more template:include/ questions
One of the problems I'm having with template:include/ and probably more with the XML nature of JSP is that html:XXX form elements have to be within html:form/html:form tags, and no in an include used by template:include/. I have common components that different forms use, but have to do either a cut-n-paste(tm) or use %@ include ... % instead? Any thoughts? -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
Re: template:put and get
I decided to do the second option, it works quite well. I don't like the idea of putting the while html:img/ tag into the template:put/ because that'd make the template file confusing for someone trying to maintain it (the image tag would never appear in the source for the included file, the developer could spend ages trying to work out what's going on) JR but replace the page= with the template:get/. Trouble is you can't nest JR tags. One options is to do the html without the jsp, and have JR img src=%= request.getContextPath() %/media/h_modify.gif width=315 height=23 border=0/ JR but that doesn't look great, but I can't think of anything else. The template:put/ JR seems to hide the content away somewhere away in the request, and I don't want to JR create 20 little html/jsp pages for the different headers that I have. JR Any ideas? You can place whole html:img tag to the template:put tag as its content and cut content and direct attributes. -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
template:put and get
I've got a situation where I want to include a header file with template:insert template=/share/inc_header.jsp template:put name=header_image content=/media/h_modify.gif direct=true/ /template:insert and then in the header file have html:img page=/media/h_modify.gif width=315 height=23 border=0/ but replace the page= with the template:get/. Trouble is you can't nest tags. One options is to do the html without the jsp, and have img src=%= request.getContextPath() %/media/h_modify.gif width=315 height=23 border=0/ but that doesn't look great, but I can't think of anything else. The template:put/ seems to hide the content away somewhere away in the request, and I don't want to create 20 little html/jsp pages for the different headers that I have. Any ideas? -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
Re: JSP expressions inside custom tags
try html:link href=%= \/\ + name + \/index.jsp\ % You need to have the %= % present the whole string, not just the single part. At 10:56 AM 8/06/01 +0200, you wrote: Hi! I am evaluating struts for a new project. One problem I wasn't able to solve was the integration of JSP expressions inside the attributes of custom tags. My code looks like this: logic:iterate id=pelement name=MyBean property=pelements % String name = ((PElement)pelement).getName(); % html:link href=/%= name %/index.jsp ^^^ Here is my problem! bean:write name=pelement property=longname/ /html:linkbr / /logic:iterate The links always contain the %= name % string instead of the evaluated expression. I tried it with different quoting techniques but without luck. What is the correct solution here? TIA, Erich -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
Re: New Windows, Struts, IE 5 and sessions dying
At 07:04 PM 1/06/01 +1200, you wrote: Summary: Any known bugs and workarounds with Multiple Windows in IE5, Servlet Sessions and Struts 1.0b1? A few things to note. Have a look to see if the redirect is forcing :80 (port 80) onto the end of the URL, this can cause the session get lost in the ether. Upgrade to struts-b2 and it'll go away. Another thing to note is that on NT you can have each IE run in a separate process, which gives it its own set of cookies, which will also loose the session value. I think it's probably the first one though, as you'd be using html:rewrite / to generate the URL for the javascript, and up until struts-b2 it would add the port into the address and confuse the browser. This happens in Netscape Navigator as well. -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
Re: Potential Security Flaw in Struts MVC
In the case at hand, nothing stops your user from logging on (so your security checks won't catch anything) and then hand typing a URL with query string parameters that maliciously or accidentally try to change things in the system. If the user is successful at doing this, it's shame on the app developer for listening to request parameters that you shouldn't. This is a good point. I'm finding my Actions and Forms have bucketloads (and that's the technical term for it) of User user; if ((user = (User) request.getSession().getAttribute(user) == null) return mapping.getMapping(index); and on, and on and on. I'd like to try and find a good way to simplify this as best that I can. (This example is required if the session times out, other examples appear when a browser auto-fills in a URL and the user submits it without the form fields. etc. Very bad karma in that case.) Of course, you need to take other defensive measures as well (like using the transaction control support to avoid accidental or malicious resubmits of the same data). I've seen this in the example application, is there any documentation on using it (as best as possible). Thanks. -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
Re: Potential Security Flaw in Struts MVC
At 11:53 PM 30/05/01 -0700, you wrote: A good way of removing the bucketloads :-} from your Action classes is to subclass ActionServlet and implement processActionPerform to do the logon check. It's not just for login though, that was the example I used, every action that generates a form needs to do this. Mostly it is checking against URL hacking. -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
Re: subclassing ActionServlet?
I made a sub-class of ActionServlet an overloaded the initXXX() method that created the connecion pool. I did this because my database manager beans needed access to the connection pool parameters. The other choice is to use something like a singleton or factory bean pattery that generates the objects (or returns an instance to an already created one) as you need it. The down side is it may not have the initialisation parameters it needs. [EMAIL PROTECTED] wrote: Hello struts-users, My webapp needs bunch of beans that should be available all the time and accessible from all JSP pages, so I guess I have to create the beans at the startup time and set them to the application context. How should I do this? My candidates are, 1) create a subclass of ActionServlet and do it there, 2) create a subclass of Servlet (unrelated to ActionServlet) and do it there. Which is better, or are there any other standard ways? I don't see any examples that use subclasses of ActionServlet, and I am wondering if there are some reasons not to. Thanks in advance, - kazumi
Re: Possible to access Init Params from JSP??
