakage & want to minimize consequences. Minimizing database-server attack surface by hiding database-server from Intranet access would be a great idea also. I'm thinking about intermediary data access service method-based security. It seems more flexible than table-based or connection-based database-server one. This approach also allows to hide database-server from public Intranet. What kind of .net technologies and best practices would you suggest? Thank in you in advance!

    Read the article

  • IF-block brackets: best practice

    - by MasterPeter
    I am preparing a short tutorial for level 1 uni students learning JavaScript basics. The task is to validate a phone number. The number must not contain non-digits and must be 14 digits long or less. The following code excerpt is what I came up with and I would like to make it as readable as possible. if ( //set of rules for invalid phone number phoneNumber.length == 0 //empty || phoneNumber.length > 14 //too long || /\D/.test(phoneNumber) //contains non-digits ) { setMessageText(invalid); } else { setMessageText(valid); } A simple question I can not quite answer myself and would like to hear your opinions on: How to position the surrounding (outermost) brackets? It's hard to see the difference between a normal and a curly bracket. Do you usually put the last ) on the same line as the last condition? Do you keep the first opening ( on a line by itself? Do you wrap each individual sub-condition in brackets too? Do you align horizontally the first ( with the last ), or do you place the last ) in the same column as the if? Do you keep ) { on a separate line or you place the last ) on the same line with the last sub-condition and then place the opening { on a new line? Or do you just put the ) { on the same line as the last sub-condition? Community wiki.

    Read the article

  • asp.net web forms best practice to retrieve dynamic server control values

    - by Andrew Florko
    I populate web form with dynamic list of exams from database. I want user to enter examination marks for each exam. There is list of exam titles and textbox near each title. I create list with repeater control: <asp:Repeater ID="rptExams" runat="server" onitemdatabound="rptExams_ItemDataBound" > <ItemTemplate> <tr> <td> <asp:Literal runat="server" ID="ltTitle"/> </td> <td> <asp:HiddenField runat="server" ID="hfId"/> <asp:Textbox runat="server" ID="tbMark"/> </td> </tr> </ItemTemplate> </asp:Repeater> And bind data to repeater on page_init: class Exam { public int Id { get; set;} public string Title { get; set;} } ... // this list is retrieved from database actually Exam[] Exams = new Exam[] { new Exam { Id = 1, Title = "Math"}, new Exam { Id = 2, Title = "History"} }; ... protected void Page_Init(object sender, EventArgs e) { rptExams.DataSource = Exams; rptExams.DataBind(); } So far so good. Then I have to retrieve data on postback. I have two ways but both of them looks ugly. Idea is to store dynamically created databounded controls on ItemDataBoundEvent in Page_Init stage, and process their values in Page_Load stage. It looks like this: private Dictionary<HiddenField, TextBox> Id2Mark = new Dictionary<HiddenField, TextBox>(); protected void rptExams_ItemDataBound(object sender, RepeaterItemEventArgs e) { ... if (IsPostBack) { var tbMark = (TextBox)e.Item.FindControl("tbMark"); var hfId = (HiddenField)e.Item.FindControl("hfId"); // store dynamically created controls Id2Mark.Add(hfId, tbMark); } ... } protected void Page_Load(object sender, EventArgs e) { if (IsPostBack) { foreach (var pair in Id2Mark) { int examId = Int32.Parse(pair.Key.Value); string mark = pair.Value.Text; // PROCESS } ... I'm completely sure there is a better way to retrieve data from dynamically created controls. Thank you in advance!

