How do I block the current thread until OnComplete has finished executing without the use of traditional threading primitives? - system.reactive

How do I block the current thread until the OnComplete handler of my observer has finished, without the use of threading primitives?
Here is my code. I want that the Console.WriteLine("Press... statement should be executed only after the OnComplete handler, namely ResetCount has finished executing.
class Program
{
private static long totalItemCount = 0;
private static long listCount = 0;
static void Main()
{
Console.WriteLine($"Starting Main on Thread {Thread.CurrentThread.ManagedThreadId}\n");
var o = Observable.Timer(TimeSpan.FromSeconds(3), TimeSpan.FromSeconds(1))
.Take(20)
.Concat(Observable.Interval(TimeSpan.FromSeconds(0.01)).Take(200))
.Buffer(TimeSpan.FromSeconds(5), TimeSpan.FromSeconds(5));
o.Subscribe(Print, onCompleted: ResetCount);
// How I make sure this line appears only after the OnComplete has fired?
// Do I have to use traditional threading primitives such as wait handles?
// Or just cause the main thread to sleep long enough? That doesn't seem right.
Console.WriteLine("\nPress any key to exit...");
Console.ReadKey();
}
private static void ResetCount()
{
if (listCount > 0)
{
Console.WriteLine($"{totalItemCount} items processed in {listCount} lists.");
}
else
{
Console.WriteLine($"{totalItemCount} items processed.");
}
Interlocked.Exchange(ref totalItemCount, 0);
Interlocked.Exchange(ref listCount, 0);
}
static void Print<T>(T value)
{
var threadType = Thread.CurrentThread.IsBackground ? "Background" : "Foreground";
if (value is IList)
{
var list = value as IList;
Console.WriteLine($"{list.Count} items in list #{Interlocked.Increment(ref listCount)}:");
foreach (var item in list)
{
Console.WriteLine($"{item.ToString()}, ({threadType} #{Thread.CurrentThread.ManagedThreadId}), Item #{Interlocked.Increment(ref totalItemCount)}");
}
Console.WriteLine();
}
else
{
Console.WriteLine($"{value.ToString()}, ({threadType} #{Thread.CurrentThread.ManagedThreadId}), Item #{Interlocked.Increment(ref totalItemCount)}");
}
}
}

On Rx we have specific schedulers to handle threading, synchronization and related.
You can read more about that here:
http://www.introtorx.com/content/v1.0.10621.0/15_SchedulingAndThreading.html
But basically what you're looking for is changing this line:
.Buffer(TimeSpan.FromSeconds(5), TimeSpan.FromSeconds(5), Scheduler.CurrentThread);
They're several ways to test/validate a Rx query. Keep in mind that wouldn't be the answer for all the problems.

Related

How do I update a gtk listbox from an async method?

So when writing UI in GTK it's generally preferrable to handle reading of files, etc. in an Async Method. things such as listboxes, are generally bound to a ListModel, the items in the ListBox updated in accordance with the items_changed signal.
So if I have some class, that implements ListModel, and has an add function, and some FileReader that holds a reference to said ListModel, and call add from an async function, how do i make that in essence triggering the items_changed and having GTK update accordingly?
I've tried list.items_changed.connect(message("Items changed!")); but it never triggers.
I saw this: How can one update GTK+ UI in Vala from a long operation without blocking the UI
but in this example, it's just the button label that is changed, no signal is actually triggered.
EDIT: (Codesample added at the request of #Michael Gratton
//Disclaimer: everything here is still very much a work in progress, and will, as soon as I'm confident that what I have is not total crap, be released under some GPL or other open license.
//Note: for the sake of readability, I adopted the C# naming convention for interfaces, namely, putting a capital 'I' in front of them, a decision i do not feel quite as confident in as I did earlier.
//Note: the calls to message(..) was put in here to help debugging
public class AsyncFileContext : Object{
private int64 offset;
private bool start_read;
private bool read_to_end;
private Factories.IVCardFactory factory;
private File file;
private FileMonitor monitor;
private Gee.Set<IVCard> vcard_buffer;
private IObservableSet<IVCard> _vCards;
public IObservableSet<IVCard> vCards {
owned get{
return this._vCards;
}
}
construct{
//We want to start fileops at the beginning of the file
this.offset = (int64)0;
this.start_read = true;
this.read_to_end = false;
this.vcard_buffer = new Gee.HashSet<IVCard>();
this.factory = new Factories.GenericVCardFactory();
}
public void add_vcard(IVCard card){
//TODO: implement
}
public AsyncFileContext(IObservableSet<IVCard> vcards, string path){
this._vCards = vcards;
this._vCards = IObservableSet.wrap_set<IVCard>(new Gee.HashSet<IVCard>());
this.file = File.new_for_path(path);
this.monitor = file.monitor_file(FileMonitorFlags.NONE, null);
message("1");
//TODO: add connect
this.monitor.changed.connect((file, otherfile, event) => {
if(event != FileMonitorEvent.DELETED){
bool changes_done = event == FileMonitorEvent.CHANGES_DONE_HINT;
Idle.add(() => {
read_file_async.begin(changes_done);
return false;
});
}
});
message("2");
//We don't know that changes are done yet
//TODO: Consider carefully how you want this to work when it is NOT called from an event
Idle.add(() => {
read_file_async.begin(false);
return false;
});
}
//Changes done should only be true if the FileMonitorEvent that triggers the call was CHANGES_DONE_HINT
private async void read_file_async(bool changes_done) throws IOError{
if(!this.start_read){
return;
}
this.start_read = false;
var dis = new DataInputStream(yield file.read_async());
message("3");
//If we've been reading this file, and there's then a change, we assume we need to continue where we let off
//TODO: assert that the offset isn't at the very end of the file, if so reset to 0 so we can reread the file
if(offset > 0){
dis.seek(offset, SeekType.SET);
}
string line;
int vcards_added = 0;
while((line = yield dis.read_line_async()) != null){
message("position: %s".printf(dis.tell().to_string()));
this.offset = dis.tell();
message("4");
message(line);
//if the line is empty, we want to jump to next line, and ignore the input here entirely
if(line.chomp().chug() == ""){
continue;
}
this.factory.add_line(line);
if(factory.vcard_ready){
message("creating...");
this.vcard_buffer.add(factory.create());
vcards_added++;
//If we've read-in and created an entire vcard, it's time to yield
message("Yielding...");
Idle.add(() => {
_vCards.add_all(vcard_buffer);
vcard_buffer.remove_all(_vCards);
return false;
});
Idle.add(read_file_async.callback);
yield;
message("Resuming");
}
}
//IF we expect there will be no more writing, or if we expect that we read ALL the vcards, and did not add any, it's time to go back and read through the whole thing again.