George Craig wrote: Yes you can. Here is how to do it: %= application.getInitParameter(SOME_KEY) % Does this actually work? I've had report that in Tomcat and JRUN, it doesn't seem to return the values. Has anyone used it?
Re: Architecture for a form-wizzard
I`m building a site that uses a application form that spans over several pages. How should I design this application? 1. With formBeans for every page 2. One formBean for the entire application form What I find I am doing is building a separate form bean for each page that just represents the data for that page, and having a bean in the session the represents the data for all the pages within that transaction. The each form has a pre-action which populates the form from the session bean, and a post-action which sets data back into the session bean. The post-action then forwards onto the next pre-action for the next form. So something like PreForm1Action - form1.jsp - PostForm1Action - PreForm2Action - form2.jsp - PostForm2Action this seems to work okay for now. -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
Re: is it a bug of CheckboxTag
Set the value to true in the ActionForm reset() method. Another thing is that you shouldn't call .jsp pages directly. They should always be called through an Action first. From the exmaple, the subscription page is called through (I think) editSubscription.do to populate the form, and then forward to the .jsp page. At the top of the .jsp page you can then put something like (and this is from memory, so I'm not sure of the exact syntax) logic:notExists name=editSubscriptionForm logic:forward forward=editSubscription/ /logic:notExists JeanX wrote: Jim Richards wrote at 2001-05-15 14:48:00, JeanX wrote: If I uncheck a checkbox, there is no corresponding parameter pair in request. So I can not get the right status of certain form. How to resolve it ? In you ActionForm reset method you need to set all the checkbox values to false, as reset() is called before the form is populated. That way, the only values that get sent are the ones that need to be set to true, and the ActionForm is correctly set. So if I wanna some checkbox default true, How can I do?
Re: FW: Design Error? Bean repopulation...
I had this problem, and worked out a solution that seemed workable.
The main thing to remember with the html:xxx/ tags is the
the name part should always be the name of the form bean, and
the property be relative to that, otherwise the BeanUtils won't
be able to repopulate the form.
I tend to store my collections in SortedMap's and use the
primary key of the database table as the index into the
SortedMap (This means, each bean has a getId() which is an integer,
and the Bean has the comparable interface to sort them). Then when
I do the iterate, I iterate over the keys of the map and return just
the Map.Entry values. Then when I do the html:xxx/ I have
something like (this if from memory, based on your code, it might
take a few hacks to get right, I think you'll need a useBean for
the retailerForm, can remember)
logic:iterate id=entry collection=%=
retailerForm.getRetailers().getEntries().iterator() %
html:text name=retailerForm property=%= retailerBean[\ +
((Map.Entry) entry).getKey()) + \] %/
%!-- or, if the bean identity id id, and so we use getId)( --%
html:text name=retailerForm property=%= retailerBean[\ +
((RetailerBean) ((Map.Entry) entry).getValue())).getId() + \] %/
/logic:iterate
I use this method all the time, it works great.
Here's another message with a copy of the code I use
http://www.mail-archive.com/struts-user%40jakarta.apache.org/msg07627.html
- I have a number of beans held in a Vector.
- In my ActionForm I have a two getters for this vector - one returns the
vector and one returns elements of the vector - so something like this:
public RetailerBean getRetailerBean( int index )
{
return (RetailerBean)retailers.elementAt( index );
}
and
public Vector getRetailers()
{
return retailers;
}
This works fine to populate using the following type of iterate tag:
logic:iterate id=retailer name=retailerForm property=retailers
tr
tdhtml:checkbox name=retailer property=delete//td
tdhtml:text name=retailer property=name//td
/a/td--
/tr
/logic:iterate
However, it doesn't repopulate the beans when I do a submit. This is
because the generated html is something like:
input type=checkbox name=delete
rather than:
input type=checkbox name=retailer[0].delete
input type=checkbox name=retailer[1].delete
which would allow my beans to be repopulated.
The outcome of this is that I have to create arrays for all properties of my
beans in my ActionForm and take the values from there. This seems like
bodge-heaven and not at all object-oriented.
If the form can be populated directly from the beans then surely it should
work the other way too.
Re: Initialization of Request-Scoped Objects
No. What I do (antd I think it this is from the example app as well, maybe subscription.jsp) is to put at the beginning of the .jsp file (this is from memory, don't have the code handy) logic:notPresent name=yourFormBean logic:forward forward=editFormCode/ /logic:notPresent Where editFormCode is set up as a forward in the struct-config.xml file. This way, if the form bean has not been set up, it forwards the code through the form bean. This technique is useful for things like checking a user bean has been set in the session as well. Jeff Trent wrote: When myAction.do is called, I have a chance within reset() to initialize my request-scoped objects. However, if myAction.jsp is called instead then I can't seem to find a way to get initialized. Are there any methods called in my Action class when the .jsp file is invoked instead of the .do file?
