Editing in a tableview without selecting
Hi all. I've got an NSTableView that I'd like to be editable.
Even if I return YES from -tableView:shouldSelectRow: and YES from
tableView:shouldEditTableColumn:row:, however, editing does not
actually commence. The row selects, and I've confirmed with logs that
both delegate methods are indeed getting called; the tableView seems
to just decide not to start editing after all. No other console logs
appear, so it's not a raise. Simply not implementing the two delegate
methods mentioned above does not make any difference. The object
value for cell that should enter editing is an NSString, and its cell
is an NSTextFieldCell. Is there some additional step I have to take
to make editing actually commence?
The tableview is set up in code. What appear to be custom
subclasses in this code are all #defined back to their superclasses,
and all of the code for these classes if #if 0'ed out, so this code is
really using only the Cocoa classes:
#define AKVariableStoreTableView NSTableView
#define AKTextFieldCell NSTextFieldCell
#define AKSwatchCell NSTextFieldCell
#define AKDividerCell NSTextFieldCell
+ (NSTableColumn *)columnObjectForIdentifier:(NSString *)identifier
{
NSTableColumn *column = [[NSTableColumn alloc]
initWithIdentifier:identifier];
if (identifier == nameColumnID)
{
AKTextFieldCell *cell = [[AKTextFieldCell alloc]
initTextCell:@""];
[column setMinWidth:AKNameColumnWidth];
[column setMaxWidth:AKNameColumnWidth];
[cell setFont:[NSFont
antikytheraFontOfSize:AK_VARIABLE_STORE_VARIABLE_FONT_SIZE]];
[column setDataCell:cell];
}
else if (identifier == equalsColumnID)
{
AKTextFieldCell *cell = [[AKTextFieldCell alloc]
initTextCell:@""];
[column setMinWidth:AKEqualsColumnWidth];
[column setMaxWidth:AKEqualsColumnWidth];
[cell setAlignment:NSCenterTextAlignment];
[cell setFont:[NSFont
antikytheraFontOfSize:AK_VARIABLE_STORE_VALUE_FONT_SIZE]];
[column setDataCell:cell];
}
else if (identifier == valueColumnID)
{
AKTextFieldCell *cell = [[AKTextFieldCell alloc]
initTextCell:@""];
[column setMinWidth:AKValueColumnWidth];
[column setMaxWidth:AKValueColumnWidth];
[cell setAlignment:NSRightTextAlignment];
[cell setFont:[NSFont
antikytheraFontOfSize:AK_VARIABLE_STORE_VALUE_FONT_SIZE]];
[column setDataCell:cell];
}
else if (identifier == swatchColumnID)
{
AKSwatchCell *cell = [[AKSwatchCell alloc] initImageCell:nil];
[column setMinWidth:AKSwatchColumnWidth];
[column setMaxWidth:AKSwatchColumnWidth];
[column setDataCell:cell];
}
else if (identifier == descriptionColumnID)
{
AKTextFieldCell *cell = [[AKTextFieldCell alloc]
initTextCell:@""];
[column setMinWidth:AKDescriptionColumnWidth];
[column setMaxWidth:1000];
[cell setTextColor:[NSColor grayColor]];
[cell setAlignment:NSRightTextAlignment];
[cell setFont:[NSFont
antikytheraFontOfSize:AK_VARIABLE_STORE_DESCRIPTION_FONT_SIZE]];
[column setDataCell:cell];
}
return column;
}
#define AK_VARIABLE_STORE_INSET 15
- (NSTableView *)makeTableViewWithWidth:(int)width
{
if (!tableView)
{
NSRect tableViewFrame = NSMakeRect(AK_VARIABLE_STORE_INSET,
AK_VARIABLE_STORE_INSET, width, 100);
tableView = [[AKVariableStoreTableView alloc]
initWithFrame:tableViewFrame];
// Set up behavior
[tableView setAllowsColumnReordering:NO];
[tableView setAllowsColumnResizing:NO];
[tableView setAllowsColumnSelection:NO];
[tableView setAllowsEmptySelection:YES];
[tableView setAllowsMultipleSelection:NO];
[tableView setAllowsTypeSelect:NO];
[tableView setAllowsColumnReordering:NO];
[tableView setAutosaveTableColumns:NO];
[tableView
setColumnAutoresizingStyle:NSTableViewLastColumnOnlyAutoresizingStyle];
[tableView setAutosaveTableColumns:NO];
// Set up appearance
[tableView setBackgroundColor:[NSColor windowBackgroundColor]];
[tableView setGridStyleMask:NSTableViewGridNone];
[tableView setIntercellSpacing:NSMakeSize(0.0, 1.0)];
[tableView setRowHeight:18];
Re: Editing in a tableview without selecting
On 1-Nov-09, at 12:42 PM, Ben Haller wrote: Hi all. I've got an NSTableView that I'd like to be editable. Even if I return YES from -tableView:shouldSelectRow: and YES from tableView:shouldEditTableColumn:row:, however, editing does not actually commence. The row selects, and I've confirmed with logs that both delegate methods are indeed getting called; the tableView seems to just decide not to start editing after all. No other console logs appear, so it's not a raise. Simply not implementing the two delegate methods mentioned above does not make any difference. The object value for cell that should enter editing is an NSString, and its cell is an NSTextFieldCell. Is there some additional step I have to take to make editing actually commence? The title of this thread is a bit wrong, sorry about that. I was originally trying to figure out a way to allow editing without selecting, but when I realized I couldn't get editing to happen at all, in any case whatsoever, my focus shifted. :-> I am letting the row get selected as normal, and it seems to be selected correctly, as my first message says. Ben Haller Stick Software ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Re: Asynchronous downloading again