if(changes_done){ //|| vcards_added == 0){
this.offset = 0;
}
this.start_read = true;
}
}
//The main idea in this class is to just bind the IObservableCollection's item_added, item_removed and cleared signals to the items_changed of the ListModel. IObservableCollection is a class I have implemented that merely wraps Gee.Collection, it is unittested, and works as intended
public class VCardListModel : ListModel, Object{
private Gee.List<IVCard> vcard_list;
private IObservableCollection<IVCard> vcard_collection;
public VCardListModel(IObservableCollection<IVCard> vcard_collection){
this.vcard_collection = vcard_collection;
this.vcard_list = new Gee.ArrayList<IVCard>.wrap(vcard_collection.to_array());
this.vcard_collection.item_added.connect((vcard) => {
vcard_list.add(vcard);
int pos = vcard_list.index_of(vcard);
items_changed(pos, 0, 1);
});
this.vcard_collection.item_removed.connect((vcard) => {
int pos = vcard_list.index_of(vcard);
vcard_list.remove(vcard);
items_changed(pos, 1, 0);
});
this.vcard_collection.cleared.connect(() => {
items_changed(0, vcard_list.size, 0);
vcard_list.clear();
});
}
public Object? get_item(uint position){
if((vcard_list.size - 1) < position){
return null;
}
return this.vcard_list.get((int)position);
}
public Type get_item_type(){
return Type.from_name("VikingvCardIVCard");
}
public uint get_n_items(){
return (uint)this.vcard_list.size;
}
public Object? get_object(uint position){
return this.get_item((int)position);
}
}
//The IObservableCollection parsed to this classes constructor, is the one from the AsyncFileContext
public class ContactList : Gtk.ListBox{
private ListModel list_model;
public ContactList(IObservableCollection<IVCard> ivcards){
this.list_model = new VCardListModel(ivcards);
bind_model(this.list_model, create_row_func);
list_model.items_changed.connect(() => {
message("Items Changed!");
base.show_all();
});
}
private Gtk.Widget create_row_func(Object item){
return new ContactRow((IVCard)item);
}
}
Heres the way i 'solved' it.
I'm not particularly proud of this solution, but there are a couple of awful things about the Gtk ListBox, one of them being (and this might really be more of a ListModel issue) if the ListBox is bound to a ListModel, the ListBox will NOT be sortable by using the sort method, and to me at least, that is a dealbreaker. I've solved it by making a class which is basically a List wrapper, which has an 'added' signal and a 'remove' signal. Upon adding an element to the list, the added signal is then wired, so it will create a new Row object and add it to the list box. That way, data is control in a manner Similar to ListModel binding. I can not make it work without calling the ShowAll method though.
private IObservableCollection<IVCard> _ivcards;
public IObservableCollection<IVCard> ivcards {
get{
return _ivcards;
}
set{
this._ivcards = value;
foreach(var card in this._ivcards){
base.prepend(new ContactRow(card));
}
this._ivcards.item_added.connect((item) => {
base.add(new ContactRow(item));
base.show_all();
});
base.show_all();
}
}
Even though this is by no means the best code I've come up with, it works very well.

checking paragraph property loses the selection

in my vsto addin i have some simple code on a timer
private void MainTimer_Elapsed(object sender, System.Timers.ElapsedEventArgs e)
{
if (!dialogopen & Application.Documents.Count > 0)
{
var doc = Application.ActiveDocument;
Word.InlineShapes shps;
Word.Paragraphs pars;
try
{
pars = doc.Paragraphs;
}
catch (Exception)
{
return;
}
var pars2 = pars.Cast<Word.Paragraph>().ToList();
foreach (var obj in pars2)
{
if (obj.OutlineLevel == Word.WdOutlineLevel.wdOutlineLevelBodyText )//PROBLEM HERE
{
};
}
}
}
as soon as it reaches the line that checks the outlinelevel, even if i dont do a thing, the selection in the activedocument gets lost
of course the user cant use a plugin which keeps on canceling his selection...
googling didnt give me a thing
thanks
EDIT1
I tried making a static function for checking the styles, but it did not help. Here's the code
static public class Helpers
{
static public Word.Paragraph checkPars(List<Word.Paragraph> pars)
{
return pars.FirstOrDefault();//(x) => x.OutlineLevel == Word.WdOutlineLevel.wdOutlineLevelBodyText);
}
}
As you can see, I had to remove the actual check, since it was causing the cursor to blink and lose selection
please advise
We use the Application.ScreenUpdating = true; and this keep the selection for all properties in Paragraph except for the Range property.
Then, we tried to access the range via Reflection and this was the solution.
Range rng = (Range)typeof(Paragraph).GetProperty("Range").GetValue(comObj);
We tried to eliminate querying ActiveDocument thinking that that might have had side-effects that were causing the problem but that was not the case.
Then, we confirmed that the selection was not "lost" and screen-updating is the only problem, so we tried restoring the UI with Application.ScreenRefresh and while it did work, it causes the screen to flicker every time the timer fires and this is not good enough.