Re: Philosophical question(s) related to STRUTS
As a profession developer I have taken to learning how to use Struts for several reasons (besides being Java, and thus way cool). It's free. It doesn't cost me anything to develop or deploy. Open Source. If the development team die in a freak accident from caffeine overdose at something like JavaOne then I can continue to apply my own patches to it. If Oracle disappeared tomorrow (or Cisco?) then who is going to apply patches, provide support and have further releases? if you end up developing your own framework, that's great but why re-invent the wheel? Does it matter if it doesn't become sanctioned? Your own framework won't become sanctioned? I developed my own framework when Turbine was being developed and discovered that amount of work it took was huge. I could have adopted Turbine an done all the development with that (I didn't because Turbine is huge and I was starting out with JavaServlets, and the best way I learn is to do it myself.) 1) Is SunMicro going to be supporting STRUTS from a financial and/or marketing standpoint ? I only saw one tiny, tiny mention of STRUTS in the JAVAONE outline of presentations I just got in the mail. That bothers me. 2) If the answer to #1 is unknown, will JSP/Java serverside professional developers take the time to learn STRUTS and employee it in their web applications ? 3) If the answer to #1 is unknown, will corporations and consulting firms decide to committ to STRUTS in lieu of their own frameworks and methodology ? I am wading thru the one-inch thick documentation and I must say I am really impressed with the work done so far. However, as a consultant, I must be concerned about spending too much time with this if it is not going to become a popular or sanctioned approach in the webdev marketplace. Thoughts / feedback anyone ?
Re: Forward to an Action
Try forward name=success path=/viewCategory.do/ instead of forward name=success path=/viewCategory/ -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
Re: Struts and DAO pattern. Expensive?
I need to confess I'm lost. The PetStore approach sounds cleaner in some sence, but also sounds too repetite in other, and mostly, sounds way to expensive (or it isn't?). Struts uses a connection pool. So when you do a getConnection() you're actually pulling one out from a pool of shared connection objects, so there is no real problem with generating a new connection. In my application I have manager objects will deal with data beans, and these managers have methods that all look the same (open connection, select data, build bean, close up, return). Yes it's a lot of cut and paste. I'd prefer to have used EJBs but the project size doesn't warrant such expense (time and effort, not money).
Re: pleae help..submit button only allows 1 parameter to be sent with the request
Do you have multiple submit buttons, and your action treats the data differently? You can use hidden fields, so that all the data is sent with the form fields. Then your action can determine which fields to use based on the valud of the submit button. If you have the same input field name and different values then you can either use some funky JavaScript to set the hidden fields values, or give them different names. This second option may break your form beans though. At 11:36 AM 2/05/01 +0200, Dudley Butt@i-Commerce wrote: I want to be able to send more than one parameter when the submit button is clicked. currently i have this... html:submit property=action value=Add --- I want more parameters to be sentis this possible? bean:message key=button.submit/ /html:submit thanx -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
Re: Why isn't ActionsErrors throwable?
Because exceptions are slow, and can only handle one error at a time. The design as it is, allows for multiple action errors to be returned to a page for display. Exceptions should be for catastrophic errors that cause the system to barf. Having a blank password doesn't really fit within that framework. The question is: Why doesn't ActionErrors extends Throwable? Does it make sense to make a change in the framework class org.apache.struts.action.ActionErrors and make it throwable?
Re: struts at webappcabaret
Michael Mok wrote: The servlet engine to use is Tomcat 3.1. Tomcat 3.1 has lots of buggies in it, you're better off using a 3.2.x version or go straight to a 4.0, although I think 4.0 is still in beta.
RE: why does struts use forward instead of redirect
In your code you can, I don't think there is a run-time parameter between them because it would affect code quite drastically if you switched from forwards to redirects. At 08:40 AM 25/04/01 +0100, you wrote: Can you switch between the two, i.e. make struts use redirect? At 11:13 PM 24/04/01 -0400, you wrote: Anyone know why struts uses forward instead of redirect? A redirect sends a request back to the browser, thus generating another connection between the client and server. Forward keeps it internal to the server, which also means you can re-use the request object. -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
Re: why does struts use forward instead of redirect
At 11:13 PM 24/04/01 -0400, you wrote: Anyone know why struts uses forward instead of redirect? A redirect sends a request back to the browser, thus generating another connection between the client and server. Forward keeps it internal to the server, which also means you can re-use the request object. -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
Re: help with JavaScript and text box.
Try document.requisition['requisitionLines.requiredOnDate'].value = "the new value"; Alex Colic wrote: Hi, I have a text box mapped to a bean via: html:text property="requisitionLines.requiredOnDate" size="10" / My form is called requisition. If I try to set the value of the text box via JavaScript I have to access the box via form.object or requisition.requsitionLine.requiredOnDate. JavaScript can't find the box due to the numerous '.'. Is there a way around this to access the box. Any help is appreciated. Regards Alex
Re: servlet context attribute?
"Young, Wayne" wrote:
Ok, that gets me closer. I created the servlet put the following code in
the init() method.
// Store the lookup service in the servlet context
LookupService lookupService = new LookupService();
getServletContext().setAttribute("LookupService",lookupService);
I need to get access to this attribute in a class that has nothing to do
with servlets. Is there a method call I can make?
Not really no, you might want to look up the definition of the
single design pattern. I don't have the references here, but the
basic idea is to not create the object as you done about but have
LookupService lookupService = LookupService.getInstance();
and then
public class LookupService {
private static LookupService instance = null;
public LookupService getInstance() {
if (instance == null)
instance = new LookupService();
return instance;
}
}
so you can grab the objcet any where and be assured there is
only one instance that you are sharing. You of course then need
to make sure that the methods are synchroni(s|z)ed.