    Read the article

  • Best practice to structure large html-based project

    - by AntonAL
    I develop Rails based website, enjoying using partials for some common "components" Recently, i faced a problem, that states with CSS interference. Styles for one component (described in css) override styles for another components. For example, one component has ... <ul class="items"> ... and another component has it too. But that ul's has different meaning in these two components. On the other hand, i want to "inherit" some styles for one component from another. For example: Let, we have one component, called "post" <div class="post"> <!-- post's stuff --> <ul class="items"> ... </ul> </div And another component, called "new-post": <div class="new-post"> <!-- post's stuff --> <ul class="items"> ... </ul> <!-- new-post's stuff --> <div class="tools">...</div> </div Post and new-post have something similar ("post's stuff") and i want to make CSS rules to handle both "post" and "new-post" New post has "subcomponents", for example - editing tools, that has also: <ul class="items"> This is where CSS rules starting to interfer - some rules, targeted for ul.items (in post and new-post) applies subcomponent of new-post, called "tools" On the one hand - i want to inherit some styles On the other hand, i want to get better incapsulation What are the best practices, to avoid such kind of problems ?

    Read the article

  • Best practice to test a web application, regarding domain name and integration with external service

    - by ycseattle
    I have run into these problems several times and was never able to find a comfortable solution. Let's say my website has the domain name MyDomain.com. When I run the tests on the test machine (a continuous integration server), I will modify the HOSTS file on this machine so the MyDomain.com is mapped to this local machine instead of the real production server. This doesn't work very well for many situations. For example, my application will create subdomain names user1.MyDomain.com dynamically but this is difficult to keep the testing flexible. Another problem is my web application will interact with Amazon S3, and sometimes other service like Amazon Simple Message Queue. I am only comfortable to include these interaction in my tests but I am never happy with my solution for mixing testing and production on Amazon services. Could somebody offer some tips on these issues? I would like to make my testing framework clean and flexible. I am sure this is a common question for all web applications and there must be a mature way to deal with these. Thanks!

    Read the article

  • html clickable layout area. best practice

    - by Andrew Florko
    I am bad in html layout but I have to produce it :) I want to make big button on a page that is implemented as div with children tags (maybe - a bad idea). I can handle click event on boundary-div with javascript but it requires javascript enabled. I can wrap boundary-div with "anchor" tag but is doesn't work in IE Please, suggest me the best way to implement this. <a href="..."> <table> <td> ... </td> <td> ... <table> ... </table> </td> </table> </a>

    Read the article

  • What is best practice about having one-many hibernate

    - by Patrick
    Hi all, I believe this is a common scenario. Say I have a one-many mapping in hibernate Category has many Item Category: @OneToMany( cascade = {CascadeType.ALL},fetch = FetchType.LAZY) @JoinColumn(name="category_id") @Cascade( value = org.hibernate.annotations.CascadeType.DELETE_ORPHAN ) private List<Item> items; Item: @ManyToOne(targetEntity=Category.class,fetch=FetchType.EAGER) @JoinColumn(name="category_id",insertable=false,updatable=false) private Category category; All works fine. I use Category to fully control Item's life cycle. But, when I am writing code to update Category, first I get Category out from DB. Then pass it to UI. User fill in altered values for Category and pass back. Here comes the problem. Because I only pass around Category information not Item. Therefore the Item collection will be empty. When I call saveOrUpdate, it will clean out all associations. Any suggestion on what's best to address this? I think the advantage of having Category controls Item is to easily main the order of Item and not to confuse bi-directly. But what about situation that you do want to just update Category it self? Load it first and merge? Thank you.

    Read the article

  • Java Best Practice for type resolution at runtime.