Finally, knowing that it's only a UI problem, I tried simply switching off Application.ScreenUpdating...
in ThisAddin
private void ThisAddIn_Startup(object sender, System.EventArgs e)
{
Timer timer = new Timer(2000);
timer.Elapsed += (s, t) =>
{
var scrnUpdating = Application.ScreenUpdating;
Application.ScreenUpdating = false;
MainTimer.onElapsed(Application, t);
if (scrnUpdating)
Application.ScreenUpdating = true;
};
timer.Start();
}
In another class library (note that it's static, I still think this is the best way)...
public static class MainTimer
{
public static void onElapsed (Word.Application state, System.Timers.ElapsedEventArgs e)
{
if (state.Documents.Count > 0)
{
var doc = state.ActiveDocument;
Word.InlineShapes shps;
Word.Paragraphs pars;
try
{
pars = doc.Paragraphs;
}
catch (Exception)
{
return;
}
var pars2 = pars.Cast<Word.Paragraph>()
.Where(p => p.OutlineLevel == Word.WdOutlineLevel.wdOutlineLevelBodyText)
.Select(p => p) // do stuff with the selected parragraphs...
.ToList();
}
}
}
And this works for me.
The selection remains and is displayed in the UI without flicker.
I also eliminated some premature enumeration from your code: you don't meed the .ToList() before the foreach.

Powershell monitoring for file system changes [duplicate]

I have an application where I am looking for a text file and if there are any changes made to the file I am using the OnChanged eventhandler to handle the event. I am using the NotifyFilters.LastWriteTime but still the event is getting fired twice. Here is the code.
public void Initialize()
{
FileSystemWatcher _fileWatcher = new FileSystemWatcher();
_fileWatcher.Path = "C:\\Folder";
_fileWatcher.NotifyFilter = NotifyFilters.LastWrite;
_fileWatcher.Filter = "Version.txt";
_fileWatcher.Changed += new FileSystemEventHandler(OnChanged);
_fileWatcher.EnableRaisingEvents = true;
}
private void OnChanged(object source, FileSystemEventArgs e)
{
.......
}
In my case the OnChanged is called twice, when I change the text file version.txt and save it.
I am afraid that this is a well-known bug/feature of the FileSystemWatcher class. This is from the documentation of the class:
You may notice in certain situations that a single creation event generates multiple Created events that are handled by your component. For example, if you use a FileSystemWatcher component to monitor the creation of new files in a directory, and then test it by using Notepad to create a file, you may see two Created events generated even though only a single file was created. This is because Notepad performs multiple file system actions during the writing process. Notepad writes to the disk in batches that create the content of the file and then the file attributes. Other applications may perform in the same manner. Because FileSystemWatcher monitors the operating system activities, all events that these applications fire will be picked up.
Now this bit of text is about the Created event, but the same thing applies to other file events as well. In some applications you might be able to get around this by using the NotifyFilter property, but my experience is says that sometimes you have to do some manual duplicate filtering (hacks) as well.
A while ago I bookedmarked a page with a few FileSystemWatcher tips. You might want to check it out.
I've "fixed" that problem using the following strategy in my delegate:
// fsw_ is the FileSystemWatcher instance used by my application.
private void OnDirectoryChanged(...)
{
try
{
fsw_.EnableRaisingEvents = false;
/* do my stuff once asynchronously */
}
finally
{
fsw_.EnableRaisingEvents = true;
}
}
Any duplicated OnChanged events from the FileSystemWatcher can be detected and discarded by checking the File.GetLastWriteTime timestamp on the file in question. Like so:
DateTime lastRead = DateTime.MinValue;
void OnChanged(object source, FileSystemEventArgs a)
{
DateTime lastWriteTime = File.GetLastWriteTime(uri);
if (lastWriteTime != lastRead)
{
doStuff();
lastRead = lastWriteTime;
}
// else discard the (duplicated) OnChanged event
}
Here is my solution which helped me to stop the event being raised twice:
watcher.NotifyFilter = NotifyFilters.FileName | NotifyFilters.Size;
Here I have set the NotifyFilter property with only Filename and size.
watcher is my object of FileSystemWatcher. Hope this will help.
Here's my approach :
// Consider having a List<String> named _changedFiles
private void OnChanged(object source, FileSystemEventArgs e)
{
lock (_changedFiles)
{
if (_changedFiles.Contains(e.FullPath))
{
return;
}
_changedFiles.Add(e.FullPath);
}
// do your stuff
System.Timers.Timer timer = new Timer(1000) { AutoReset = false };
timer.Elapsed += (timerElapsedSender, timerElapsedArgs) =>
{
lock (_changedFiles)
{
_changedFiles.Remove(e.FullPath);
}
};
timer.Start();
}
This is the solution I used to solve this issue on a project where I was sending the file as attachment in a mail.
It will easily avoid the twice fired event even with a smaller timer interval but in my case 1000 was alright since I was happier with missing few changes than with flooding the mailbox with > 1 message per second.
At least it works just fine in case several files are changed at the exact same time.
Another solution I've thought of would be to replace the list with a dictionary mapping files to their respective MD5, so you wouldn't have to choose an arbitrary interval since you wouldn't have to delete the entry but update its value, and cancel your stuff if it hasn't changed.
It has the downside of having a Dictionary growing in memory as files are monitored and eating more and more memory, but I've read somewhere that the amount of files monitored depends on the FSW's internal buffer, so maybe not that critical.
Dunno how MD5 computing time would affect your code's performances either, careful =\
My scenario is that I have a virtual machine with a Linux server in it. I am developing files on the Windows host. When I change something in a folder on the host I want all the changes to be uploaded, synced onto the virtual server via Ftp. This is how I do eliminate the duplicate change event when I write to a file ( which flags the folder containing the file to be modified as well ) :
private Hashtable fileWriteTime = new Hashtable();
private void fsw_sync_Changed(object source, FileSystemEventArgs e)
{
string path = e.FullPath.ToString();
string currentLastWriteTime = File.GetLastWriteTime( e.FullPath ).ToString();
// if there is no path info stored yet
// or stored path has different time of write then the one now is inspected
if ( !fileWriteTime.ContainsKey(path) ||
fileWriteTime[path].ToString() != currentLastWriteTime
)
{
//then we do the main thing
log( "A CHANGE has occured with " + path );
//lastly we update the last write time in the hashtable
fileWriteTime[path] = currentLastWriteTime;
}
}
Mainly I create a hashtable to store file write time information. Then if the hashtable has the filepath that is modified and it's time value is the same as the currently notified file's change then I know it is the duplicate of the event and ignore it.