Re: Newbie datasource Q
What specific operation are you trying to do? Since you don't have autocommit set to true, you may be having transaction problems. I found it helpful to create a Test.java file with the same operations as creating the datasource (with the same parameters), a connection and doing a sample select statement. It helped to confirm where errors might be happening. JOEL VOGT wrote: later on in a servlet I have javax.sql.DataSource dataSource = servlet.findDataSource(null); con = dataSource.getConnection(); When I try to use this however, I get an java.sql exception : 'no data found'
never ending iterate tag questions
I've now got the logic:iterate ... tag to work, but I'm confused about how it works with the collection attribute. I've seen examples for people where they have logic:iterate id="bean" collection="%= someObject.getValues() %" ... do something with bean ... /logiv:interate but the documentation http://jakarta.apache.org/struts/api/org/apache/struts/taglib/logic/package-summary.html#package_description indicates that it should be used thus: logic:iterate id="bean" collection="% someObject.getValues() %" ... do something with bean ... /logiv:interate The difference here being the "=" inside the scriptlet for the collection attribute. When I don't have the "=" as the documentation suggests, I get Unable to convert a String to java.lang.Object for attribute collectio and if I do use the "=", which seem to be the correct way, looking at the generated Java code, I get the correct response. -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
Re: Confusion : Usage of Struts
What happens to javascript validations? Where do we fit them in the comlete picture? They don't happen at the moment. Part of the Struts 1.1 plan is to look at some forms of autogeneration of JavaScript validation. A few people have implemented their own system for this. Who takes care of whether is it a normal application or struts application. does the container do this ? There is no real difference, struts is normal ;-) When the web container is set up it need to know that a path ending in .do is part of struts, as are .jsp pages. If we have to code the struts-jsp and a corresponding bean for every form then where is the convenience ? Are we not trading one problem with another. I'm not sure what you mean, but there is some shortcut methods you can handle with the form input/output part. If you read the taglib documentation you'll see examples of this. Under what circumstances do I need to extend ActionServlet and use that ? When you need something that it doesn't provide. For example I extend ActionServlet to implement setting up singleton objects for database access. I have a gut feeling that struts application will be slow as it involves one extra component at the server side for every single GUI developed ? Milliseconds slower ;-) Can some body who is deeply involved with struts development reply. I'n not that deeply involved, but they are valid questions. -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
RE: Alternative Frameworks - continued
At 10:25 PM 20/03/01 -0800, you wrote: Eric, Have you looked into ASP.net vs ...? The main problem there is that a lot of the development tools are beta, and finding an ISP that will host it is also difficult. It will be at least 12 months before it has settled down. -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
Re: PostgreSQL problems.
I've finally got the problem sorted out. It was rather obscure, but I think I could have only solved it because of the "Joy of Open Source(tm)". When I built my postgresql.jar file, I only put .class files into it, not realising there were some error_xx.properties file to be put in as well. I was getting an error during ActionServlet loading because there are documented "not supported" errors that would throw an SQLException, but since the error_xx.properties was not there an extra NullPointerException was being thrown (because the ResourceBundle was null, because there were no .proerties file) and this rippled through the system and broke init(). I've rebuilt my postgresql.jar file and the system loads, I can access the database and we'll all live happily ever after. At 11:30 PM 14/03/01 +1100, Jim Richards wrote: At 01:28 PM 14/03/01 +0100, you wrote: We are currently using PostgreSQL as a database in a productive system together with struts. We tried with Interbase before, but ran into many many difficulties. So we switched to PostgreSQL and everything went well from then on. What specific kind of problems do you have? See http://nagoya.betaversion.org/bugzilla/show_bug.cgi?id=827 -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
PostgreSQL problems.
I'm still getting problems with PostgreSQL (documented previously several times). Has anyone used it successfully as a database. Yes, I've got the settings correct, the JDBC string correct, lots of other things. I really need someone who actually has it working to put their hand up and say "yes, it does work, this is how I did it." Thanks. -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
Re: PostgreSQL problems.
At 01:28 PM 14/03/01 +0100, you wrote: We are currently using PostgreSQL as a database in a productive system together with struts. We tried with Interbase before, but ran into many many difficulties. So we switched to PostgreSQL and everything went well from then on. What specific kind of problems do you have? See http://nagoya.betaversion.org/bugzilla/show_bug.cgi?id=827 -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
Re: PostgreSQL problems.
