I have just tried downloading the new version from http://foobar.bowron.us/ and though the modified timestamp on foo_playlist_tree.zip has changed, the files inside and the changelog still seem to be version 1.7.9

Try it again... I re ran my uploading script and downloaded it and it looks like 1.7.10 is on there now... Thanks for the heads up...

Thanks for that "remove duplicates when sending / adding to a playlist".

I just want to let you know that duplicates are still sent to the browser.

For now, I'm not a browser user but that @browse_as<> may be quite useful.

Let me ask this one:

- does the depth (number of levels) of a tree affect performance (loading/refreshing time) or it is just the number of records x the numbers of queries that makes my FB taking 25 secs at start/refresh?

If the depth matters, will the presence of the browser matters too?

I mean, with the new dynamic browsing we can see the browser as a tree level or even as more than one so it would be useful which set up would work faster.

Thanks.

Is is possible to hide duplicate entries in the browser. e.g I want to list albums and have one row for each album rather than it repeated for every database entry for the album.

Hello, I've retryed this query with the last version, and now, i have with:
Query: artist IS "Ravi Shankar"
[x] foo_playlist_gen format
Format: $upper($cut(@any<album artist,artist>,1))|@any<album artist,artist>|%album%|%title%

I have files with :
- ARTIST = Ravi Shankar
- ARTIST = Ravi Shankar\Ali Akbar Khan
- ARTIST = Zakir Hussain ; ALBUM ARTIST = Ravi Shankar

The result displayed :
A/Ali Akbar Khan/From India
A/From India
R/From India
R/Ravi Shankar/From India
R/Ravi Shankar/Rough Guide To Ravi Shankar
R/Rough Guide To Ravi Shankar

I have no entry on Z for Zakir Hussain;
I continue having albums under the two levels.

If i remove the first level display (=first leter), i don't have Zakir Hussain too.

What I logically hope to have is something like this:
A/Ali Akbar Khan/From India
R/Ravi Shankar/From India
R/Ravi Shankar/Rough Guide To Ravi Shankar
Z/Zakir Hussain/Rough Guide To Ravi Shankar

So, is it a problem in my request or in the @any function?

Have you tried amending the query to something along the lines of:



It looks like based on your query and your file that the file in question does not match the query.

OK, your right!

Sorry for this stupid error.

But, i still having albums displayed in the two levels.

In fact, I still having exactly the same problem than in previous versions...

With this formating (same files, and functionning query) :

$upper($cut(@any<album artist,artist>,1))|
@any<album artist,artist>|
[%date%' - ']%album%|$num(%tracknumber%,2)' - '%title%

I have :
R\Rough Guide To Ravi Shankar
R\Ravi Shankar\Rough Guide To Ravi Shankar
R\Zakir Hussain\Rough Guide To Ravi Shankar
Z\Ravi Shankar\Rough Guide To Ravi Shankar
Z\Zakir Hussain\Rough Guide To Ravi Shankar

I understand why I have "Ravi Shankar" under Z and "Zakir Hussain" under R, but how to do? (I've tryed some stricmp onto @parent and $cut(@any<album artist,artist>,1) but, donsn't work).
And albums displayed in the first level is it a bug?


edit: one interresting thing: under Z (so when i have an ARTIST and an ALBUM ARTIST, and when is the ARTIST matching) I don't have the album displayed in the first level (so, it's working as expected in this precise case).

RELEASE 1.7.11

@parent<>, @query<> can be used in library name

TAGZ allowed in number value for @parent<> (i.e. you can do something like $sub(%_nestlevel%,3))

#tag# literally replaced by $meta(tag,1), $meta(tag,2), then formatted
- fixes some issues people were having (I'm looking at you fabiospark)

globals back in tree. No more funky @globalformat...
- can be used in labels with a @format<> or @function that will perform formatting
- can be used in a query format
- can be used in a playlist_gen type query in a @format<>

globals now formatted in browser... be sure to use | to break variables and use a | at the end
of your globals.

Mmmh..

Trying this one:



I'm still having all the performers under the group "pippo", even the (voice - woman) lot. And no "women" group.

Did I understand something wrong, didn't I?

$stricmp checks whether strings are equal ignoring case.

You want $strstr - finds whether a string contains another string.