I have created a Git repo with a class that extends FileSystemWatcher to trigger the events only when copy is done. It discards all the changed events exept the last and it raise it only when the file become available for read.
Download FileSystemSafeWatcher and add it to your project.
Then use it as a normal FileSystemWatcher and monitor when the events are triggered.
var fsw = new FileSystemSafeWatcher(file);
fsw.EnableRaisingEvents = true;
// Add event handlers here
fsw.Created += fsw_Created;
Try with this code:
class WatchPlotDirectory
{
bool let = false;
FileSystemWatcher watcher;
string path = "C:/Users/jamie/OneDrive/Pictures/Screenshots";
public WatchPlotDirectory()
{
watcher = new FileSystemWatcher();
watcher.Path = path;
watcher.NotifyFilter = NotifyFilters.LastAccess | NotifyFilters.LastWrite
| NotifyFilters.FileName | NotifyFilters.DirectoryName;
watcher.Filter = "*.*";
watcher.Changed += new FileSystemEventHandler(OnChanged);
watcher.Renamed += new RenamedEventHandler(OnRenamed);
watcher.EnableRaisingEvents = true;
}
void OnChanged(object sender, FileSystemEventArgs e)
{
if (let==false) {
string mgs = string.Format("File {0} | {1}",
e.FullPath, e.ChangeType);
Console.WriteLine("onchange: " + mgs);
let = true;
}
else
{
let = false;
}
}
void OnRenamed(object sender, RenamedEventArgs e)
{
string log = string.Format("{0} | Renamed from {1}",
e.FullPath, e.OldName);
Console.WriteLine("onrenamed: " + log);
}
public void setPath(string path)
{
this.path = path;
}
}
I know this is an old issue, but had the same problem and none of the above solution really did the trick for the problem I was facing. I have created a dictionary which maps the file name with the LastWriteTime. So if the file is not in the dictionary will go ahead with the process other wise check to see when was the last modified time and if is different from what it is in the dictionary run the code.
Dictionary<string, DateTime> dateTimeDictionary = new Dictionary<string, DateTime>();
private void OnChanged(object source, FileSystemEventArgs e)
{
if (!dateTimeDictionary.ContainsKey(e.FullPath) || (dateTimeDictionary.ContainsKey(e.FullPath) && System.IO.File.GetLastWriteTime(e.FullPath) != dateTimeDictionary[e.FullPath]))
{
dateTimeDictionary[e.FullPath] = System.IO.File.GetLastWriteTime(e.FullPath);
//your code here
}
}
One possible 'hack' would be to throttle the events using Reactive Extensions for example:
var watcher = new FileSystemWatcher("./");
Observable.FromEventPattern<FileSystemEventArgs>(watcher, "Changed")
.Throttle(new TimeSpan(500000))
.Subscribe(HandleChangeEvent);
watcher.EnableRaisingEvents = true;
In this case I'm throttling to 50ms, on my system that was enough, but higher values should be safer. (And like I said, it's still a 'hack').
I spent some significant amount of time using the FileSystemWatcher, and some of the approaches here will not work. I really liked the disabling events approach, but unfortunately, it doesn't work if there is >1 file being dropped, second file will be missed most if not all times.
So I use the following approach:
private void EventCallback(object sender, FileSystemEventArgs e)
{
var fileName = e.FullPath;
if (!File.Exists(fileName))
{
// We've dealt with the file, this is just supressing further events.
return;
}
// File exists, so move it to a working directory.
File.Move(fileName, [working directory]);
// Kick-off whatever processing is required.
}
I have a very quick and simple workaround here, it does work for me, and no matter the event would be triggered once or twice or more times occasionally, check it out:
private int fireCount = 0;
private void inputFileWatcher_Changed(object sender, FileSystemEventArgs e)
{
fireCount++;
if (fireCount == 1)
{
MessageBox.Show("Fired only once!!");
dowork();
}
else
{
fireCount = 0;
}
}
}
Here is a new solution you can try. Works well for me. In the event handler for the changed event programmatically remove the handler from the designer output a message if desired then programmatically add the handler back. example:
public void fileSystemWatcher1_Changed( object sender, System.IO.FileSystemEventArgs e )
{
fileSystemWatcher1.Changed -= new System.IO.FileSystemEventHandler( fileSystemWatcher1_Changed );
MessageBox.Show( "File has been uploaded to destination", "Success!" );
fileSystemWatcher1.Changed += new System.IO.FileSystemEventHandler( fileSystemWatcher1_Changed );
}
The main reason was
first event's last access time was current time(file write or changed time).
then second event was file's original last access time.
I solve under code.
var lastRead = DateTime.MinValue;
Watcher = new FileSystemWatcher(...)
{
NotifyFilter = NotifyFilters.FileName | NotifyFilters.LastWrite,
Filter = "*.dll",
IncludeSubdirectories = false,
};
Watcher.Changed += (senderObject, ea) =>
{
var now = DateTime.Now;
var lastWriteTime = File.GetLastWriteTime(ea.FullPath);
if (now == lastWriteTime)
{
return;
}
if (lastWriteTime != lastRead)
{
// do something...
lastRead = lastWriteTime;
}
};
Watcher.EnableRaisingEvents = true;
This code worked for me.