    - by Brian
    I'm trying to define a class (or set of classes which implement the same interface) that will behave as a loosely typed object (like JavaScript). They can hold any sort of data and operations on them depend on the underlying type. I have it working in three different ways but none seem ideal. These test versions only allow strings and integers and the only operation is add. Adding integers results in the sum of the integer values, adding strings concatenates the strings and adding an integer to a string converts the integer to a string and concatenates it with the string. The final version will have more types (Doubles, Arrays, JavaScript-like objects where new properties can be added dynamically) and more operations. Way 1: public interface DynObject1 { @Override public String toString(); public DynObject1 add(DynObject1 d); public DynObject1 addTo(DynInteger1 d); public DynObject1 addTo(DynString1 d); } public class DynInteger1 implements DynObject1 { private int value; public DynInteger1(int v) { value = v; } @Override public String toString() { return Integer.toString(value); } public DynObject1 add(DynObject1 d) { return d.addTo(this); } public DynObject1 addTo(DynInteger1 d) { return new DynInteger1(d.value + value); } public DynObject1 addTo(DynString1 d) { return new DynString1(d.toString()+Integer.toString(value)); } } ...and similar for DynString1 Way 2: public interface DynObject2 { @Override public String toString(); public DynObject2 add(DynObject2 d); } public class DynInteger2 implements DynObject2 { private int value; public DynInteger2(int v) { value = v; } @Override public String toString() { return Integer.toString(value); } public DynObject2 add(DynObject2 d) { Class c = d.getClass(); if(c==DynInteger2.class) { return new DynInteger2(value + ((DynInteger2)d).value); } else { return new DynString2(toString() + d.toString()); } } } ...and similar for DynString2 Way 3: public class DynObject3 { private enum ObjectType { Integer, String }; Object value; ObjectType type; public DynObject3(Integer v) { value = v; type = ObjectType.Integer; } public DynObject3(String v) { value = v; type = ObjectType.String; } @Override public String toString() { return value.toString(); } public DynObject3 add(DynObject3 d) { if(type==ObjectType.Integer && d.type==ObjectType.Integer) { return new DynObject3(Integer.valueOf(((Integer)value).intValue()+((Integer)value).intValue())); } else { return new DynObject3(value.toString()+d.value.toString()); } } } With the if-else logic I could use value.getClass()==Integer.class instead of storing the type but with more types I'd change this to use a switch statement and Java doesn't allow switch to use Classes. Anyway... My question is what is the best way to go about something thike this?

    Read the article

  • SQL Query - 20mil records - Best practice to return information

    - by eqiz
    I have a SQL database that has the following table: Table: PhoneRecords ID(identity Seed) FirstName LastName PhoneNumber ZipCode Very simple straight forward table. This table has over 20million records. I am looking for the best way to do queries that pull out records based off area codes from the table. For instance here is an example query that I have done. SELECT phonenumber, firstname FROM [PhoneRecords] WHERE (phone LIKE '2012042%') OR (phone LIKE '2012046%') OR (phone LIKE '2012047%') OR (phone LIKE '2012083%') OR (phone LIKE '2012088%') OR (phone LIKE '2012841%') As you can see this is an ugly query, but it would get the job done (if I wasn't running into timeout issues) Can anyone tell me the best way for speed/optimization to do the above query to display the results? Currently that query above takes around 2 hours to complete on a 9gb 1600mhz ram, i7 930 quadcore OC'd 4.01ghz. I obviously have the computer power required to do such a query, but still takes too long for queries.

    Read the article

  • Opinions regarding C++ programming practice

    - by Sagar
    I have a program that I am writing, not too big. Apart from the main function, it has about 15 other functions that called for various tasks at various times. The code works just fine all in one file, and as it is right now. However, I was wondering if anyone had any advice on whether it is smarter/more efficient/better programming to put those functions in a separate file different from where main is, or whether it even matters at all. If yes, why? If no, why not? I am not new at C++, but definitely not an expert either, so if you think this question is stupid, feel free to tell me so. Thanks for your time!