Sorry. That was just stupid by me thinking to stricmp as an HAS instead as of an IS.

Sorry again for that.

----------

Of course, this one splits the groups:



but, under "groups" I'm still getting ALL the other performers, even those with no "voice" in the field (but with at least one voice in the whole %performer%).

So I'm still to having to wrap the whole thing with an "$if(strstr(#performer#,voice)......)", but, wouldn't this be the work that "performer HAS voice" in the Query box is meant for?

the "Performer HAS voice" in the query will filter out all the files that do not have a voice performer, but the #performer# in the format field will loop through all the PERFORMER tags.

coulpe of question, they may be stupid?

bare with me im at work and dont have foobar on this pc so it may be hard to explain.

1. if i want to view all my music in the structure it is on my hard drive i can add it to a folder but that then becomes static. is there a way to have it dynamically so it updates itself?

2. is it possible to add your CD rom drives as nodes? i have then in there at the mo (dragged them from My computer into a folder), but then when u insert a cd, you have to expand the node, and then expand another folder before you get to the contents.

it would be cool if you could just add folders (i mean add as in navigate to folders, select from a open dialog window, not drag, if that makes sense)

Dynamically in the sense that when you restart foobar2000 it gets refreshed, yes.
Dynamically in the sense that it reflects the directory tree whenever you change a file, no.

When you drop a folder, if you hold down control, rather than create a folder, it will create a query with a @drop<> command. Then whenever the query is refreshed, it will look in the directory for updated files.

I set up my files in the directory structure like this:

E:\Music\B\Bob Dylan\2005 - No Direction Home Soundtrack

so I set up my tree with drop commands for each letter of the alphabet (@drop<E:\Music\B>). I could also set up a drop like @drop<E:\Music> but I have some extra cruft in that directory that I want to exclude.



You could also set up a @drop query for your cd drives, but that will still require you to refresh the query manually when you change cds.

You may want to look into foo_uie_explorer

Thanks for the explanation. I will live with some "ignore me" nodes under some query.
By the way: is there a way to really hide a subnode? I tried with hide/unhide but the label is still there. (Well, I could hide the cross and put a blank in its name... I could be an easy solution.)

--------------

Sorry to touch that key again, but I'm still unable to make the @drop<> of my playlist folder work: any clue? I remind you I'm using fpl files but it doesn't work with m3u too.

If I load them or import drag them from foo_explorer they work.

You can really hide hidden nodes under the menu under components/playlist tree/settings/show hidden nodes.

Are there any unicode characters in the path to the fpl file?

Do you have ignore nested playlists checked under config?

With jazz it is usual having the same tune many times, so I'd like to have a tree like that:



sorted from the higher number to the lower.

Is there a way to get this?

Thanks.

You're lucky that I have been working on my covers/originals playlist and after reading your post I thought it might be interesting to sort it based on the number of tracks in each folder.

Yes it is possible.

Here is how I set mine up:

Format: %_trimmed_title% '(@format<%_itemcount%>)'|%artist%

Display Sort Order: $if(%_isleaf%,%_name%,$sub(999,%_itemcount%)%_name%))

%_trimmed_title% is just a little global I have that strips of potential parenthesis and $trims it.

Is there a way to extract only the items counts and have them as groups?




I mean something as giving an argument to itemcount without having to show it.
By the way: which is (are) the correct syntax of %_itemcount%? Does it count the items on its same level? How can it know which items to count?

Thanks.

Would you consider adding an "Apply" button to the query edit window? It would be quite useful when testing new functions and their syntax.

An maybe a quick way to export the text of the whole query in a single txt file. And the import too, of course.

And, why not an undo?.

Thanks.

Not currently, because the itemcount gets calculated after the tree has been filled. What would be necessary for it to work without displaying anything else is a way to specifiy criteria for a folder, but not have that criteria actually be displayed. It could be done by adding a new local function, maybe @noshow<> or something....



%_itemcount% is variable that exists for folders, and returns the number of tracks that are contained in it and all it subfolders.

The trick for this is that %_itemcount% really isnt known until the tree has been totally filled with the tracks, then %_itemcount% gets evaluated when the tree is drawn and/or sorted. This works with the newer version of PT that support the display sort order because that sort is done *AFTER* the tree has been totally populated.