On Oct 31, 2009, at 6:50 PM, DKJ wrote: All of the files have to be downloaded before the app can do anything. I get the connectionDidFinishLoading delegate method of my one and only NSURLConnection to call a downloadFinished method at the end of the synchronous downloads, so the app knows everything is in place and can start processing the files. All NSURLConnection delegate methods are called on the main thread. (Actually on the same thread from which you started the NSURLConnection.) This is true of pretty much everything in Cocoa: the only time any of your code runs on a different thread is when you explicitly request it, by using NSThread or NSOperationQueue (or the dispatcher API in 10.6.) Don't go back to using synchronous connections. They will block the main thread, and there's no sense in using two different techniques to solve the same problem. Your approach of starting each connection when the previous one finishes is fine. If all of these downloads are coming from the same server, and you can't do anything till they're all done, you're probably not going to gain anything by doing them in parallel: either way, the limiting factor will be the speed of the user's Internet connection. —Jens___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Re: Bold and Italic For Dictionary Font Attributes (Hillegass Chapter 20, Challenge 2)
On Oct 31, 2009, at 12:44 PM, Russell Finn wrote: -- i.e. key, object, key, object; this is what I tried to correct this in my post, as shown above. Yeah, sorry about that; my mistake. (I'm so used to using my $dict, which puts the keys/values in the 'correct' order...) —Jens ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
objectAtIndex
Hi, When I ran Build and Analyze on my code, it has pointed as a possible leak the following statement: NSObject *anObject = [anArray objectAtIndex:i]; anObject wasn't released and here's the question: whether anObject is a copy of actual object, that is a member of anArray or it's the object itself? Should I release it? Thanks, Nava ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Re: objectAtIndex
That line alone does not indicate any memory leaks occurring due to over-retaining objects. You'll need to provide more context (the surrounding code, etc). The object returned by -[NSArray objectAtIndex:] is the actual object (ie. pointer to the actual object's space in memory) stored in the array, not a copy. When you add objects to an array, as well, the object stored in the array is not a copy but the object itself. - Bryan On Nov 1, 2009, at 2:44:07 PM, Nava Carmon wrote: Hi, When I ran Build and Analyze on my code, it has pointed as a possible leak the following statement: NSObject *anObject = [anArray objectAtIndex:i]; anObject wasn't released and here's the question: whether anObject is a copy of actual object, that is a member of anArray or it's the object itself? Should I release it? Thanks, Nava ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/bryanhenry%40mac.com This email sent to bryanhe...@mac.com ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Re: objectAtIndex
On Sun, Nov 1, 2009 at 2:44 PM, Nava Carmon wrote: > > NSObject *anObject = [anArray objectAtIndex:i]; ... > Should I release it? No. You didn't create anObject with +alloc or a method beginning with -copy, nor did you send it a -retain message. Therefore you do not own it and must not release it. sherm-- -- Cocoa programming in Perl: http://camelbones.sourceforge.net ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Does [NSNetService getInputStream:outputStream:] really return retain objects and where does it say that
Hi,
I'm puzzled by a comment in the WiTap iPhone SDK example project
that I can't confirm through documentation or header comments. The
method and comment is this (AppController.m line 214 onwards):
- (void) browserViewController:(BrowserViewController *)bvc
didResolveInstance:(NSNetService *)netService
{
if (!netService) {
[self setup];
return;
}
// note the following method returns _inStream and _outStream with a
retain count that the caller must eventually release
if (![netService getInputStream:&_inStream
outputStream:&_outStream]) {
[self _showAlert:@"Failed connecting to server"];
return;
}
[self openStreams];
}
Note the comment above "if (![netService getInputStream...".