At 01:51 PM 14/03/01 +0100, you wrote: This does not sound at all like a PostgreSQL problem. Doesn't the error also show up, when you remove the datasource section? No. That's the catch, it loads and runs if there's no datasource section in the struts-config.xml file (although you need to comment out the database code, or not access it). -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
Re: Business Object Bean Persistence Strategies
My approach is to have: JSP - Struts Form/Action - Business Rules - Database The busines rules are set up as singleton object on servlet startup. I've subclassed the ActionServlet and overloaded the datasource init method, there I set up the four business rule classes and put them into application scope, on creation the business classes get their datasource reference. The objects I pass around are data beans that model the database closely (similar to entity EJBs). So for logon, I have a User() object where I set the user_id and password from the logon form, then the action class will call User user = UserManager.logon(User user) which will either find the user and return a fully filled user object or throw and exception (I might change that though). UserManager has other methods for logoff, update and create. So the action code doesn't need to send the dataconnection because that ties the Action too closely to the data store (what happens if the authentication changes to LDAP or unix password/username or database password/username?). It also means I can reuse the business rules for another project that might not be struts based. (But whay would I do that ;-) Michael McCallister wrote: What strategies or patterns do people use to manage persistence of business object data in a Struts application when there is no EJB layer and there is a desire to keep the business objects as independent of the web portion of the application as possible? Do you use the Struts DataSource and pass either it or a Connection as a parameter to bean methods that take responsibility for managing persistence? Do you follow the J2EE blueprint and create separate Data Access Objects to support persistence? Are you using an open-source framework to manage persistence? This seems like a common problem, but I haven't seen much talk about common solutions. Mike
Database pooling code
Has anyone got the database pooling code/DataSource to work or specifically had it fail? I've been having problems (documented previously) and I've read about a few others who can't seem to get the pooling to work as well, and I want to get an idea of specifically where the problem might lie in order to fix it. You are the simulacrum of my dreams and the signifier of my desires Mr Grumpy ... come and visit his home! http://www.cyber4.org/members/grumpy/index.html
Re: startup problems
I've now got the beta-1 build and it works fine with the PostgreSQL driver, which I'm very happy about. Mosy things seem to be working fine, I've got my own ActionServlet extending the base one, overloaded the initDataSources() method so I can create my own database singleton objects for reference within each Action, to access the database. All is well. For now. I think what might be useful for me now is getting some time together to read all the new documentation again. Jim Richards wrote: I'm sure earlier today when I started with the new version I didn't get the load error, and something I've done has freaked it out. But as you'll see I haven't done that much to freak it out.
Re: Stupd question about Struts and EJB.
EJBs come in two flavours (as opposed to flavors), session beans and entity beans. Session beans represent business logic and rules, entity beans represent a row in the database. The model that you'd use is to have Actions do the basic processing of form elements into data beans (just regular beans, nothing special), but no business rules. The Action then calls a session bean with the data bean to do something (eg. usdate user details in the database). The session bean connects to the database with an entity bean (or more than one depending on the database structure). Visually, you'd have JSP = Action = Session Bean= Entity Bean = Database Hope that helps. At 06:37 AM 27/02/01 -0500, you wrote: I am in the process of convincing my company to go with MVC and Struts. As with any thing new, people are sometimes resistant to change. One of the opposition questions I am getting is, how will this work with EJB(problem is I don't know too much about EJB currently). Do the EJB's references just go in the Form and Action Bean and every thing else behaves the same? Are there any obvious points that I can bring up on the subject? Any help on this subject would be much appreciated! Thanks, Cameron Ingram -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
Re: more than one formclass for a one jsp ?
At 07:37 AM 23/02/01 -0500, you wrote: Hi, I wanted to have more than one form class for a one jsp file. Because in my one jsp file, i have a information of 4 tables. So i want to transfer the information to 4 formclasses? Is it possible? If what I think you are doing is wanting to have the four form classes handle all the data from a form in one go, no I don't think so. A form can only have one action attribute, so there's no way to split the data appart. You can have four forms within the jsp page, and each one has it's own action, but only one can be used at a time. You won't be able to submit to all four at once. -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
Re: database locks.. (OT)
The usual thing to check for are: Are you using the database conneciton pool? Are you closing your database connection? Do use you transactions at all? Do you commit your work? Do you have autocommit turned on? Anand Raman wrote: I am trying to implement a multi page form and in some scenario the application is hanging because of a database lock which is not going away after some DML.
Re: startup problems
I'm sure earlier today when I started with the new version I didn't get the load error, and something I've done has freaked it out. But as you'll see I haven't done that much to freak it out. I think I'm onto something now. I commented out the database connection code I had: data-sources data-source autoCommit="false" description="Cyberlancers Data Source Configuration" driverClass="org.postgresql.Driver" maxCount="10" minCount="1" user="grumpy" password="grumpy" url="jdbc:postgresql://dbase.home.cyber4.org/cyberlancers" / /data-sources And now it doesn't get the error. Which is really odd because the startup comment with Tomcat indicated it loaded without a problem. The database is running, and the postgresql driver is in the CLASSPATH of tomcat. New org.apache.struts.util.GenericDataSource Set org.apache.struts.util.GenericDataSource properties Call org.apache.struts.action.ActionServlet.addDataSource(org.apache.struts.action.DATA_SOURCE, GenericDataSource[activeCount=0, autoCommit=false, closed=false, description=Cyberlancers Data Source Configuration, driverClass=org.postgresql.Driver, loginTimeout=0, maxCount=10, minCount=1, password=grumpy, readOnly=false, url=jdbc:postgresql://dbase.home.cyber4.org/cyberlancers, useCount=0, user=grumpy]) Pop org.apache.struts.util.GenericDataSource The url and username/password works with the JDBC test that come with PostgreSQL 7.0.3 that I'm using, and other simple tests that I've done. You are the simulacrum of my dreams and the signifier of my desires Mr Grumpy ... come and visit his home! http://www.cyber4.org/members/grumpy/index.html
Re: AW: JDBC Connection Pooling
I'm trying to do the same thing here, but I'm not sure on a few
of the details.
