Out of Memory error?

daniel wrote:I was talking to Sadia on our way to Sheridan about the case in which we validate the input URL, instead of parsing through all of the URLs.

nested try catch is legal, HOWEVER, it's not recommended in term of style, especially when it is part of the functionality of the program.

The reason for this is the following. Whenever/everytime an error is caught, the program creates a dump file (a file containing the entire state of the program) for debugging purposes. Normally, this wouldn't affect us, since our dump file is less than a few KB big. however when we program bigger programs, like photoshop big, it can get up to 500MB. This accumulates in the computer, and the user will not be happy lol. Furthermore, it prints a stack trace to the console, which overshadows all of the console output we may output.

The way we can validate invalid URLs (by invalid we mean malformed, no connection etc), is to do the following:

In the FMLinkFinder, I noticed that the constructor does the following:

Code:


Document mainPage = Jsoup.connect(currURL).get();
link = mainpage.select("a[href]");


what we can do is this:

1. Create a global private Connection conn; in FMLinkFinder.
2. modify the constructor to the following:

Code:


conn = Jsoup.connect(currURL);
if (conn != null) { // We need to find out if indeed, Jsoup.connect(URL) returns null IF no connection can be made
    Document mainPage = conn.get();
    link = mainpage.select("a[href]");
}


Create a GET method for the connection object:

Code:


public Connection getConnection() {
    return conn;
}


in the UI:

instead of the try catch block, try this:

Code:


if (fm.getConnection() == null) {
    statusText.setText("[error message]");
} else {
//do whatever is currently coded



We can apply this exact concept to all of the errors Lenard wrote, in order to prevent error.

=)