You could always remove all the children of the query, then right click and File\Save This as a collection. Its a text based file, with all the query info, which can be imported.

RELEASE 1.7.12

faster context menu

some changes to loading procedure

option: config::Convert Old style tags to new style automatically
- turning it off will turn off conversion of $tag() to new style @tag<> and should be faster

changed query tree population - fewer string copies == hopefully faster...

@ignore tag... if a query format contains "@ignore " after being formatted it will not be
added to the tree. must be followed by a space...
- for example if you have multiple genres tags but you do not want to generate a rap genre
folder you could use this format:
$if($strstr(#genre#,rap),@ignore ,#genre#)|%album%|%title%

@hidden tag. hides a node. can be used in a query to specify that one of the nodes created
should not really be shown in the tree... (there must be a space following @hidden)

@noshow<> hides all the text between it in a folder name. Allows you to split a query into
different folders without actually displaying the split criteria...

Thanks for the update. Nice work.

I tried unchecking that "auto convert tags etc..." but when I click on a node or a subnode its label/title disappear leaving a "?". After checking tha box again and after a refresh I get the labels back. I don't have any code in the label box, nor browse or icon or else.

-----------------

Do you mind explaining a bit more about the sort boxes?
I mean: exact sintax, allowed functions and, maybe, some tricks.

In the meantime let me show a little trick of mine.

For instance, I have this query:


With this sorting I get:


Under each subnode but the "many" one I have the %instrument% content sorted alphabetically.

If I put


I get the %instrument% content correctly sorted under "many" too but I get the first level (Solo, Duo etc) alphabetically sorted too : how to force the content of "many" to be correctly sorted too?

I tried some different Pop and disp sorting strings but couldn't come across the right one.

Then you came out with that useful @noshow<>.....
So, after a quick test this is my new (perfectly working) string:



I think this could be of general use when people want to create a custom "manual" sorting.

If you know a more straightforward way (or a more elegant) please let me know.

Thanks.

(One day we should embark into a discussion about performances:
- better three separate queries or a @multiple three headed one?
etc, etc)
Thanks.

Somewhere in your configurations you probably have the old style %name% tag instead of %_name%. (and similarly for other tags) Look under the preferences under folders.

I cannot go into exact syntax, etc. but this is how query population works

1 - get all matching tracks

2 - sort tracks by Population Sort Order if specified

3 - remove tracks that exceed maximum if specified

4 - For each track:

4 a - apply formatting to each item in the list in order

4 b - add item to tree based on formatting results, splitting each | into a seperate folder.

5 - After entire tree is made, sort by Display Sort Order if specified.

Because the tracks are sorted in step 2, before any tree is created only functions local to the track can be used as a sort order here.

Because at step 5 the entire tree has been populated and every folder has its contents, you can use folder level functions in the display sort order (%_itemcount%, etc).

I have a problem, which, as I found, was described in a year-old problems list for foo_playlist_tree:
• invalid treatment of subsong index:
The subsong index gets discarded when saving .col files. This breaks some (container) formats that allow multiple songs in one file, including Ogg Vorbis, Matroska and cue sheets.
All versions of foo_playlist_tree that I used, including 1.7.12, do this. Many of my albums are in APE+CUE format, and I cannot access any of them through the tree because of this error. Will it ever be fixed?

Thanks.

The problem you refer to of not saving the subsong index in collection files, was fixed. According to the changelog it was December 12, 2004, prior to the release of 1.0.0. Something else funny could be going on though.

Do the items show up correctly if you were to drag and drop the files? Does the problem only occur after load a tree?

If you can find a small example to email me or put something up somewhere where I can download it, I will take a look.

On further investigation, the indices are stripped on reading, rather than writing.

I drug a cue file into the tree, and all the songs show up properly. I then save it - so far so good. But when I reopen, it stops working. I save again, and see the indices missing.

If you fix this, the plugin will be flawless, for me. Thanks for all the work.

What it looks like is happening is that when I was cleaning up some of the write_leaf code, I inadvertantly made it write two dividing characters where there should be only one. I will send you a preview release of 1.7.13 via PM and you can let me know if that fixes it.

Thanks.

That seems to be just it. Problem solved.

