AltME groups: search

Cool James, but it can be yet more elegant, using a single iterated 
face instead of four (one for each column), but it's not a concern 
for your app, as you have a fixed, small number of columns (4).  
If it was a grid of large or "infinite" width, then you would better 
program the list's pane function yourself.

Cool. Searching for an my old drag-drop-demo now :) And the button-list 
is cool too btw :) Maybe that complete state-saving should be supported 
by rebol, and focus could include a focal-index .

Someone had contacted me a month or so ago and asked for my modified 
text-list and was looking to make some updates.  I don't recall who 
it was but if your here - can you tell me if you improved such.

I'm working on a different implementation with a more advanced list 
system instead

Ok, I've tried... I have a list of toggles with a scroller attached. 
For the life of me I can't get the "dragger" to size properly. The 
slider button is not proportional to the what is being shown in the 
list. In my case, I'm seeing about 20 of the 24 rows so the dragger 
should be about 80% the size of the panel, no? Thanks in advance 
to anyone who knows.

I'm expecting you're doing a show every time there is a scroller 
movement. What you want to do is check if the current row (ie. the 
one at the top of your list) has changed If so, show the list. If 
not, ignore. Not all the shows are necessary.

James, this example might help:

do http://www.lexicon.net/antonr/rebol/gui/iterated/demo-vid-list-scroller.r

I am sure Cyphre has this on his to do list already.

Generic newbie-like question: I was spying on source request-dir 
and am baffled as to how it works? There are many function names, 
I can't seem to access directly, such as F-LIST and DIROUT. Where 
do they come from?

request-dir: func [
    "Requests a directory using a popup list."
    /title "Change heading on request." title-line
    /dir "Set starting directory" where [file!]
    /keep "Keep previous directory path"
    /offset xy
][
    if block? dirout [
        dirout: layout dirout
        max-dirs: to-integer f-list/size/y - 4 / f-txt/size/y
        center-face dirout
    ]
    set-face f-title any [title-line "Select a directory:"]
    if not all [keep path] [
        path: copy either where [clean-path where] [what-dir]
    ]

    if all [not empty? path slash = last path] [remove back tail path]
    last-path: path
    result: none
    show-dir
    either offset [inform/offset dirout xy] [inform dirout]
    result
]

REBOL [
  title: "Multi-File Search"
  author: "Henrik Mikael Kristensen"
]

search-function: func [/local files results] [
  results: copy []
  status/color: red
  show status
  nof-files: length? files: read to-file get-face search-dir
  forall files [
    if not dir? first files [
      data: read/lines first files
      status/text: join "Searching: " first files
      set-face status-progress divide index? files nof-files  
      show [status status-progress]
      forall data [
        if found? find/any first data get-face search-string [

          insert tail results reduce [first files index? data first data]
        ]
      ]
    ]
  ]
  status/color: gray
  results-list/data: results
  status/text: reform ["Results:" length? results "found"]
  show [results-list status]
]

main: layout [
  style header button 100x20 gray edge [size: 1x1]
  backdrop effect [gradient 0x1 200.200.200 180.180.180]
  across space 2 origin 8

  label white gray 300x20 "Search Multiple Files" font [align: 'center 
  size: 14] return
  search-string: field (as-pair 228 22) btn 70 "Search" [
    if not empty? search-string/text [search-function]
  ] return

  search-dir: field (as-pair 228 22) btn 70 "Directory..." [set-face 
  search-dir request-dir] return

  status: label white gray 300x20 "Results" font [align: 'center size: 
  14] return
  space 0x0

  header "File" header 50 "Line" header 132 "Text" box gray 18x20 return
  results-list: text-list 300x200 data [] [print "test"] return
  status-progress: progress gray white 300x10
]

set-face search-dir %.

inform center-face main

crude prototype, but it helped me find the file I had been looking 
for :-) I don't know how  to do actual multi columns with the internal 
text-list. the plan was to allow scrolling (I can't get that working) 
and when you click on a search result, the file pops up in the viewtop 
editor. feel free to perfect it...

I suppose a "Save" feature is getting close to the top of the list.

Anybody here with already done notes on text-list internals (lc, 
sn, cnt...) ???

Hello, anyone have any ideas how I might improve on the integrity 
of a program that sends email to a list of people. I just realized 
I should be using '/only' but I have to check its behavior as I don't 
want each person to get a list of the other recipients. My problem 
is I often get a time out and I want to trap that. Yesterday during 
my first test it got halfway through the list and errored out which 
then caused the program to exit. I figured I should at least keep 
a running log so I can go back.