private void OnChanged(object source, FileSystemEventArgs e)
{
string fullFilePath = e.FullPath.ToString();
string fullURL = buildTheUrlFromStudyXML(fullFilePath);
System.Diagnostics.Process.Start("iexplore", fullURL);
Timer timer = new Timer();
((FileSystemWatcher)source).Changed -= new FileSystemEventHandler(OnChanged);
timer.Interval = 1000;
timer.Elapsed += new ElapsedEventHandler(t_Elapsed);
timer.Start();
}
private void t_Elapsed(object sender, ElapsedEventArgs e)
{
((Timer)sender).Stop();
theWatcher.Changed += new FileSystemEventHandler(OnChanged);
}
mostly for future me :)
I wrote a wrapper using Rx:
public class WatcherWrapper : IDisposable
{
private readonly FileSystemWatcher _fileWatcher;
private readonly Subject<FileSystemEventArgs> _infoSubject;
private Subject<FileSystemEventArgs> _eventSubject;
public WatcherWrapper(string path, string nameFilter = "*.*", NotifyFilters? notifyFilters = null)
{
_fileWatcher = new FileSystemWatcher(path, nameFilter);
if (notifyFilters != null)
{
_fileWatcher.NotifyFilter = notifyFilters.Value;
}
_infoSubject = new Subject<FileSystemEventArgs>();
_eventSubject = new Subject<FileSystemEventArgs>();
Observable.FromEventPattern<FileSystemEventArgs>(_fileWatcher, "Changed").Select(e => e.EventArgs)
.Subscribe(_infoSubject.OnNext);
Observable.FromEventPattern<FileSystemEventArgs>(_fileWatcher, "Created").Select(e => e.EventArgs)
.Subscribe(_infoSubject.OnNext);
Observable.FromEventPattern<FileSystemEventArgs>(_fileWatcher, "Deleted").Select(e => e.EventArgs)
.Subscribe(_infoSubject.OnNext);
Observable.FromEventPattern<FileSystemEventArgs>(_fileWatcher, "Renamed").Select(e => e.EventArgs)
.Subscribe(_infoSubject.OnNext);
// this takes care of double events and still works with changing the name of the same file after a while
_infoSubject.Buffer(TimeSpan.FromMilliseconds(20))
.Select(x => x.GroupBy(z => z.FullPath).Select(z => z.LastOrDefault()).Subscribe(
infos =>
{
if (infos != null)
foreach (var info in infos)
{
{
_eventSubject.OnNext(info);
}
}
});
_fileWatcher.EnableRaisingEvents = true;
}
public IObservable<FileSystemEventArgs> FileEvents => _eventSubject;
public void Dispose()
{
_fileWatcher?.Dispose();
_eventSubject.Dispose();
_infoSubject.Dispose();
}
}
Usage:
var watcher = new WatcherWrapper(_path, "*.info");
// all more complicated and scenario specific filtering of events can be done here
watcher.FileEvents.Where(x => x.ChangeType != WatcherChangeTypes.Deleted).Subscribe(x => //do stuff)
Try this, It's working fine
private static readonly FileSystemWatcher Watcher = new FileSystemWatcher();
static void Main(string[] args)
{
Console.WriteLine("Watching....");
Watcher.Path = #"D:\Temp\Watcher";
Watcher.Changed += OnChanged;
Watcher.EnableRaisingEvents = true;
Console.ReadKey();
}
static void OnChanged(object sender, FileSystemEventArgs e)
{
try
{
Watcher.Changed -= OnChanged;
Watcher.EnableRaisingEvents = false;
Console.WriteLine($"File Changed. Name: {e.Name}");
}
catch (Exception exception)
{
Console.WriteLine(exception);
}
finally
{
Watcher.Changed += OnChanged;
Watcher.EnableRaisingEvents = true;
}
}
I wanted to react only on the last event, just in case, also on a linux file change it seemed that the file was empty on the first call and then filled again on the next and did not mind loosing some time just in case the OS decided to do some file/attribute change.
I am using .NET async here to help me do the threading.
private static int _fileSystemWatcherCounts;
private async void OnChanged(object sender, FileSystemEventArgs e)
{
// Filter several calls in short period of time
Interlocked.Increment(ref _fileSystemWatcherCounts);
await Task.Delay(100);
if (Interlocked.Decrement(ref _fileSystemWatcherCounts) == 0)
DoYourWork();
}
I think the best solution to solve the issue is to use reactive extensions
When you transform event into observable, then you can just add Throttling(..) (originally called Debounce(..))
Sample code here
var templatesWatcher = new FileSystemWatcher(settingsSnapshot.Value.TemplatesDirectory)
{
NotifyFilter = NotifyFilters.LastWrite,
IncludeSubdirectories = true
};
templatesWatcher.EnableRaisingEvents = true;
Observable.FromEventPattern<FileSystemEventHandler, FileSystemEventArgs>(
addHandler => templatesWatcher.Changed += addHandler,
removeHandler => templatesWatcher.Changed -= removeHandler)
.Throttle(TimeSpan.FromSeconds(5))
.Subscribe(args =>
{
_logger.LogInformation($"Template file {args.EventArgs.Name} has changed");
//TODO do something
});
You could try to open it for write, and if successful then you could assume the other application is done with the file.
private void OnChanged(object source, FileSystemEventArgs e)
{
try
{
using (var fs = File.OpenWrite(e.FullPath))
{
}
//do your stuff
}
catch (Exception)
{
//no write access, other app not done
}
}
Just opening it for write appears not to raise the changed event. So it should be safe.
FileReadTime = DateTime.Now;
private void File_Changed(object sender, FileSystemEventArgs e)
{
var lastWriteTime = File.GetLastWriteTime(e.FullPath);
if (lastWriteTime.Subtract(FileReadTime).Ticks > 0)
{
// code
FileReadTime = DateTime.Now;
}
}
Sorry for the grave dig, but I've been battling this issue for a while now and finally came up with a way to handle these multiple fired events. I would like to thank everyone in this thread as I have used it in many references when battling this issue.
