Up until now, when I needed to update the content of any pages, I have always had to update the source code directly and re-deploy the whole application. Right now, I want to implement a feature such that I can update the content of any HTML pages dynamically without having to re-deploy the application.
I tried to implement the feature with PrimeFaces's <p:editor> component but it does not work. To be more precise, my functions can correctly update the required page. When I goes to the source code folder, I can actually see my changes. However, subsequent requests for the page still render the old content.
I'd be very grateful if you could show me what I have done wrong. I'd also appreciate it very much if you could show me any other ways to achieve the same goal.
I think you are editing your work-space from your deployment. :)
You have 2 places with the code. One is deployed, and the other in your "working space".
First, it sounds to me like you want your working space to be the deployment. This way whenever you are editing something, you will be changing the deployment directly. For that, simply create a new project in your IDE and point it to the deployment folder.
I bet that :
C:\\Users\\James\\Documents\\NetBeansProjects\\MyProject\\MyProject-war\\web\\
points to your work-space and not the deployment. so effectively, your deployment is editing your work-space.
I think you are looking for this one:
FacesContext.getCurrentInstance().getExternalContext().getRealPath("/")
and if you want the location of the WEB-INF
use the following
String fullpath = FacesContext.getCurrentInstance().getExternalContext().getRealPath("/")+File.separator+"WEB-INF";
and so on...
My code actually was working perfectly. From the above answer of user1068746, I did some research and found this article. The solution is very simple: creating a virtual directory mapping to my hard-disk's directory. As a result, any updates to my files on the hard-disk will immediately be visible to future requests.
Related
It is as simple as an application that can update itself.
More specific, maybe a launcher that loads a jar file from the web and replaces it with the current jar.
The approach that I want is:
A launcher which is coded in Java.
That launcher overall uses classloader e.t.c.
I want to keep the data from the previous jar. Maybe the data from the previous jar are written on a .txt file and pass them to the new one? (If possible)
I have no clue about classloader or any direction to follow.
So my question is "How can i use classloader to create an auto-update Java application?".
P.S. I want java web start to be the last choice. Some lines of code would help me very much.
Edit: I understand completely what are you saying in theory. What i am looking for is specific links or code which i can take as reference.
The question isn't a duplicate because i can't find links or code but theory.
You can use the URLClassLoader to load a jar file from an URL. Mind, URLs may also be local files, and I'd recommend that you download the jar file locally before loading it in, instead of loading it directly from the web. The reason for this is that you want the application to remain usable, even if the user cannot reach your server.
However, I'd also recommend against taking this approach. What if you want to update the launcher? It's better to download a seperate updater application from the main application, launch the seperate updater application, have that kill the main application and update it.
My JSP project is the back-end of a fairly simple site with the purpose to show many submissions which I want to present on the website. They are organized in categories, basically similar to a typical forum.
The content is loaded entirely from a database since making separate files for everything would be extremely redundant.
However, I want to give the users the possibility to navigate properly on my site and also give unique links to each submission.
So for example a link can be: site.com/category1/subcategory2/submission3.jsp
I know how to generate those links, but is there a way to automatically redirect all the theoretically possible links to the main site.com/index.jsp ?
The Java code of the JSP needs access to the original link of course.
Hope someone has an idea..
Big thanks in advance! :)
Alright, in case someone stumbles across this one day...
The way I've been able to solve this was by using a Servlet. Eclipse allows their creation directly in the project and the wizard even allows you to set the url-mapping, for example /main/* so you don't have to mess with the web.xml yourself.
The doGet function simply contains the redirection as follows:
request.getRequestDispatcher("/index.jsp").forward(request,response);
This kind of redirection unfortunately causes all relative links in the webpage to fail. This can be solved by hardlinking to the root directory for example though. See the neat responses here for alternatives: Browser can't access/find relative resources like CSS, images and links when calling a Servlet which forwards to a JSP
I’m coding my first java Desktop application using eclipse and I’m having difficulty deploying it. My project uses JavaFX2 and the e(fx)clipse plugin, the latter is in charge of generating the build.xml file.
ABOUT THE APP
The app, amongst other things, provides an interface where the user can create categories and associate these with labels. All modifications are saved within a single file (the data is stored as a serialized object.) and are supposed to be loaded automatically when the app is reopened.
THE PROBLEM
When build as an executable jar using a .xml file (Ant), the project runs fine within the folder where it gets created. I can run my application, modify data, and save everything once I’m done. When I reopen the app everything gets loaded as it should.
However, if I copy the folder elsewhere, I can no longer save any data. Everything else seems to work; the app will even load the data that was saved when it was in its original directory. I assume that this means that the app can still see the data file, but can no longer write to it.
WHAT IVE TRIED
I’ve read that warping the .jar around an installer may fix the issue; however, one of the goals for the app was to make it as portable as possible. Meaning that it should be possible to move it around from one directory/computer to another, ideally in a manner that is cross-platform friendly, without the need of installing it.