The documentation doesn't back this assertion, neither does the header
comment in NSNetServices.h. According to the memory management rules
the method should return autoreleased streams.
Sure enough the author of the example project believes what he says
because the project doesn't retain the streams anywhere. And
surprisingly the project doesn't crash which if should at some point
if the streams where auto-released - so they must be retained.
So, am I blind? I can't find where it says they are.
Looking a bit further the same kind of comment pops up in on other
example too (Cocoa SimpleNetworkStream) where it also references a
radar (rdar://problem/6868813), so this is known. Why not add a note
to getInputStream:outputStream: then?
Regards
Markus
--
__
Markus Spoettl
smime.p7s
Description: S/MIME cryptographic signature
___
Cocoa-dev mailing list (Cocoa-dev@lists.apple.com)
Please do not post admin requests or moderator comments to the list.
Contact the moderators at cocoa-dev-admins(at)lists.apple.com
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com
This email sent to arch...@mail-archive.com
Re: Editing in a tableview without selecting
I'm new to Obj-c / Cocoa but this was helpful in figuring out editable table views I was trying to create with IB. Maybe it can help you in your case. http://www.idevgames.com/forum/showpost.php?p=94784&postcount=3 I know I wasn't putting a dictionary object into an array with the same identifier that I had created in IB. So I could "edit" my data in my tableview but as soon as I was done editing it never updated my array so it my data never changed. ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Re: Editing in a tableview without selecting
On 1-Nov-09, at 4:22 PM, Scott Waters wrote: I'm new to Obj-c / Cocoa but this was helpful in figuring out editable table views I was trying to create with IB. Maybe it can help you in your case. http://www.idevgames.com/forum/showpost.php?p=94784&postcount=3 I know I wasn't putting a dictionary object into an array with the same identifier that I had created in IB. So I could "edit" my data in my tableview but as soon as I was done editing it never updated my array so it my data never changed. Thanks for the reference, but the problem I'm seeing is that editing doesn't even begin on my tableview. I double-click on a cell, and nothing whatsoever happens. Ben Haller Stick Software ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Re: Editing in a tableview without selecting
On 02/11/2009, at 4:42 AM, Ben Haller wrote: But where? I've been hunting for a -setEditable:YES method somewhere that I need to call, but I haven't found it... Hi Ben, Have you tried -setEditable:YES on the textfield cell you're using for the table column? (Inherited from NSCell). I'm not sure what its default state is, documentation doesn't say. --Graham ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Re: Does [NSNetService getInputStream:outputStream:] really return retain objects and where does it say that
On Nov 1, 2009, at 1:30 PM, Markus Spoettl wrote: The documentation doesn't back this assertion, neither does the header comment in NSNetServices.h. According to the memory management rules the method should return autoreleased streams. Yes; but that doesn't necessarily mean the streams will go away soon if you don't retain them. An NSStream object is usually scheduled on a runloop, which keeps a reference to it as long as its open, so I believe it will stay around even if not explicitly retained by the app. Sure enough the author of the example project believes what he says because the project doesn't retain the streams anywhere. Does it release the streams? That would be the true test. My belief is that it can get away without retaining them, as long as it never accesses them again after they close; but if it explicitly releases them, it would crash. That said, I think it's safer to keep your own references to the stream objects by retaining and releasing them. —Jens ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Commandline Tool Help - NSDistantObject and NSConnection