Here is my complete code. It uses a dictionary to track the date and time of the last write of the file. It compares that value, and if it is the same, it suppresses the events. It then sets the value after starting the new thread.
using System.Threading; // used for backgroundworker
using System.Diagnostics; // used for file information
private static IDictionary<string, string> fileModifiedTable = new Dictionary<string, string>(); // used to keep track of our changed events
private void fswFileWatch_Changed( object sender, FileSystemEventArgs e )
{
try
{
//check if we already have this value in our dictionary.
if ( fileModifiedTable.TryGetValue( e.FullPath, out sEmpty ) )
{
//compare timestamps
if ( fileModifiedTable[ e.FullPath ] != File.GetLastWriteTime( e.FullPath ).ToString() )
{
//lock the table
lock ( fileModifiedTable )
{
//make sure our file is still valid
if ( File.Exists( e.FullPath ) )
{
// create a new background worker to do our task while the main thread stays awake. Also give it do work and work completed handlers
BackgroundWorker newThreadWork = new BackgroundWorker();
newThreadWork.DoWork += new DoWorkEventHandler( bgwNewThread_DoWork );
newThreadWork.RunWorkerCompleted += new RunWorkerCompletedEventHandler( bgwNewThread_RunWorkerCompleted );
// capture the path
string eventFilePath = e.FullPath;
List<object> arguments = new List<object>();
// add arguments to pass to the background worker
arguments.Add( eventFilePath );
arguments.Add( newEvent.File_Modified );
// start the new thread with the arguments
newThreadWork.RunWorkerAsync( arguments );
fileModifiedTable[ e.FullPath ] = File.GetLastWriteTime( e.FullPath ).ToString(); //update the modified table with the new timestamp of the file.
FILE_MODIFIED_FLAG.WaitOne(); // wait for the modified thread to complete before firing the next thread in the event multiple threads are being worked on.
}
}
}
}
}
catch ( IOException IOExcept )
{
//catch any errors
postError( IOExcept, "fswFileWatch_Changed" );
}
}
Event if not asked, it is a shame there are no ready solution samples for F#.
To fix this here is my recipe, just because I can and F# is a wonderful .NET language.
Duplicated events are filtered out using FSharp.Control.Reactive package, which is just a F# wrapper for reactive extensions. All that can be targeted to full framework or netstandard2.0:
let createWatcher path filter () =
new FileSystemWatcher(
Path = path,
Filter = filter,
EnableRaisingEvents = true,
SynchronizingObject = null // not needed for console applications
)
let createSources (fsWatcher: FileSystemWatcher) =
// use here needed events only.
// convert `Error` and `Renamed` events to be merded
[| fsWatcher.Changed :> IObservable<_>
fsWatcher.Deleted :> IObservable<_>
fsWatcher.Created :> IObservable<_>
//fsWatcher.Renamed |> Observable.map renamedToNeeded
//fsWatcher.Error |> Observable.map errorToNeeded
|] |> Observable.mergeArray
let handle (e: FileSystemEventArgs) =
printfn "handle %A event '%s' '%s' " e.ChangeType e.Name e.FullPath
let watch path filter throttleTime =
// disposes watcher if observer subscription is disposed
Observable.using (createWatcher path filter) createSources
// filter out multiple equal events
|> Observable.distinctUntilChanged
// filter out multiple Changed
|> Observable.throttle throttleTime
|> Observable.subscribe handle
[<EntryPoint>]
let main _args =
let path = #"C:\Temp\WatchDir"
let filter = "*.zip"
let throttleTime = TimeSpan.FromSeconds 10.
use _subscription = watch path filter throttleTime
System.Console.ReadKey() |> ignore
0 // return an integer exit code
In my case need to get the last line of a text file that is inserted by other application, as soon as insertion is done. Here is my solution. When the first event is raised, i disable the watcher from raising others, then i call the timer TimeElapsedEvent because when my handle function OnChanged is called i need the size of the text file, but the size at that time is not the actual size, it is the size of the file imediatelly before the insertion. So i wait for a while to proceed with the right file size.
private FileSystemWatcher watcher = new FileSystemWatcher();
...
watcher.Path = "E:\\data";
watcher.NotifyFilter = NotifyFilters.LastWrite ;
watcher.Filter = "data.txt";
watcher.Changed += new FileSystemEventHandler(OnChanged);
watcher.EnableRaisingEvents = true;
...
private void OnChanged(object source, FileSystemEventArgs e)
{
System.Timers.Timer t = new System.Timers.Timer();
try
{
watcher.Changed -= new FileSystemEventHandler(OnChanged);
watcher.EnableRaisingEvents = false;
t.Interval = 500;
t.Elapsed += (sender, args) => t_Elapsed(sender, e);
t.Start();
}
catch(Exception ex) {
;
}
}
private void t_Elapsed(object sender, FileSystemEventArgs e)
{
((System.Timers.Timer)sender).Stop();
//.. Do you stuff HERE ..
watcher.Changed += new FileSystemEventHandler(OnChanged);
watcher.EnableRaisingEvents = true;
}
Here is another approach. Instead of propagating the first event of a quick succession of events and suppressing all that follow, now all are suppressed except from the last one. I think that the scenarios that can benefit from this approach are more common.
To make this happen we must use a sliding delay. Every incoming event cancels the timer that would fire the previous event, and restarts the timer. This opens the possibility that a never-ending series of events will delay the propagation forever. To keep things simple, there is no provision for this abnormal case in the extension methods below.
public static class FileSystemWatcherExtensions
{
public static IDisposable OnAnyEvent(this FileSystemWatcher source,
WatcherChangeTypes changeTypes, FileSystemEventHandler handler, int delay)
{
var cancellations = new Dictionary<string, CancellationTokenSource>(
StringComparer.OrdinalIgnoreCase);
var locker = new object();
if (changeTypes.HasFlag(WatcherChangeTypes.Created))
source.Created += FileSystemWatcher_Event;
if (changeTypes.HasFlag(WatcherChangeTypes.Deleted))
source.Deleted += FileSystemWatcher_Event;
if (changeTypes.HasFlag(WatcherChangeTypes.Changed))
source.Changed += FileSystemWatcher_Event;
if (changeTypes.HasFlag(WatcherChangeTypes.Renamed))
source.Renamed += FileSystemWatcher_Event;
return new Disposable(() =>
{
source.Created -= FileSystemWatcher_Event;
source.Deleted -= FileSystemWatcher_Event;
source.Changed -= FileSystemWatcher_Event;
source.Renamed -= FileSystemWatcher_Event;
});
async void FileSystemWatcher_Event(object sender, FileSystemEventArgs e)
{
var key = e.FullPath;
var cts = new CancellationTokenSource();
lock (locker)
{
if (cancellations.TryGetValue(key, out var existing))
{
existing.Cancel();
}
cancellations[key] = cts;
}
try
{
await Task.Delay(delay, cts.Token);
// Omitting ConfigureAwait(false) is intentional here.