I've got a servlet called DatabaseContainer, which will handle all
the database connections and populating Beans to send back
to the Actions.
What I can't work out is, once I've get the DataSource, how does the
Action know about the DatabaseContainer? At the moment I've
got
getServletContext().setAttribute("DatabaseContainer", this);
and so in the Action, I can call
DatabaseContainer database = (DatabaseContainer)
servlet.getServletContext().getAttribute("DatabaseContainer");
But this doesn't seem right. Since Craig doesn't like singleton objects
hanging around, and I don't feel like I want to extend ActionServlet,
this seems like the only way (since getServlet() is depreciated in the
2.2 spec).
Any ideas?
The problem I'm facing is a bit diffrent as I want to access the
Connection Pool from a servlet that is started upon startup
of the servlet engine (similar to the struts example application).
In order to be able to access the DataSource, I tried to subclass
the ActionServlet class and to access the data source that way.
However, my servlet engine (Tomcat 3.2.1) complains that it is unable
to start this new Action servlet...
Any ideas how to solve this problem?
There are three things you will need to do:
* Make sure that both the Struts controller servlet and your own
servlet are marked load-on-startup.
* Use values for the load-on-startup attribute that cause the
Struts controller servlet to be initialized first (for example, use a
"1" as the value for the controller servlet, and "2" for yours).
I'm not positive that Tomcat 3.2 respects this, but it is supposed to.
* The Struts controller servlet makes the data source it initializes
available as a servlet context attribute, as well as available via a
method call. You can therefore grab it by calling:
DataSource ds = (javax.sql.DataSource)
getServletContext().getAttribute(Action.DATA_SOURCE_KEY);
in any other servlet (or JSP page) that is part of the same web app.
--
Kumera - a new Open Source Content Management System
for small to medium web sites written in Perl and using XML
http://www.cyber4.org/kumera/index.html
Re: Help with Exception
Sorry about not cutting out the previous messages, but the discussion is sort-of relevant to what I'm having problems with. I'm getting javax.servlet.ServletException: Cannot find ActionMappings or ActionFormBeans collection I can run the example, and I can run a test page with nothing but a few tags (eg: html:html) but no forms/beans/etc. And I get the above error. I've checked struts-config.xml. I can only think I've missed something. Have you configured the controller servlet to be load-on-startup? If you do not do this, none of the servlet context attributes created by the controller servlet (including the two mentioned in the exception) will have been created if the first request you submit is to a JSP page. If that's not the issue, what servlet container are you running on. Some have problems with load-on-startup -- check the installation documentation for platform specific details. Yes, sorry. I'm using Tomcat 3.2.1 on Solaris, with J2SE 1.3. I have load-on-startup set to 1 (and my database container set to 2). I get an odd error when Tomcat starts in that it says (and this is from memory, don't have the extact message here) it can't load servlet: action. Now I assume this is the action servlet that is the struts one in the web.xml. But, it was a cut and paste job with modifications from the example web.xml, so I'd expect it to work. The servlet mapping seems correct.
Re: EJB references
You probably need to create a helper class that does all this work, keeps the references as public variables and create it as a singleton in the init of the servlet, and store it as an application wide variable so you can access it easily and have references to all the session beans. I can't remember the deal though with synchronisation and EJBs, you had better check on that as there will be a high possibility of multiple threads (servlets) accessing it at the same time. At 08:37 AM 6/12/00 -0800, you wrote: Hi Jim and Alex, I think both of you understood me wrong. I am not planning to access EJBs from JSP pages at all. I will have a lot of session EJBs to manipulate Entity EJBs. But, the suggested approach accessing them from servlets was: [snip ...] -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
Re: EJB references
I'll be using ELB's for my application when I start building it in a few weeks time (still in the design stage). I expect to be using only EJBs from a action servlet, that are session beans, and those session beans use entity beans. From previous work I found there was too much overhead using the entity beans directly in the generation of a html page, and you rarely needed that direct conncetion to the database anyway. As for the references, I'm not sure what you mean. I'll be putting things like the JNDI and connection information into to web.xml as startup parameters. "Boulatian, Misak" wrote: Hi all, The application we design is quite large. I am still trying to figure out the question on EJB references. From the Craig Mcc.'s comment I could put those in one startup servlet. Is it a viable solution? I am going to have many session references. Where am I going to store those references (in servlet context)? How can I make those available to action classes? Or I can put only the handle in the servlet context and call lookup in every action class (any scalability problems with this approach?). I cannot believe that none of you worked in large applications and will not be using EJBs for business logic and did not go over this kind of issue. I appreciate response, Misak Boulatian
Re: Question about errors tag
[big snip ...] Is there another way to accomplish the same thing, without tacking the error on to the redirect URL? Probably not, no. I can't see where you'd store the errors. In the session would cause problems with synchronisation, and on the redirect URL could cause buffer problems (with more then 1024 bytes of errors). You could store it in the session, if you used a unique identifier to store them, and pass that identifier through to the redirected URL. I'd think you need to look at *why* you don't want the action URL displayed in the first place. You could always be sneaky and use a hidden frame to force the URL to always be the same.