hey Folks, quick question. I'm building an app that I'm providing a command line tool for. I have it hooked up and working. However the strange thing is that it works the first time it's run, but then just throws exceptions. I'm using a proxy object, which is exposed through NSConnection, and through NSDistantObject it exposes one method to open a document at a path. The application is using the document architecture, and I subclass NSDocumentController. The cliproxy is setup in the document controller code because it's initialized only once. here's some code: The commandline tool code: http://pastebin.com/m6c1bb952 Here's the proxy code h: http://pastebin.com/me651e88 m: http://pastebin.com/m57e4ab4e The (simplified) doc controller: http://pastebin.com/m6adf5d96 So what's strange is that it works fine on the first time I run the command line tool; it opens the app, and opens a document as expected. But if the cocoa application is already running - i'm seeing different behavior. here's a quick sequence:: *first run* $ ./build/Debug/gity 2009-11-01 16:01:33.895 gity[10625:903] proxy: 2009-11-01 16:01:33.898 gity[10625:903] proxy: 2009-11-01 16:01:33.899 gity[10625:903] doc: *second run (app is running)* $ ./build/Debug/gity 2009-11-01 16:01:54.777 gity[10807:903] proxy: 2009-11-01 16:01:54.780 gity[10807:903] doc: 2009-11-01 16:01:54.784 gity[10807:903] *** Terminating app due to uncaught exception 'NSInvalidArgumentException', reason: '[NOTE: this exception originated in the server.] *** -[NSDistantObject methodSignatureForSelector:]: remote object does not recognize selector: 'isEqual:'' *** Call stack at first throw: ( 0 CoreFoundation 0x7fff854245a4 __exceptionPreprocess + 180 1 libobjc.A.dylib 0x7fff80d31313 objc_exception_throw + 45 2 Foundation 0x7fff83b21277 -[NSConnection sendInvocation:internal:] + 4304 3 CoreFoundation 0x7fff853f737c ___forwarding___ + 860 4 CoreFoundation 0x7fff853f3458 _CF_forwarding_prep_0 + 232 5 gity0x00011740 main + 471 6 gity0x0001134c start + 52 7 ??? 0x0001 0x0 + 1 ) terminate called after throwing an instance of 'NSException' Abort trap *third run (app is running, but no windows are open)* $ ./build/Debug/gity 2009-11-01 16:02:24.232 gity[10904:903] proxy: 2009-11-01 16:02:24.235 gity[10904:903] doc: 2009-11-01 16:02:24.315 gity[10904:903] *** Terminating app due to uncaught exception 'NSObjectInaccessibleException', reason: '[NOTE: this exception originated in the server.] NSDistantObject (0x101e2e030) is invalid (no connection)' *** Call stack at first throw: ( 0 CoreFoundation 0x7fff854245a4 __exceptionPreprocess + 180 1 libobjc.A.dylib 0x7fff80d31313 objc_exception_throw + 45 2 Foundation 0x7fff83b21277 -[NSConnection sendInvocation:internal:] + 4304 3 CoreFoundation 0x7fff853f737c ___forwarding___ + 860 4 CoreFoundation 0x7fff853f3458 _CF_forwarding_prep_0 + 232 5 gity0x00011740 main + 471 6 gity0x0001134c start + 52 7 ??? 0x0001 0x0 + 1 ) terminate called after throwing an instance of 'NSException' Abort trap So there are a couple things that I'm unsure of here. One of the errors is complaining that the distant object doesn't respond to isEqual. and another error is saying that the object is inaccessible, but the NSLog's above it clearly log out an object.. any help is appreciated. ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Re: Does [NSNetService getInputStream:outputStream:] really return retain objects and where does it say that
On Nov 2, 2009, at 12:51 AM, Jens Alfke wrote: Yes; but that doesn't necessarily mean the streams will go away soon if you don't retain them. An NSStream object is usually scheduled on a runloop, which keeps a reference to it as long as its open, so I believe it will stay around even if not explicitly retained by the app. True. I did one additional test in that I delayed opening the streams by 5 seconds (using performSelector:withObject:afterDelay:), giving the run loop enough time to release them if they were in fact auto- released (I know there's no guarantee they would actually be release in that period of time). The app doesn't crash and keeps working which indicates they are retained. Sure enough the author of the example project believes what he says because the project doesn't retain the streams anywhere. Does it release the streams? That would be the true test. My belief is that it can get away without retaining them, as long as it never accesses them again after they close; but if it explicitly releases them, it would crash. Yes, the example releases the streams. They're removed from the run loop first, the author apparently forgot to close them though (not sure if that can have an effect on the retain count). It appears that Apple already knows about this (see radar in comment), they just forgot to document it. How else would someone who writes example projects for the SDK know this specific function has this very special behavior - not that I know anything about who writes such examples. Regards Markus -- __ Markus Spoettl smime.p7s Description: S/MIME cryptographic signature ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Re: Editing in a tableview without selecting