// Continuing in the captured context is desirable.
}
catch (TaskCanceledException)
{
return;
}
lock (locker)
{
if (cancellations.TryGetValue(key, out var existing)
&& existing == cts)
{
cancellations.Remove(key);
}
}
cts.Dispose();
handler(sender, e);
}
}
public static IDisposable OnAllEvents(this FileSystemWatcher source,
FileSystemEventHandler handler, int delay)
=> OnAnyEvent(source, WatcherChangeTypes.All, handler, delay);
public static IDisposable OnCreated(this FileSystemWatcher source,
FileSystemEventHandler handler, int delay)
=> OnAnyEvent(source, WatcherChangeTypes.Created, handler, delay);
public static IDisposable OnDeleted(this FileSystemWatcher source,
FileSystemEventHandler handler, int delay)
=> OnAnyEvent(source, WatcherChangeTypes.Deleted, handler, delay);
public static IDisposable OnChanged(this FileSystemWatcher source,
FileSystemEventHandler handler, int delay)
=> OnAnyEvent(source, WatcherChangeTypes.Changed, handler, delay);
public static IDisposable OnRenamed(this FileSystemWatcher source,
FileSystemEventHandler handler, int delay)
=> OnAnyEvent(source, WatcherChangeTypes.Renamed, handler, delay);
private struct Disposable : IDisposable
{
private readonly Action _action;
internal Disposable(Action action) => _action = action;
public void Dispose() => _action?.Invoke();
}
}
Usage example:
myWatcher.OnAnyEvent(WatcherChangeTypes.Created | WatcherChangeTypes.Changed,
MyFileSystemWatcher_Event, 100);
This line combines the subscription to two events, the Created and the Changed. So it is roughly equivalent to these:
myWatcher.Created += MyFileSystemWatcher_Event;
myWatcher.Changed += MyFileSystemWatcher_Event;
The difference is that the two events are regarded as a single type of event, and in case of a quick succession of these events only the last one will be propagated. For example if a Created event is followed by two Changed events, and there is no time gap larger than 100 msec between these three events, only the second Changed event will be propagated by invoking the MyFileSystemWatcher_Event handler, and the previous ones will be discarded.
I have changed the way I monitor files in directories. Instead of using the FileSystemWatcher I poll locations on another thread and then look at the LastWriteTime of the file.
DateTime lastWriteTime = File.GetLastWriteTime(someFilePath);
Using this information and keeping an index of a file path and it's latest write time I can determine files that have changed or that have been created in a particular location. This removes me from the oddities of the FileSystemWatcher. The main downside is that you need a data structure to store the LastWriteTime and the reference to the file, but it is reliable and easy to implement.
I was able to do this by added a function that checks for duplicates in an buffer array.
Then perform the action after the array has not been modified for X time using a timer:
- Reset timer every time something is written to the buffer
- Perform action on tick
This also catches another duplication type. If you modify a file inside a folder, the folder also throws a Change event.
Function is_duplicate(str1 As String) As Boolean
If lb_actions_list.Items.Count = 0 Then
Return False
Else
Dim compStr As String = lb_actions_list.Items(lb_actions_list.Items.Count - 1).ToString
compStr = compStr.Substring(compStr.IndexOf("-") + 1).Trim
If compStr <> str1 AndAlso compStr.parentDir <> str1 & "\" Then
Return False
Else
Return True
End If
End If
End Function
Public Module extentions
<Extension()>
Public Function parentDir(ByVal aString As String) As String
Return aString.Substring(0, CInt(InStrRev(aString, "\", aString.Length - 1)))
End Function
End Module
This solution worked for me on production application:
Environment:
VB.Net Framework 4.5.2
Set manually object properties: NotifyFilter = Size
Then use this code:
Public Class main
Dim CalledOnce = False
Private Sub FileSystemWatcher1_Changed(sender As Object, e As IO.FileSystemEventArgs) Handles FileSystemWatcher1.Changed
If (CalledOnce = False) Then
CalledOnce = True
If (e.ChangeType = 4) Then
' Do task...
CalledOnce = False
End If
End Sub
End Sub
Try this!
string temp="";
public void Initialize()
{
FileSystemWatcher _fileWatcher = new FileSystemWatcher();
_fileWatcher.Path = "C:\\Folder";
_fileWatcher.NotifyFilter = NotifyFilters.LastWrite;
_fileWatcher.Filter = "Version.txt";
_fileWatcher.Changed += new FileSystemEventHandler(OnChanged);
_fileWatcher.EnableRaisingEvents = true;
}
private void OnChanged(object source, FileSystemEventArgs e)
{
.......
if(temp=="")
{
//do thing you want.
temp = e.name //name of text file.
}else if(temp !="" && temp != e.name)
{
//do thing you want.
temp = e.name //name of text file.
}else
{
//second fire ignored.