    Read the article

  • Best practice accessing an array set within a class

    - by user350599
    I have created a basic class for a customer. I haven't done this before and want to know the best way to access the data. Should I have a get() method for every field in the customer array or should I simply pass the customer array back and access with the page. i.e. Just return the array class Customer { protected $id; protected $customer; public function __construct($customer_id) { $this->id = $customer_id; $this->set_customer(); } protected function set_customer() { $query = mysql_query("SELECT * FROM customer WHERE id = '$this->id'"); $this->customer = mysql_fetch_row($query); } public function get_customer() { return $this->customer; } } versus create a method for each item in the array class Customer { protected $id; protected $customer; public function __construct($customer_id) { $this->id = $customer_id; $this->set_customer(); } protected function set_customer() { $query = mysql_query("SELECT * FROM customer WHERE id = '$this->id'"); $this->customer = mysql_fetch_row($query); } public function get_customer_name() { return $this->customer->customer_name; } ... ... } versus option 3 based on Tobias' feedback: (not sure if syntax is correct) class Customer { protected $id; protected $customer; public function __construct($customer_id) { $this->id = $customer_id; return $this->set_customer(); } protected function set_customer() { $query = mysql_query("SELECT * FROM customer WHERE id = '$this->id'"); return mysql_fetch_row($query); } }

    Read the article

  • What is the best practice to segment c#.net projects based on a single base project

    - by Anthony
    Honestly, I can't word my question any better without describing it. I have a base project (with all its glory, dlls, resources etc) which is a CMS. I need to use this project as a base for othe custom bake projects. This base project is to be maintained and updated among all custom bake projects. I use subversion (Collabnet and Tortise SVN) I have two questions: 1 - Can I use subversion to share the base project among other projects What I mean here is can I "Checkout" the base project into another "Checked Out" project and have both update and commit seperatley. So, to paint a picture, let's say I am working on a custom project and I modify the core/base prject in some way (which I know will suit the others) can I then commit those changes and upon doing so when I update the base project in the other "Checked out" resources will it pull the changes? In short, I would like not to have to manually deploy updated core files whenever I make changes into each seperate project. 2 - If I create a custom file (let's say an webcontrol or aspx page etc) can I have it compile seperatley from the base project Another tricky one to explain. When I publish my web application it creates DLLs based on the namespaces of projects attached to it. So I may have a number of DLLs including the "Website's" namespace DLL, which could simply be website. I want to be able to make a seperate, custom, control which does not compile into those DLLs as the custom files should not rely on those DLLS to run. Is it as simple to set a seperate namespace for those files like CustomFiles.ProjectName for example? Think of the whole idea as adding modules to the .NET project, I don't want the module's code in any of the core DLLs but I do need for module to be able to access the core dlls. (There is no need for the core project to access the module code as it should be one way only in theory, though I reckon it woould not be possible anyway without using JSON/SOAP or something like that, maybe I am wrong.) I want to create a pluggable environment much like that of Joomla/Wordpress as since PHP generally doesn't have to be compiled first I see this is the reason why all this is possible/easy. The idea is to allow pluggable themes, modules etc etc. (I haven't tried simply adding .NET themes after compile/publish but I am assuming this is possible anyway? OR does the compiler need to reference items in the files?)

    Read the article

  • Learn Prolog Now! DCG Practice Example

    - by Timothy
    I have been progressing through Learn Prolog Now! as self-study and am now learning about Definite Clause Grammars. I am having some difficulty with one of the Practical Session's tasks. The task reads: The formal language anb2mc2mdn consists of all strings of the following form: an unbroken block of as followed by an unbroken block of bs followed by an unbroken block of cs followed by an unbroken block of ds, such that the a and d blocks are exactly the same length, and the c and d blocks are also exactly the same length and furthermore consist of an even number of cs and ds respectively. For example, ε, abbccd, and aaabbbbccccddd all belong to anb2mc2mdn. Write a DCG that generates this language. I am able to write rules that generate andn, b2mc2m, and even anb2m and c2mndn... but I can't seem to join all these rules into anb2mc2mdn. The following are my rules that can generate andn and b2mc2m. s1 --> []. s1 --> a,s1,d. a --> [a]. d --> [d]. s2 --> []. s2 --> c,c,s2,d,d. c --> [c]. d --> [d]. Is anb2mc2mdn really a CFG, and is it possible to write a DCG using only what was taught in the lesson (no additional arguments or code, etc)? If so, can anyone offer me some guidance how I can join these so that I can solve the given task?

    Read the article

  • What should I do to practice?