Pekr, it actually didn't do a list. Instead it sent individual emails. 
Also I probed send and there is a '/show'' option that will replace 
the to: address. checking...

Anyway, setting header/to: none works but it puts the entire list 
there as you mentioned. Hmmm, maybe I have to do it individually.

I've been grumbling over a few observations I did on early versions 
of Canvas and ROAM regarding SHOW on three different machines: A 
slow PC linux laptop, my fast Windows PC with a Radeon 9500 graphics 
card and my mac mini with medium speed Radeon 9200 graphics chip. 
For my tests I used the ROAM object browser available in the desktop 
under Rebol/REBOL Tools.


When I maximize ROAM to full screen on my laptop and mac, they are 
rather slow at redrawing the highlight bar. But if I move the mouse 
quickly to the top/bottom of the list, the highlight skips those 
entries that can't be highlighted in time, and therefore it remains 
fairly responsive even if the frame rate is low.

However on my fastest machine with my powerful Radeon 9500, the highlight 
bar moves at a snails pace behind the mouse pointer, insisting on 
redrawing every highlight. It takes at least a full second to reach 
the current mouse position and the event system is locked and REBOL 
is fully concentrating on SHOWing the list face.


The amount of time SHOW takes to display the list view is dramatically 
dependent on the size of the window. A small 400x400 window is fast 
enough for normal use, but a 1280x1024 window is very slow.


It looks to me like a frame rate problem: Somehow my fast machine 
calculates a specific very high framerate (say 50 fps) that SHOW 
should handle for the list face, but can't keep up at all. Therefore 
50 SHOW operations take 2-3 seconds longer than they should and the 
delay occurs and it drowns out the event handling. This framerate 
is apparently tuned properly on the Linux and the Mac so it never 
becomes a problem there. How is this framerate calculated?


Also when I run the console benchmark program, I get remarkably bad 
performance on text output on my normally fast Windows machine, while 
the mac and linux consoles output text 5-10 times faster. Reports 
from other Windows users say my console is very slow.


I'm really just wondering if others with Radeon graphics cards and 
Catalyst drivers have similar problems.


Not all is bad: This framerate problem can be solved within REBOL. 
I did a simple delay system for Canvas which sped things up quite 
a bit on my machine and I've just done a dynamic method for ROAM. 
The delay system simply implements a DELAY-SHOW function which does 
not invoke SHOW unless a specific amount of time has passed. The 
method I did for ROAM is dynamic in that it measures how long a SHOW 
takes and adjusts the delay accordingly plus a tiny safety margin 
of 0.01 seconds. It works quite well. At the cost of a reduced framerate 
and sometimes missing a redraw at the end of a mouse move, I get 
a much more responsive ROAM, and the framerate adjusts nicely between 
large and small windows.

See if you can tell the difference between:

http://www.rebol.com/view/tools/roam.r
and
http://www.hmkdesign.dk/rebol/roam.r

A couple of questions: Is there a switch to start view in the console 
mode (CLI)? I know I can do something in the user.r but I'd like 
to do it from a prompt on demand. And also, anyone have any idea 
why my emails sent via a rebol program take so long to arrive. Any 
special settings in the header or hints. I 've created a program 
that sends emails to a list but I notice that it takes a long time 
(30 minutes or more) to receive the email. I checked by sending from 
Outlook and through the command line in rebol and it  was instantaneous. 
There's something happening and I wonder if any on you  have seen 
this.

Purpose: {Mimic DO EVENT (which is called in wake-event) with a mezzanine 
function of our own.
	The mimicry is good for two things:
	1) as a document of how DO EVENT works inside

 2) as a way of achieving crazy magical tricks not otherwise possible, 
 like 
	   - faces transparent to events

    - restoring and saving facets of iterated faces before and after 
    events are sent to them 
	     (see shell-list.r)
}

Currently shell-list, with the help of mimic-do-event, is iterating 
buttons and toggles, and they operate just like standalone buttons 
and toggles. If you've tried putting buttons and toggles in a list 
you know it is not easy to achieve.

I've just submitted a patch to RAMBO which should allow progress, 
slider and scroller to also be iterated without problems.