}
}
I had to combine several ideas from the posts above and add file locking check to get it working for me:
FileSystemWatcher fileSystemWatcher;
private void DirectoryWatcher_Start()
{
FileSystemWatcher fileSystemWatcher = new FileSystemWatcher
{
Path = #"c:\mypath",
NotifyFilter = NotifyFilters.LastWrite,
Filter = "*.*",
EnableRaisingEvents = true
};
fileSystemWatcher.Changed += new FileSystemEventHandler(DirectoryWatcher_OnChanged);
}
private static void WaitUntilFileIsUnlocked(String fullPath, Action<String> callback, FileAccess fileAccess = FileAccess.Read, Int32 timeoutMS = 10000)
{
Int32 waitMS = 250;
Int32 currentMS = 0;
FileInfo file = new FileInfo(fullPath);
FileStream stream = null;
do
{
try
{
stream = file.Open(FileMode.Open, fileAccess, FileShare.None);
stream.Close();
callback(fullPath);
return;
}
catch (IOException)
{
}
finally
{
if (stream != null)
stream.Dispose();
}
Thread.Sleep(waitMS);
currentMS += waitMS;
} while (currentMS < timeoutMS);
}
private static Dictionary<String, DateTime> DirectoryWatcher_fileLastWriteTimeCache = new Dictionary<String, DateTime>();
private void DirectoryWatcher_OnChanged(Object source, FileSystemEventArgs ev)
{
try
{
lock (DirectoryWatcher_fileLastWriteTimeCache)
{
DateTime lastWriteTime = File.GetLastWriteTime(ev.FullPath);
if (DirectoryWatcher_fileLastWriteTimeCache.ContainsKey(ev.FullPath))
{
if (DirectoryWatcher_fileLastWriteTimeCache[ev.FullPath].AddMilliseconds(500) >= lastWriteTime)
return; // file was already handled
}
DirectoryWatcher_fileLastWriteTimeCache[ev.FullPath] = lastWriteTime;
}
Task.Run(() => WaitUntilFileIsUnlocked(ev.FullPath, fullPath =>
{
// do the job with fullPath...
}));
}
catch (Exception e)
{
// handle exception
}
}

How can I create an Rx observable which stops publishing events when the last observer unsubscribes?

I'll create an observable (through a variety of means) and return it to interested parties, but when they're done listening, I'd like to tear down the observable so it doesn't continue consuming resources. Another way to think of it as creating topics in a pub sub system. When no one is subscribed to a topic any more, you don't want to hold the topic and its filtering around anymore.
Rx already has an operator to suit your needs - well two actually - Publish & RefCount.
Here's how to use them:
IObservable xs = ...
var rxs = xs.Publish().RefCount();
var sub1 = rxs.Subscribe(x => { });
var sub2 = rxs.Subscribe(x => { });
//later
sub1.Dispose();
//later
sub2.Dispose();
//The underlying subscription to `xs` is now disposed of.
Simple.
If I have understood your question you want to create the observable such that when all subscribers have disposed their subscription i.e there is no more subscriber, then you want to execute a clean up function which will stop the observable from production further values.
If this is what you want then you can do something like below:
//Wrap a disposable
public class WrapDisposable : IDisposable
{
IDisposable disp;
Action act;
public WrapDisposable(IDisposable _disp, Action _act)
{
disp = _disp;
act = _act;
}
void IDisposable.Dispose()
{
act();
disp.Dispose();
}
}
//Observable that we want to clean up after all subs are done
public static IObservable<long> GenerateObs(out Action cleanup)
{
cleanup = () =>
{
Console.WriteLine("All subscribers are done. Do clean up");
};
return Observable.Interval(TimeSpan.FromSeconds(1));
}
//Wrap the observable
public static IObservable<T> WrapToClean<T>(IObservable<T> obs, Action onAllDone)
{
int count = 0;
return Observable.CreateWithDisposable<T>(ob =>
{
var disp = obs.Subscribe(ob);
Interlocked.Increment(ref count);
return new WrapDisposable(disp,() =>
{
if (Interlocked.Decrement(ref count) == 0)
{
onAllDone();
}
});
});
}
//Usage example:
Action cleanup;
var obs = GenerateObs(out cleanup);
var newObs = WrapToClean(obs, cleanup);
newObs.Take(6).Subscribe(Console.WriteLine);
newObs.Take(5).Subscribe(Console.WriteLine);

Using yield to iterate over a datareader might not close the connection?

Here is a sample code to retrieve data from a database using the yield keyword that I found in a few place while googling around :
public IEnumerable<object> ExecuteSelect(string commandText)
{
using (IDbConnection connection = CreateConnection())
{
using (IDbCommand cmd = CreateCommand(commandText, connection))
{
connection.Open();
using (IDbDataReader reader = cmd.ExecuteReader())
{
while(reader.Read())
{
yield return reader["SomeField"];
}
}
connection.Close();
}
}
}
Am I correct in thinking that in this sample code, the connection would not be closed if we do not iterate over the whole datareader ?
Here is an example that would not close the connection, if I understand yield correctly..
foreach(object obj in ExecuteSelect(commandText))
{
break;
}
For a db connection that might not be catastrophic, I suppose the GC would clean it up eventually, but what if instead of a connection it was a more critical resource?
The Iterator that the compiler synthesises implements IDisposable, which foreach calls when the foreach loop is exited.
The Iterator's Dispose() method will clean up the using statements on early exit.
As long as you use the iterator in a foreach loop, using() block, or call the Dispose() method in some other way, the cleanup of the Iterator will happen.
Connection will be closed automatically since you're using it inside "using" block.
From the simple test I have tried, aku is right, dispose is called as soon as the foreach block exit.
#David : However call stack is kept between call, so the connection would not be closed because on the next call we would return to the next instruction after the yield, which is the while block.
My understanding is that when the iterator is disposed, the connection would also be disposed with it. I also think that the Connection.Close would not be needed because it would be taken care of when the object is disposed because of the using clause.
Here is a simple program I tried to test the behavior...
class Program
{
static void Main(string[] args)
{
foreach (int v in getValues())
{
Console.WriteLine(v);
}
Console.ReadKey();
foreach (int v in getValues())
{
Console.WriteLine(v);
break;
}
Console.ReadKey();
}
public static IEnumerable<int> getValues()
{
using (TestDisposable t = new TestDisposable())
{
for(int i = 0; i<10; i++)
yield return t.GetValue();
}
}
}
public class TestDisposable : IDisposable
{
private int value;
public void Dispose()
{
Console.WriteLine("Disposed");
}
public int GetValue()
{
value += 1;
return value;
}
}
Judging from this technical explanation, your code will not work as expected, but abort on the second item, because the connection was already closed when returning the first item.
#Joel Gauvreau : Yes, I should have read on. Part 3 of this series explains that the compiler adds special handling for finally blocks to trigger only at the real end.