    - by simion
    I start a year long industrial placement in September where i will be coding in Java predominantly. I am going to use the summer to brush up on my Java as in year one of the degree Java was the main language taught for OOP modules. However this year i have had no Java exposure except for an algorithms module, which was one of eight, so as you can see i am probably getting really rusty!. What i wanted to know is, how does the "real world" java programming differ from university coding and what do you suggest i brush up on that would be different to my normal workings. As a start I definitely need to get familiar with a professional IDE like NetBeans, opposed to having used BlueJ throughout but more specifically what coding practices should I get more familiar with. I appreciate they wont expect me to be a qualified full developer and will give me time, but I would like to hit the ground running as it were, with me having full hopes to secure a permanent position after I finish my degree.

    Read the article

  • Best Practice to write Connection string for heavy traffic ASP.NET Web Application

    - by hungrycoder
    What is the best way to define connection string for a web application that has minimum 500 live users on the internet in terms of connection pooling. And load factors to consider? Suppose I have connection string as follows: initial catalog=Northwind; Min Pool Size=20;Max Pool Size=500; data source=localhost; Connection Timeout=30; Integrated security=sspi" as Max Pool Size is 500 and as live users exceed 500 say 520 will the remaining 20 users experience slower page load?? Or what if I have connection string as follows which doesn't talks anything about pooling or Connection time out? How the application behaves then? initial catalog=Northwind; data source=localhost; Integrated security=sspi" I'm using "Using statements" however to access the MYSQL database.

    Read the article

  • dynamic searchable fields, best practice?

    - by boblu
    I have a Lexicon model, and I want user to be able to create dynamic feature to every lexicon. And I have a complicate search interface that let user search on every single feature (including the dynamic ones) belonged to Lexicon model. I could have used a serialized text field to save all the dynamic information if they are not for searching. In case I want to let user search on all fields, I have created a DynamicField Model to hold all dynamically created features. But imagine I have 1,000,000,000 lexicon, and if one create a dynamic feature for every lexicon, this will result creating 1,000,000,000 rows in DynamicField model. So the sql search function will become quite inefficient while a lot of dynamic features created. Is there a better solution for this situation? Which way should I take? searching for a better db design for dynamic fields try to tuning mysql(add cache fields, add index ...) with current db design

    Read the article

  • Best practice for using Wcf service by silverlight?

    - by bonefisher
    How would you structure the code for calling a wcf service in silverlight application? Using only-once instanciated wcf service-proxy (aka singleton) and using it across the whole SL app? If so, how did you solve the unsubscribing controls from ws-call-completed event? or creating the wcf service-proxy for each ws-call? Where do you close the proxy then?

    Read the article

  • Are protected constructors considered good practice?

    - by Álvaro G. Vicario
    I'm writing some little helper classes to handle trees. Basically, I have a node and a special root node that represents the tree. I want to keep it generic and simple. This is part of the code: <?php class Tree extends TreeNode{ public function addById($node_id, $parent_id, $generic_content){ if( $parent = $this->findNodeById($parent_id) ){ $parent->addChildById($node_id, $generic_content); } } } class TreeNode{ public function __construct($node_id, $parent_id, $generic_content){ // ... } protected function addChildById($node_id, $generic_content){ $this->children[] = new TreeNode($this->node_id, $node_id, $generic_content); } } $Categories = new Tree; $Categories->addById(1, NULL, $foo); $Categories->addById(2, NULL, $bar); $Categories->addById(3, 1, $gee); ?> My questions: Is it sensible to force TreeNode instances to be created through TreeNode::addById()? If it's so, would it be good practise to declare TreeNode::__construct() as private/protected?