Gregg: probably you noticed it already, but when one presses the 
up or down arrow buttons while in the search field in the Word Browser 
it crashes with the following statement:


** Script Error: index? expected series argument of type: series 
port
** Where: pick-next
** Near: sync-funcs-list index? f
show-word first
>>

so it's not possible to navigate the words via the coursor keys

Here is SHELL-LIST showing the VID styles BUTTON BTN TOGGLE and TOG, 
iterated without modification of them:

do http://www.lexicon.net/antonr/rebol/gui/demo-shell-list.r

site: select load-thru http://www.reboltech.com/index.r[folder "Anton"]
clear find site %index.r
foreach url [
	site/gui/mimic-do-event.r
	site/library/window.r
	site/library/stack.r
	site/gui/shell-list.r
	site/gui/percent-progress.r
][
	read-thru/update do url
]
do http://www.lexicon.net/antonr/rebol/gui/demo-shell-list.r

... and which is on hold because I wanted to use shell-list for it.....

My original thought process was that I don't want to automate the 
deleting of files in the cache, on my computer or someone else's. 
I think if the user wants to delete their files, they should do it 
themselves. I don't want to get involved in any data loss.

So I thought it preferable to do an intelligent update. But for that 
I need an intelligent interface... -> shell-list.

Shell-list is designed to be highly flexible and configurable, for 
maximum code re-use. I didn't like the thought of writing a new custom 
list for every program.

Note to everyone: shell-list is nowhere near finished, but the hook 
system I use to iterate those styles is so significant to me I have 
to make a demo of it.

What I need, to continue with shell-list, is some extra hooks in 
the View system. I need to be able to hook each call to feel/redraw. 
This means a hook in SHOW and probably other places too.

I might also need access to view system's internal "timer-list" - 
this is the list of faces with rates which need time events sent. 
I at least need to know when this list changes, which as far as I 
know only happens when SHOW is called.

Yes, the demo is not very impressive looking, I admit. There are 
no flashing lights or fancy graphics.

(though I can easily add some interesting variations for each iteration.)

It doesn' t look any different from a window with some normal buttons 
or toggles in it.

But, in a way, that's the point. You, the user of shell-list, should 
not have to care at all what you throw into the iterated-layout. 
It should just work as it would inside a normal layout. That's the 
beauty. The buttons look and interact like normal buttons.

Good question. The idea of shell-list is to make things simpler. 
It should be *easy* to make lists of things.
Why do I publish it now, when it's still in such an early stage ?

I guess I am hoping for a bit of attention. I will need some changes 
to the View system or it will not be able to work perfectly. Right 
now seems not a good time with rebcode just released, but this seems 
so fundamental to user interfaces to me I think it's more important. 
Everything always seems to lead back to the list.

Although, maybe it could. The facets I have discovered (and will 
continue to discover) which are used in shell-list ("default-iterates" 
block), could be used for set-face/get-face of panels.

yes. and i guess your list-filling does something similar, to save/restore 
facets?

I thought about this pretty hard for shell-list.

shell-list allows functions to be placed in the map dialect block, 
which is kind of like your /explicit refinement above. The automatic 
iterates are found in the default-iterates block.

I get a seperate list with the hits and can jump from there. emacs.

I'm trying to solve a performance problem with lists that uses a 
supply function. Mouseovers are dog-slow because every time the mouse 
gets over a row, the entire list view is regenerated by the supply 
function. the supply function redraws the list and truncates values 
so they fit in cells, paints backgrounds according to cell contents, 
etc. Lots of stuff and slow.

I'm looking for a way to discern between that I want to show the 
entire list and when I'm just doing a mouse over, so the supply function 
only changes the colors of the affected rows. Ideas?

Henrik, can you show us the code (or a cut down demo) of your list 
?

I threw the code out. It's been replaced by the standard LIST function 
in layout

it was way too memory hungry and the code was about 800 lines. it 
also took about 25 lines to initialize one list.

Gosh. Was the old code an attempt at direct pane programming ? Is 
the standard LIST performing ok ?

The standard list has a few problems, notably in that you are forced 
to update the entire listview, when you want to refresh a single 
entry. That makes it slow for mouseovers.