On 1-Nov-09, at 6:01 PM, Graham Cox wrote: On 02/11/2009, at 4:42 AM, Ben Haller wrote: But where? I've been hunting for a -setEditable:YES method somewhere that I need to call, but I haven't found it... Hi Ben, Have you tried -setEditable:YES on the textfield cell you're using for the table column? (Inherited from NSCell). I'm not sure what its default state is, documentation doesn't say. Just tried it; makes no difference. I think I will have to construct a test case to see if I can reproduce this in a simple, isolated way. If that test case reproduces the problem, then I will post it somewhere and follow up on the list... Ben Haller Stick Software ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Update an NSMenu / NSMenuItem while it's being displayed on key down?
Hey all, quick question. What's the best way to go about updating an NSMenu that's currently being shown. For example, I have a button that triggers an NSMenu popup call and I want to detect the option key press and update some of the items in the menu. anyone done this? thanks ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Re: Update an NSMenu / NSMenuItem while it's being displayed on key down?
On Sun, Nov 1, 2009 at 7:31 PM, aaron smith wrote: > What's the best way to go about updating an NSMenu that's currently being > shown. For example, I have a button that triggers an NSMenu popup call and I > want to detect the option key press and update some of the items in the > menu. For this specific case, use menu alternates. In general, NSEvent has the information you want in the form of a modifiers-changed event. --Kyle Sluder ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Re: Editing in a tableview without selecting
On 02/11/2009, at 1:04 PM, Ben Haller wrote: I think I will have to construct a test case to see if I can reproduce this in a simple, isolated way. If that test case reproduces the problem, then I will post it somewhere and follow up on the list... Well, setting up a table view in IB, which takes but a few minutes, works correctly. The question you need to answer is a) what's different in your case and b) why not just use IB? --Graham ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Re: Editing in a tableview without selecting
On Sun, Nov 1, 2009 at 9:42 AM, Ben Haller wrote: > - (void)tableView:(NSTableView *)aTableView setObjectValue:(id)anObject > forTableColumn:(NSTableColumn *)aTableColumn row:(NSInteger)rowIndex You might want to double-check that you didn't typo this in your data source's @implementation. --Kyle Sluder ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Create an NSCharacterSet constant
Hi All, Is it possible to create an NSCharacterSet constant? Attempting to do so with the methods for creating an NSCharacterSet naturally fails to compile with "initializer element is not constant" I would ideally like to be able to do this. Is it best just to declare a global constant string, then initialize an NSCharacterSet during app initialization? It certainly seems that is the only real option. Am I missing something? If I am correct in what I am seeing, would there not be a valid use case for this as an extension of NSCharacterSet? (BTW, thanks GCC for the less than clear error message which should say "you cannot run functions or methods in a global initializer" or something better.) ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Re: Create an NSCharacterSet constant
On Sun, Nov 1, 2009 at 9:19 PM, John Joyce
wrote:
> Is it possible to create an NSCharacterSet constant?
No. By definition, constants are assigned values at compile time.
Objects other than strings can only be created at runtime. There is
no way around this fact.
> Is it best just to declare a global constant string, then initialize an
> NSCharacterSet during app initialization? It certainly seems that is the
> only real option.
Answers will vary. You could use a static global variable which is
assigned in +initialize, or you could use a static variable in a
category on NSCharacterSet, or a number of other things. Basically
what you really want is a singleton, not a constant.
I'm partial to the following construct:
+ (Foo *)sharedFoo {
static Foo *sharedFoo;
dispatch_once_t once;
dispatch_once(&once, ^{ sharedFoo = [[foo alloc] init]; }
return sharedfoo;
}
--Kyle Sluder
___
Cocoa-dev mailing list (Cocoa-dev@lists.apple.com)
Please do not post admin requests or moderator comments to the list.