    Read the article

  • python list/dict property best practice

    - by jterrace
    I have a class object that stores some properties that are lists of other objects. Each of the items in the list has an identifier that can be accessed with the id property. I'd like to be able to read and write from these lists but also be able to access a dictionary keyed by their identifier. Let me illustrate with an example: class Child(object): def __init__(self, id, name): self.id = id self.name = name class Teacher(object): def __init__(self, id, name): self.id = id self.name = name class Classroom(object): def __init__(self, children, teachers): self.children = children self.teachers = teachers classroom = Classroom([Child('389','pete')], [Teacher('829','bob')]) This is a silly example, but it illustrates what I'm trying to do. I'd like to be able to interact with the classroom object like this: #access like a list print classroom.children[0] #append like it's a list classroom.children.append(Child('2344','joe')) #delete from like it's a list classroom.children.pop(0) But I'd also like to be able to access it like it's a dictionary, and the dictionary should be automatically updated when I modify the list: #access like a dict print classroom.childrenById['389'] I realize I could just make it a dict, but I want to avoid code like this: classroom.childrendict[child.id] = child I also might have several of these properties, so I don't want to add functions like addChild, which feels very un-pythonic anyway. Is there a way to somehow subclass dict and/or list and provide all of these functions easily with my class's properties? I'd also like to avoid as much code as possible.

    Read the article

  • Multiset of shared_ptrs as a dynamic priority queue: Concept and practice

    - by Sarah
    I was using a vector-based priority queue typedef std::priority_queue< Event, vector< Event >, std::greater< Event > > EventPQ; to manage my Event objects. Now my simulation has to be able to find and delete certain Event objects not at the top of the queue. I'd like to know if my planned work-around can do what I need it to, and if I have the syntax right. I'd also like to know if dramatically better solutions exist. My plan is to make EventPQ a multiset of smart pointers to Event objects: typedef std::multi_set< boost::shared_ptr< Event > > EventPQ; I'm borrowing functions of the Event class from a related post on a multimap priority queue. // Event.h #include <cstdlib> using namespace std; #include <set> #include <boost/shared_ptr.hpp> class Event; typedef std::multi_set< boost::shared_ptr< Event > > EventPQ; class Event { public: Event( double t, int eid, int hid ); ~Event(); void add( EventPQ& q ); void remove(); bool operator < ( const Event & rhs ) const { return ( time < rhs.time ); } bool operator > ( const Event & rhs ) const { return ( time > rhs.time ); } double time; int eventID; int hostID; EventPQ* mq; EventPQ::iterator mIt; }; // Event.cpp Event::Event( double t, int eid, int hid ) { time = t; eventID = eid; hostID = hid; } Event::~Event() {} void Event::add( EventPQ& q ) { mq = &q; mIt = q.insert( boost::shared_ptr<Event>(this) ); } void Event::remove() { mq.erase( mIt ); mq = 0; mIt = EventPQ::iterator(); } I was hoping that by making EventPQ a container of pointers, I could avoid wasting time copying Events into the container and avoid accidentally editing the wrong copy. Would it be dramatically easier to store the Events themselves in EventPQ instead? Does it make more sense to remove the time keys from Event objects and use them instead as keys in a multimap? Assuming the current implementation seems okay, my questions are: Do I need to specify how to sort on the pointers, rather than the objects, or does the multiset automatically know to sort on the objects pointed to? If I have a shared_ptr ptr1 to an Event that also has a pointer in the EventPQ container, how do I find and delete the corresponding pointer in EventPQ? Is it enough to .find( ptr1 ), or do I instead have to find by the key (time)? Is the Event::remove() sufficient for removing the pointer in the EventPQ container? There's a small chance multiple events could be created with the same time (obviously implied in the use of multiset). If the find() works on event times, to avoid accidentally deleting the wrong event, I was planning to throw in a further check on eventID and hostID. Does this seem reasonable? (Dumb syntax question) In Event.h, is the declaration of dummy class Event;, then the EventPQ typedef, and then the real class Event declaration appropriate? I'm obviously an inexperienced programmer with very spotty background--this isn't for homework. Would love suggestions and explanations. Please let me know if any part of this is confusing. Thanks.

    Read the article

< Previous Page | 15 16 17 18 19 20 21 22 23 24 25 26  | Next Page >