Re: Nesting form inside ifParameterEquals
You can use java script. I posted a message about this to the struts-dev mailing list a few days ago. I don't have the email, but you can find it in the mail archives, which I don't have a reference for either. Nikolaus Rumm wrote: Hello, what I wanted to do is... struts:ifParameterEquals name="action" value="create" struts:form action="/html/user/storeUser.do?action=create" name="userForm" type="com.kephera.scivent.client.participant.UserForm" /struts:ifParameterEquals struts:ifParameterEquals name="action" value="edit" struts:form action="/html/user/storeUser.do?action=edit" name="userForm" type="com.kephera.scivent.client.participant.UserForm" /struts:ifParameterEquals However, under struts 0.5 the tag's body and thus the form tag is not evaluated, but the whole page is skipped. Is there a way to dynamically alter the action of a struts:form tag ? Regards Nikolaus
Re: Article on JavaWorld
If you're prepared to be patient, I need both of these, and can work on them but I won't be able to start unitl January on the development of it. The AutoBean idea (as an idea) is something I use already in Cold Fusion, and have done before in PL/SQL. The client side validator, as I posted before I have a good JavaScript validation library I can use, and am happy for it to be included. "Craig R. McClanahan" wrote: The "AutoBean" concept has been requested several times, often in the guise of "beans with dynamic properties" so that you don't have to create a form bean with individual getters and setters. I'd like to explore this concept in depth in Struts 1.1, which will give us time to make sure that a complete and coherent set of support for such beans can be included (for example, all the custom tags that know how to extract bean properties should now how to extract them from an AutoBean, without the page developer having to do anything). The "Validator" concept is interesting. Besides the server-side format checking that is being done here, I would also like to see the option for the form tags to generate client-side JavaScript code (if requested) to check as many things like this as you can on the client side, to improve the user experience.
Re: Submitting forms to alternate actions
With JavaScript you can specify the form action, so you'd have something like: form name="aForm" action="defaultAction.do" onSubmit="return checkSubmit(this);" input type="submit" name="button1" value="use default action" input type="submit" name="button2" value="use other action" onClick="document.aForm.action='otherAction.do';" /form You can also mess around with the form onSubmit action, but I was having trouble trying to determine which button was clicked through that method. This seemed a lot cleaner. Steven Peterson wrote: Hi all: I have used struts to set up a simple application. My problem comes up when I want to submit a form to different actions, depending on which button the user presses. If the contents of the form were not important, I could simply GET a different URL (each mapped to a different action) for each button. However, when I need the contents of the form so that the form must be POSTed, then I am stuck -- as far as I can tell, there is only one url that a form can be posted to. Is there a way to override the form's coded post action based on which button is pressed? - Steve -- Steven Peterson, President Frontier Productions, Inc. 310 Wesley Drive Chapel Hill, NC 27516 http://www.frontierproductions.net Tel: 919-942-1386 Fax: 919-933-2677
Re: book ideas
I'd prefer (and this is personal preference only) more on the development of models and application frameworks then actual introduction code. Coverage of PostgreSQL as well (it has transaction, which are quite crucial to the EJBs, otherwise you get MySQL's serialised table locking which is not great). Use of EJBs through the MVC framework. Where to use session beans (both types) and when to use entity beans (ie. should entity beans be attached to the session object (no, not really) but only accessed through the session beans, and the session beans return ordinary Java Beans to the servlet. Security context in relation to attaching a profile to the whole application. Not a reprint of the documentation that comes with any of the main downloads, as this just duplicates content and kills more trees. Other EJB/J2EE containers (eg. Enhydra/Jonas). Wait till Tomcat 4.0 is stable and Struts 1.0, otherwise you'll have major dead space in the book. (Although a paper book will always be somewhat out of date) If you focus on frameworks/ideas/concepts rather then deep examples you'll find the book lasts a lot longer, which is why Knuth's books are still valid resources today but my Java 1.0 Beta2 books are wasted space. At 10:59 AM 30/11/00 +0100, you wrote: Hi Everybody, I'm (already) working in a book in JSP and Tag Libraries development for New Riders Publishing. The idea of the book is: - An introduction on Servlets and JSP - Deep in Tag coding - Design and Architectural issues (major challenges when designing web apps with Servlets, JSP, Pet Shop Model 2 approach, Struts Model 2 approach, the role of tags in process encapsulation and reuse, how tags affect system design and how the Servlet/JSP architecture responds to that, etc.). EJBs are treated as subsidiary (calm down!) since it's too broad a subject and very well covered in many other books. - Deep in Struts (a couple of chapters to discuss the model, the usage scenarios and of course the taglibs) - Jakarta Taglibs (a broad coverage) - Commercial taglibs (the commercial case) - Appendixes on TOMCAT (I focus in TOMCAT), MySQL, Caucho Resin and Jrun (examples are provided also on these), jBoss - Many, many, MANY examples on everything. Questions are: 1. What do you think of this coverage ? 2. What would you like to see on paper ? My sources on the design issues are a number of articles (for instance that one from Jason Hunter questioning JSP), some reference books and the lists (not to mention experience, of course). 3. On the side of Struts, what do you like to see there ? What do you think is not yet said or well said about Struts ? 4. And on the design issues on Servlet/JSP architecture(s) ? Thank you all for your attention on this, Wellington Silva UN/FAO -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
Re: HttpSession.getAttribute() missing?