Contact the moderators at cocoa-dev-admins(at)lists.apple.com
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com
This email sent to arch...@mail-archive.com
Re: Create an NSCharacterSet constant
On 02/11/2009, at 4:19 PM, John Joyce wrote:
Is it possible to create an NSCharacterSet constant?
Attempting to do so with the methods for creating an NSCharacterSet
naturally fails to compile with "initializer element is not constant"
I would ideally like to be able to do this.
Is it best just to declare a global constant string, then initialize
an NSCharacterSet during app initialization? It certainly seems that
is the only real option.
Am I missing something?
If I am correct in what I am seeing, would there not be a valid use
case for this as an extension of NSCharacterSet?
You can't create any non-scalar constant (except nil) using a static
initializer.
However, what you want to do is super-easy, and yes, it's a case for
an extension to NSCharacterSet. Just add a class method as a category:
@interface NSCharacterSet (MyCharacterSetConstant)
+ (NSCharacterSet*) theConstantCharacterSet;
@end
@implementation NSCharacterSet (MyCharacterSetConstant)
+ (NSCharacterSet*) theConstantCharacterSet
{
static NSCharacterSet* s_tccs = nil;
if( s_tccs == nil )
s_tccs = [[self characterSetWithCharactersInString:@"..."]
retain];
return s_tccs;
}
@end
--Graham
___
Cocoa-dev mailing list (Cocoa-dev@lists.apple.com)
Please do not post admin requests or moderator comments to the list.
Contact the moderators at cocoa-dev-admins(at)lists.apple.com
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com
This email sent to arch...@mail-archive.com
Re: Create an NSCharacterSet constant
On Sun, 1 Nov 2009 21:29:25 -0800, Kyle Sluder said:
>I'm partial to the following construct:
>
>+ (Foo *)sharedFoo {
> static Foo *sharedFoo;
> dispatch_once_t once;
> dispatch_once(&once, ^{ sharedFoo = [[foo alloc] init]; }
> return sharedfoo;
>}
Why (I'm really asking, not arguing; I don't know anything about GCD) is
this better than the traditional:
+ (Foo*) sharedFoo {
static Foo* sharedFoo = nil;
if (nil == sharedFoo) sharedFoo = [[Foo alloc] init];
return sharedFoo;
}
m.
--
matt neuburg, phd = m...@tidbits.com,
Re: Create an NSCharacterSet constant
On Nov 1, 2009, at 11:32 PM, Graham Cox wrote:
On 02/11/2009, at 4:19 PM, John Joyce wrote:
Is it possible to create an NSCharacterSet constant?
Attempting to do so with the methods for creating an NSCharacterSet
naturally fails to compile with "initializer element is not constant"
I would ideally like to be able to do this.
Is it best just to declare a global constant string, then
initialize an NSCharacterSet during app initialization? It
certainly seems that is the only real option.
Am I missing something?
If I am correct in what I am seeing, would there not be a valid use
case for this as an extension of NSCharacterSet?
You can't create any non-scalar constant (except nil) using a static
initializer.
However, what you want to do is super-easy, and yes, it's a case for
an extension to NSCharacterSet. Just add a class method as a category:
@interface NSCharacterSet (MyCharacterSetConstant)
+ (NSCharacterSet*) theConstantCharacterSet;
@end
@implementation NSCharacterSet (MyCharacterSetConstant)
+ (NSCharacterSet*) theConstantCharacterSet
{
static NSCharacterSet* s_tccs = nil;
if( s_tccs == nil )
s_tccs = [[self characterSetWithCharactersInString:@"..."]
retain];
return s_tccs;
}
@end
--Graham
Wow, you all responded wonderfully and super fast!
I had accidentally sent it to the list hours ago from the wrong email,
so it was naturally rejected, then I just now sent this.
Thanks to everyone, when I get a block of time I'll dive into each
approach and see how they all feel.
Anyway, I digress...
Category, yeah, that's totally one way I thought about, but I was
thinking more of an enhancement request in Radar. I figured the nature
of it inheriting from NSObject implies there are some kind of
optimizations under the hood (otherwise we'd just use NSString and
NSScanner right?).