RELEASE 1.7.13

fixed double \b\b error in write_leaf (subsong index fix)

@fakelevel directive. Allows you to use a folder to limit the number of subitems, etc but
the level is not displayed and all its children are displayed as children of its parent
node.

Example:
Name: Random Beatles Album (@first<album>)
Query: artist HAS beatles
Format: @fakelevel %album%|$num(%tracknumber%,2) - %title%
Max: 1 subfolder
PSO: $rand()
DSO: %_name%

This query gets a random beatles album, and each title is displayed directly under the query
node.

Dman, does that means that this will get a random album and displayed in the correct order(track1, track2, track 3, etc...)? Sorry if this is obvious, but I don't have my PC working the last months so I'm a little out in the development of PT...

I've been trying to wrap my head around this:-

Basically I want a query to list songs that have not been played in over 30 days from the system date, using last_played.

My last_played field is formatted like this: YYYY-MM-DD HH:MM:SS

*scratches head*

Can anyone help me with this?

Cheers.

Update: I've got this atm, but it's still missing something:-

Since I have moved on to the more recent (1.7.12) version of the playlist tree (had been using 1.7.7), it takes a very long time to refresh queries, even after editing existing queries. Also, when I try to drag and drop queries within the tree, it locks up fb2k.

Any thoughts on why this might be happening? Let me know if there is anything I need to post, in terms of settings.

Thanks.

Yep.

When you say it locks up... You mean that its doing some processing and you cannot do anything while its processing it, but after some time it goes back to normal right?

I've been working on some of the stuff with queries... some of it has made it faster, some of it has made it slower, but more correct.

What sorts of queries are you using?

It did go back to normal, but it took a few minutes for that to happen. The queries I have are fairly basic, and would refresh very quickly with the older plugin I had been using. To note, my library does have over 2900 files of FLAC and MP3 each. My music library hasn't changed for quite some time.

As an example, here is one of the queries, labeled "Album by Year"
Query

Format


The Population Sort Order is "@default", and the foo_playlist_gen format box is checked.

Even if I do not make any changes to the query, but still click OK to close the edit window, it will take a minute or so to refresh the query. If I drag and drop a query to a different position my tree, then it takes even longer to refresh.

Whenever you click OK it refreshes the tree... If you dont make any changes, click cancel and it will not refresh...

I dont think there is anything going wrong, per se, but its just taking a long time to refresh the query... I am looking for ways to decrease refresh time in general, so hopefully the times will decrease in the future... if the time is unbearable, revert to an older version, they are available from the website...

Also you might try instead of using a population sort order of @default, leaving Population sort order blank and setting the display sort order to %_name%... it might help, i dont know...

What exactly is the difference between Population Sort order and Display sort order?

Population sort order happens before the tree is created.

Display Sort Order happens after the tree has been populated, but before it is shown.

See this post: http://www.hydrogenaudio.org/forums/index....ndpost&p=334427 for more info.

good plugin

i made this:



with the beautyful icons of everaldo.com
http://www.everaldo.com/crystal.html

hope you like it

Hello.

I continue having problems with my (impossible ?) request...

format : $upper($cut(#artist#,1))|#artist#|%album%' ('%date%')'|[%disc%.]$num(%tracknumber%,2) - %title%

I've puted out the ALBUM ARTIST field, and now, i'm sorting just on ARTIST, but i have a problem with display sorting, it seems to be unworking with this format.
Also with simple criterias like %artist% or #artist# the displayed order seems to be totally randomed.

Chris, just a quick check: does the "#tag#" function work with "quicktag SQL" fields?

To me it seems it doesn't.

I also tried the "tag" function in "albumlist" and it doesn't too.

Surely they are related.

Thanks.

What are you using for the Display Sort Order? I would think that %_name% should work.

No, I dont believe that it will.

Ok, that works well with %_name%, thank you.

how can i get rid of the white background in windows controls (these [+] [-] things there)? any help?

It looks like that's the fault of the graphics from your Windows theme (not using a transparent background). The theme I'm using (VXP) doesn't have this problem:

Hi,

Is there a known limit for the maximum megs allowable in a query? Once I get much over 2000, the results are unpredictable. eg, a limit of a 2040 gives me 2038 MB, while a limit of 2050 returns nothing. Incidentally, much higher values usually give me a figure between 1300 and 1900 MB.


thanks.