I followed the Watching a Directory for Changes Java7 nio2 tutorial to recursively monitor the entire contents of a directory using the code sample WatchDir.java.
The code looks like this:
// Get list of events for the watch key.
for (WatchEvent<?> event : key.pollEvents()) {
// This key is registered only for ENTRY_CREATE events, but an OVERFLOW event
// can occur regardless if events are lost or discarded.
if (event.kind() == OVERFLOW) {
continue;
}
// Context for directory entry event is the file name of entry.
@SuppressWarnings("unchecked")
WatchEvent<Path> ev = (WatchEvent<Path>)event;
Path fileName = ev.context();
Path fullPath = dir.resolve(fileName);
try {
// Print out event.
System.out.print("Processing file: " + fileName);
processed = fileProcessor.processFile(fullPath);
System.out.println("Processed = " + processed);
if (processed) {
// Print out event.
System.out.println(" - Done!");
}
}
catch (FileNotFoundException e) {
System.err.println("Error message: " + e.getMessage());
}
catch (IOException e) {
System.err.println("Error processing file: " + fileName.toString());
System.err.println("Error message: " + e.getMessage());
}
Ok, so the problem (where I am sure doing something stupid) is here:
processed = fileProcessor.processFile(fullPath);
And what it does is something like this:
public synchronized boolean processFile(Path fullPath) throws IOException {
String line;
String[] tokens;
String fileName = fullPath.getFileName().toString();
String fullPathFileName = fullPath.toString();
// Create the file.
File sourceFile = new File(fullPath.toString());
// If the file does not exist, print out an error message and return.
if (sourceFile.exists() == false) {
System.err.println("ERROR: " + fullPathFileName + ": No such file");
return false;
}
// Check file extension.
if (!getFileExtension(fullPathFileName).equalsIgnoreCase("dat")) {
System.out.println(" - Ignored.");
return false;
}
// Process source file.
try (BufferedReader bReader = new BufferedReader(new FileReader(sourceFile))) {
int type;
// Process each line of the file.
while (bReader.ready()) {
// Get a single line.
line = bReader.readLine();
// Get line tokens.
tokens = line.split(delimiter);
// Get type.
type = Integer.parseInt(tokens[0]);
switch (type) {
// Type 1 = Salesman.
case 1:
-> Call static method to process tokes.
break;
// Type 2 = Customer.
case 2:
-> Call static method to process tokes.
break;
// Type 3 = Sales.
case 3:
-> Call static method to process tokes.
break;
// Other types are unknown!
default:
System.err.println("Unknown type: " + type);
break;
}
}
PrintStream ps = null;
try {
// Write output file.
// Doesn't matter.
}
finally {
if (ps != null) {
ps.close();
}
}
return true;
}
}
The first time I handle an event, it all works fine! Even if there is more than 1 file to process. But in the consecutive times, I get this error message:
The process cannot access the file because it is being used by another process
What am I doing wrong here? What can I do to process the consecutive files with sucess?
Two important notes that I forgot to mention:
EDIT: If I add a sleep before I try to use the file, it works:
Thread.sleep(500);
// Process source file.
try (BufferedReader bReader = new BufferedReader(new FileReader(sourceFile))) {
So, is it possible to be Windows that is not unlocking the files in time? How can I fix that (in a proper way)?
Ok, I found a solution. I don't know if it is the best way to do this, but it works. Unfortunately, file.canRead() and file.canWrite() both return true, even if the file still locked by Windows. So I discovered that if I try to "rename" it with the same name, I know if Windows is working on it or not. So this is what I did:
while(!sourceFile.renameTo(sourceFile)) {
// Cannot read from file, windows still working on it.
Thread.sleep(10);
}