Seems like a useful one considering the
characterSetWithCharactersInString: method. Even if it were some sort
of singleton class. Particularly useful in input validation routines
(things like highly structured serial numbers and other identifiers,
etc...)
If there is consensus from multiple people on an enhancement request
and the functionality and justification for it, I'll be happy to file
it.
___
Cocoa-dev mailing list (Cocoa-dev@lists.apple.com)
Please do not post admin requests or moderator comments to the list.
Contact the moderators at cocoa-dev-admins(at)lists.apple.com
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com
This email sent to arch...@mail-archive.com
Re: Create an NSCharacterSet constant
On Sun, Nov 1, 2009 at 9:49 PM, Matt Neuburg wrote: > Why (I'm really asking, not arguing; I don't know anything about GCD) is > this better than the traditional: It avoids a race condition if +sharedFoo is called simultaneously from different threads, and it's wicked fast (Cmd-DoubleClick dispatch_once to see why). It's also got "once" in the name, and is roughly the same shape as the traditional construction, so switching to it entails minimal mental burden. --Kyle Sluder ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Re: Create an NSCharacterSet constant
On Sun, Nov 1, 2009 at 9:50 PM, John Joyce wrote: > Category, yeah, that's totally one way I thought about, but I was thinking > more of an enhancement request in Radar. I figured the nature of it > inheriting from NSObject implies there are some kind of optimizations under > the hood (otherwise we'd just use NSString and NSScanner right?). An NSCharacterSet embodies the concept of "a set of characters." An NSString and NSScanner pair embodies the concept of "how I'm gonna find some characters in a string." The first is a value type, the second an implementation detail. NSCharacterSets are also useful for other things than just finding them in a string or other stream of data. > Seems like a useful one considering the characterSetWithCharactersInString: > method. Even if it were some sort of singleton class. Particularly useful in > input validation routines (things like highly structured serial numbers and > other identifiers, etc...) I'm confused about what enhancement could be made that would not be highly specific to your application. --Kyle Sluder ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Re: Create an NSCharacterSet constant
On 02/11/2009, at 4:50 PM, John Joyce wrote: Category, yeah, that's totally one way I thought about, but I was thinking more of an enhancement request in Radar. I figured the nature of it inheriting from NSObject implies there are some kind of optimizations under the hood (otherwise we'd just use NSString and NSScanner right?). Seems like a useful one considering the characterSetWithCharactersInString: method. Even if it were some sort of singleton class. Particularly useful in input validation routines (things like highly structured serial numbers and other identifiers, etc...) If there is consensus from multiple people on an enhancement request and the functionality and justification for it, I'll be happy to file it. What is your proposed character set going to contain? Is it of ultra- widespread utility? Otherwise, this is what categories are for - to add enhancements yourself in a few moments instead of waiting (possibly years) for Apple to do it for you. I don't really see that there is a case here for a radar enhancement request, but it would depend on what you have in mind. As for optimisations, I don't quite follow your argument - NSCharacterSet is not equivalent to NScanner + NSString, though it sometimes will get used with those classes. I expect the common built- in "constant" sets are there because they are of very widespread utility - e.g. stripping whitespace, but are probably not using much in the way of magic optimisations. --Graham ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Re: Create an NSCharacterSet constant
On Nov 2, 2009, at 12:04 AM, Graham Cox wrote: On 02/11/2009, at 4:50 PM, John Joyce wrote: Category, yeah, that's totally one way I thought about, but I was thinking more of an enhancement request in Radar. I figured the nature of it inheriting from NSObject implies there are some kind of optimizations under the hood (otherwise we'd just use NSString and NSScanner right?). Seems like a useful one considering the characterSetWithCharactersInString: method. Even if it were some sort of singleton class. Particularly useful in input validation routines (things like highly structured serial numbers and other identifiers, etc...) If there is consensus from multiple people on an enhancement request and the functionality and justification for it, I'll be happy to file it. What is your proposed character set going to contain? Is it of ultra- widespread utility? Otherwise, this is what categories are for - to add enhancements yourself in a few moments instead of waiting (possibly years) for Apple to do it for you. I don't really see that there is a case here for a radar enhancement request, but it would depend on what you have in mind. As for optimisations, I don't quite follow your argument - NSCharacterSet is not equivalent to NScanner + NSString, though it sometimes will get used with those classes. I expect the common built-in "constant" sets are there because they are of very widespread utility - e.g. stripping whitespace, but are probably not using much in the way of magic optimisations. --Graham Indeed this is the point of Categories. ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Re: Editing in a tableview without selecting