Then you've got the wrong .jar file in your classpath, or earlier in your classpath. You can either edit your CLASSPATH environment variable setting, or remove the offending .jar file from your system. You'll then need to include the correct .jar file (depends on your server setup) that includes the latest classes. What server are you using? The HttpSession it's going after is in the Jswdk version 1.0.1.Could it mean putValue() instead?
Re: need some explanations
It's really only for an example. It you want real object persistence, you'll need to develop your owne layer. Something like EJB's will make this much, much easier. Or some direct JDBC code perhaps. Vilavanh Messien wrote: I analyzed the struts example in order to develop my own small application. Unfortunatelaty, I don't really understand how the user's information are saved in the "database.xml" file. The DatabaseServlet must write the information in the "database.xml" when the destroy() method is called. My question is how to make the destroy() method perform ? How can I do to pass some arguments to the DatabaseServlet ? Thanks Vilavanh
Re: Struts + Design Philosophy
users. What I am asking is, will STRUTS gracefully support a 2 stage approach to forms - 1 stage to construct and fill in values, and a second one to gather responses and process ? This is something I am looking at, have a look at the simple JavaScript validation page I posted yesterday (well, yesterday my time). I think we can improve on the ActionForm.validate() style of doing things.
Re: thought on design
In 1.0, the validation model for the server side has been significantly enhanced. The idea of auto-generating JavaScript validation code for the things you can check on the client side is attractive, but won't make Struts 1.0 (unless someone really hurries and creates it :-). Hmm ... I am using a nightly build from about a week ago, I'll have to dig into it to see how this enhanced validation works (or is it not in there yet). I would caution you that, even if you implement client-side JavaScript validations, your server side logic should not rely on them. After all, the user can easily turn off JavaScript, and/or write an application that simulates a browser's interactions with the server but tries to bypass your edits. The client side validation is only for simple things, that validate() would do, such as checking for number format, empty strings, valida credit card number, post code and so on. The action bean would still need to check that, say a credit card was valid (rather then just a number). Also, I couldn't work out how to get the request attributes to be sticky between the next and previous pages (think CGI.pm and it's param() functions). It's probably something to do with the response, but I don't have the docs handy. Sticky checkboxes have been fixed in 1.0 by adding a reset() method to the ActionForm class. You should reset any boolean properties to false in the reset() method, which is called before the property populator -- which will set the checkbox back to true if it was indeed checked. Hmm, I was thinking more about text fields and so on between multipage forms, not rebuilding the current form with the previous data. I will take a look at your example, but there is a subtle philosophical issue that has affected the way ActionForm beans are specified in Struts. Basically, an ActionForm is part of the presentation layer, not the model layer. Their primary purpose is to encapsulate the current set of inputs that a user last entered -- whether or not they are valid -- so that you can create an error page that faithfully represents what *was* entered, and points out what needs to be changed (just like GUI apps do). Otherwise, you're going to force the user to have to re-enter some of the fields, and that will not make you a popular person with your user community. I suppose what I'd like is rather then having validate() being called, when each property has it's setXXX method called, the setXXX method does the simple validation, and throws a ProperyVetoException with the error message text as part of the Exception. Then if a try block cathces this, stacks up the error messages and calls the page again with these errors. It seems that from what I've seen, most validate() methods are the same.
Re: Pages being cached
(This has been cross posted to tomcat-user, as it seems to be a tomcat issue rather then a struts issue.) I have found that this behaviour is *not* replicated when using Tomcat's built in HTTP server. I need to do some deep experimentation, and install tomcat-3.2b8 as well. I'll let you all know how it goes. I'm getting weird behaviour with my struts code, and with the example application. Once during testing I noticed that pages seemed to be served at random.. Then I shut down httpd and then shut down tomcat. Thereafter I restarted tomcat and then restarted htttpd. Then everything worked. So I've concluded that when I want to restart tomcat, I've to first stop httpd, then tomcat, then start tomcat and then start httpd. I'm not sure if I should restart tomcat when I just want to restart httpd. I have to restart Tomcat and Apache to get anything to work. It doesn't help. Once my application has compiled the three pages it has in it, (that is viewed them) it starts freaking out. Going through the documentation doesn't seem a problem, so long as that's all I do. It happens with the example application, with everything. -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
Pages being cached
I'm getting weird behaviour with my struts code, and with the example application. Currently I've got Tomcat 3.2b6 running, with the latest nightly build of struts (17-11-2000, that's the 17th November, 2000) and the examples, documentation and test data all run. (Which I must say, took a very long time to get a configuration other then a standard install right.) JDK 1.3, Solaris7 ... all the latest patches. (On a side note, I don't think tomcat loads .war files unless then are in the $TOMCAT/webapps directory and the AutoLoad interceptor is turned on.) I have virtual hosts set up, and contexts usually set as something sensible (usually /context). I've got Apache web server and mod_jk. I've found that after loading about 2 or 3 pages from the example, or my own (three page) application, pages start coming out of order, just by clicking on links or submitting forms. The weirdest behaviour is when I might be viewing the documentation on one virtual host and context, and a page from my example application (on a different virtual host and context) will appear in its place. Or the example application will appear instead of my application. It's not the browser because I have the cache set to zero, and for it to compare the document on the network to local every time. Anyone seen similar? -- Kumera - a new Open Source Content Management System for small to medium web sites written in Perl and using XML http://www.cyber4.org/kumera/index.html