I’ve tried various things to get it to work. I’ve shifted the whole project to Netbeans (to produce a different build), I’ve modified the save/load method file path to make sure the right document is targeted, I’ve tweaked the .xml file the best I could, and I even tried to build the project using the javafxpackager. No matter what I do, when the build works, I get the same results.
Right now, I’m thinking that there may still be something wrong with the .xml file but I’ve got a hard time understanding how to modify it. Perhaps the problem is somehow caused by the way the data gets serialized. I know that at one point when I moved things around within my project, both the save and load methods could no longer interact with the data.
What I find strange is that when the project is moved the load method still works. If the problem is caused by changing the file path, how come only the save method ceases to function?
APP FILES AND STRUCTURE
+src
-(Main.java)
+controller
-(misc.javas)
+modelData
-(Library.java) -->the object that is serialized
+modelLogic
-(misc.javas)
+view
-(misc.javas)
+files
-(library.data) -->the file where the serialized object is saved
+lib
-(empty.empty)
So, is their anything I can do to solve my problem?
Thanks in advance.
In the end, the problem was rather simple. When I was moving the app, I was always putting it on the desktop, which, in the case of my PC, sets by default all content as read-only (in relation to the app's privileges). Because of this, the app could not modify any files that were within folders on the desktop.
Therefore, all I had to do was to move the app to another directory, such as C:\randomFolder, and the problem was solved.
So, if anyone has a similar problem, moving the app elsewhere may be the solution. Alternatively, taking full ownership of the folder and its content can also work.
When possible, it is usually better to have the app ask its user for a specific location to save its data.
I want to get the absolute path of my app and I'm using the code below:
String pathToSave = FacesContext.getCurrentInstance()
.getExternalContext().getRealPath("/");
the result is:
/home/ronaldo/workspace/.metadata/.plugins/org.eclipse.wst.server.core/tmp0/wtpwebapps/Odontonew/
But I'm waiting for something like: /home/ronaldo/workspace/Odontonew/
What is wrong ?
ExternalContext#getRealPath() returns the path relative to where the webapp is been deployed, not where the webapp is been developed orso, as you incorrectly seemed to expect.
But, the variable name pathToSave indicates a much bigger problem: you seem to intend to save files in there. This is a bad idea for the reasons mentioned in the following answer: Uploaded image only available after refreshing the page. In a nutshell, save them to a fixed local disk file system path instead. Note that some servers offer ways to make this step easier configurable, e.g. JBoss.
Just stop using getRealPath(). You never need it in real world. In the 10 years I developed Java web applications (it's under the covers coming from ServletContext#getRealPath()), that method was never been useful for anything. Just ignore that method altogether.
See also:
What does servletcontext.getRealPath("/") mean and when should I use it
Why do I get:
java.lang.NullPointerException org.owasp.csrfguard.CsrfGuardFilter.doFilter(CsrfGuardFilter.java:53)
when trying to run the 0wasp.CsrfGuard.Test example app on Tomcat 5.5?
I'm using the example app exactly as it came in the download but with the following changes to get around listener errors:
I downloaded the CsrfGuard 3.0.0 package and copied the Owasp.CsrfGuard.jar to WEB-INF/lib to get around the listeners not being found.
I changed the web.xml file to use the listener that comes with the Owasp.CsrfGuard.jar rather than the two listeners in the original web.xml, which don't exist.
When I start the app there are no errors but when I point my browser to the /Owasp.CsrfGuard.Test/index.html page, I get the above error.
Any ideas to try?
Update...
Several hours later and I got it working without error. I don't understand why but the filter errors seem to stick around even after stopping and restarting a webapp (in Tomcat at least). I had to rename that webapp's directory, wait 10 seconds for Tomcat to auto-undeploy it, rename it back to its original directory name, and wait 10 seconds for Tomcat to recognize it as a new webapp. I also removed the "x-requested-with" init-param from the web.xml file since I don't use AJAX. Don't know if that really needed to come out but it works so I'm leaving it out.
Now, the "protect.html" file is indeed protected but it is protected too well. I can't get to it at all. If I try to go directly to the protect.html page, it should not allow that, which it doesn't. However, it should allow me to get there through the proper webpage flow with all the pages getting the proper token(s).
The question now is how can I get to the protect.html page?
Thanks.
Got it! In case someone else has problems, here is what I figured out to get JavaScript DOM Manipulation working.
All of the references in the .properties file MUST BE absolute with reference to the web container (Tomcat in my case). Even if you place the .js and .jar files in a specific webapp and modify that webapp's web.xml file with the proper settings for CSRFGuard (with relative references by the way), the entries in the .properties file must include references to the webapp name. For example, it must be something like:
org.owasp.csrfguard.unprotected.Index=/someapp/index.html
rather than just
org.owasp.csrfguard.unprotected.Index=index.html
This is not intuitive since unless you are dealing with the entire Tomcat container (i.e. the master web.xml file) you can generally drop the webapp name since it is implied for that webapp.
Now I'm going to look into the other CSRFGuard method of JSP Tag Injection since the JavaScript mechanism won't work if the user has JavaScript turned off in their browser.