On Sun, Nov 1, 2009 at 9:42 AM, Ben Haller wrote:\ > The only unusual thing about this tableview is that I set it up in code > instead of in IB, so I assume the root of the problem is in my programmatic > setup code. But where? I've been hunting for a -setEditable:YES method > somewhere that I need to call, but I haven't found it... Oh! You forgot to call -setEditable: on the table columns. :) --Kyle Sluder ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Re: Create an NSCharacterSet constant
On Nov 1, 2009, at 11:59 PM, Kyle Sluder wrote: On Sun, Nov 1, 2009 at 9:50 PM, John Joyce wrote: Category, yeah, that's totally one way I thought about, but I was thinking more of an enhancement request in Radar. I figured the nature of it inheriting from NSObject implies there are some kind of optimizations under the hood (otherwise we'd just use NSString and NSScanner right?). An NSCharacterSet embodies the concept of "a set of characters." An NSString and NSScanner pair embodies the concept of "how I'm gonna find some characters in a string." The first is a value type, the second an implementation detail. NSCharacterSets are also useful for other things than just finding them in a string or other stream of data. Seems like a useful one considering the characterSetWithCharactersInString: method. Even if it were some sort of singleton class. Particularly useful in input validation routines (things like highly structured serial numbers and other identifiers, etc...) I'm confused about what enhancement could be made that would not be highly specific to your application. --Kyle Sluder Not my application per se, but for more applications. I see it as a useful thing to have predefined character sets that don't need to be managed. It's still just as simple to create them, but it sure would be nifty to not have to and to manage them in a constants file. At the very least, I will probably submit a radar on adding locale- specific standard sets. The standard set is woefully Latin-centric. Many languages do not use whitespace the same way (generally as word boundaries). Many languages also do not have the concept of capitalization. Japanese is one in which I happen to work. There would be a multitude of sets that would make sense in Japanese for a multitude of super common purposes. I know, some of this leads to things like tokenization. Point being, if we look beyond English and Western Europe, and consider how much of programming is pushing text around, it certainly seems like a logical expansion that would be of benefit to a lot of people. Maybe it's just me, not trying to hog the list. If anybody has some good sets for consideration, I'll be happy to include them when I create that radar. Beyond that, I'd be willing to produce and host open source sets for use, if anybody wants to contribute them (( send them off list )). ( anything locale-specific should ideally include localized documentation as well ) ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Re: objectAtIndex
Thank you, that answers my question, Nava On Nov 1, 2009, at 10:24 PM, Sherm Pendley wrote: On Sun, Nov 1, 2009 at 2:44 PM, Nava Carmon wrote: NSObject *anObject = [anArray objectAtIndex:i]; ... Should I release it? No. You didn't create anObject with +alloc or a method beginning with -copy, nor did you send it a -retain message. Therefore you do not own it and must not release it. sherm-- -- Cocoa programming in Perl: http://camelbones.sourceforge.net ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
Re: Create an NSCharacterSet constant
On Sun, Nov 1, 2009 at 11:03 PM, John Joyce wrote: > I see it as a useful thing to have predefined character sets that don't need > to be managed. So you just want more things like +whitespaceCharacterSet? That makes sense. > At the very least, I will probably submit a radar on adding locale-specific > standard sets. > The standard set is woefully Latin-centric. > Many languages do not use whitespace the same way (generally as word > boundaries). > [snip] > I know, some of this leads to things like tokenization. Well, in that specific instance, you should not be doing word splitting yourself. You should instead use CFStringTokenizer. > Point being, if we look beyond English and Western Europe, and consider how > much of programming is pushing text around, it certainly seems like a > logical expansion that would be of benefit to a lot of people. > Maybe it's just me, not trying to hog the list. At some point you have to decide that it isn't worth adding more API. I certainly would not want NSCharacterSet to expose the power set of all Unicode characters. --Kyle Sluder ___ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