The old code was much faster at scrolling and doing mouse overs as 
I could have full control over which part of the list I wanted to 
show, but everything was done manually, but it was not direct pane 
programming.

but the standard list is very elegantly done and allows you to have 
a simple iteration function. It's a bit too simple, but then again 
I can't find proper docs on doing LIST with SUPPLY

5 seconds might not be enough time for the first time user to allow 
the user to respond to a software firewall (to add rebol to the allow 
list) and still get a connection.

do http://www.lexicon.net/antonr/rebol/gui/iterated/demo-vid-list-scroller.r

hope to release my new list-view for VID some time early next year... 
basing it on the standard LIST instead now

whoo, so many lists now :-) btw I wonder why Carl did not include 
his newer version of list, which was imo bacwards compatible, improved, 
and was done at days of IOS 1.3 View initiative ...

list: Petr, any reference to this improved version? I think I have 
it somewhere but can't remember the name of the script.

question: I'm using a text-list, which I update with some data after 
the use selected an entry. I want the slider to go to the top and 
show the first entry in the list as well. Positioning only the slider 
leaves the entries somewhere in teh middle.

How to add scroll-wheel support to a text-list?

Sorry, Robert. There is something I want to announce. :-)


http://hmkdesign.dk/list-test.png<--- a picture of the list view 
I'm building.


Currently about half done and quite usable at this time: It's resizable. 
Values are stored as blocks of blocks. All columns can be sorted. 
Input columns can be filtered so you can show only some columns. 
Columns can be freely reordered (but not in the GUI yet). One arbitrary 
column can be resized.

It has the normal range of series manipulation functions available 
in REBOL. There is also possibility for inline editing, by doubleclicking 
a line. Changed values are automatically stored in the list. All 
such operations are "bundled" in the list view VID code and you only 
need to provide whatever functions needed to store the list data 
in an external place. If a text entry is too wide, it'll be neatly 
cut with ellipsis (...).

Filtering function, to filter input by rows. Also has a scroll-to-selected-line 
function.

It's about as fast as the current LIST in VID, since it really is 
LIST with just a whole bunch of extra functions to make general list 
views easy. There are functions possible for clicking and double 
clicking and functions for retrieving rows and columns.


Current limitations: No mouse over indication (can't make it fast 
enough). Only one resizable column. No keyboard navigation. No horizontal 
scrolling. No scroll-wheel support. It doesn't integrate 100% with 
VID yet. I'm using some of my own widgets and bitmap graphics from 
a pretty big GUI library. Stripe look, font and coloring is locked. 
No standard settings yet for the list view.

All code is about 250 lines.


Planning: Reordering columns via drag'n'drop. Column resizing, if 
I can figure it out. Format the font object conditionally from list 
input (make this line bold if the age column is > 45 years, etc.). 
Grid drawing. Images in list rows. And if I can get around to it: 
Single cell in-line editing ala spreadsheets. :-)

Yes, you can freely do that in runtime, although there is no method 
for doing that within the list. All viewing and filtering operations 
are non-destructive.

columns have input and output description blocks. so if you have 
a dataset with [name height weight age sex] you can select to output 
[name age] and also rearrange that to [age name] and simply refresh 
the list. also adding and removing columns can be done on the fly.

I prefer having grid with nice and open/extensible functionality, 
than with knowing it is based upon 'list or other style - make it 
from scratch, if necessary :-)

The caching stuff you mentioned is something I partially did for 
my old listview. Correctly the listview was very fast, but also very 
memory hungry, took a long time to initialize and the code was 5 
times bigger and a lot harder to maintain.


This one simply doesn't paint anything other than what is seen, so 
if you have a million rows, only 15 will be painted, if you have 
15 visible rows on screen. The backside to the LIST which this one 
is based on is that it uses iterated faces. This means that if I 
only do a small change to one face, it has to repaint the entire 
list view with all rows. That's why mouse over became too slow to 
be practically useful. I'm not sure yet how to fix this problem.

Henrik, I advise building this style from FACE (ie. not an existing 
style, like LIST). Makes it less dependant on other code, more portable 
(ie. to RebGUI), and you will be forced to write your own pane function, 
which you will need to do anyway . :)

I based a lot of styles on existing styles, and this is a great way 
to learn, but I found as existing styles were changed (bugs fixed 
or ottherwise), my styles would break. Then it's a run-around to 
do a version-dependant fix for the problems.

I've got an area field and am using ^S to send text in my chat client. 

I clear the field with face/line-list: none, and face/text: copy 
""

but the cursor ends up not at the left edge.  How does one fix that?

anton, I read it (great stuff), but I currently have problems with 
which method to correctly group faces in the main pane (header, scroller 
and list), and get each face properly initalized. I know LAYOUT does 
this, but I can't use this with already made faces.

ok, haven't gotten very far yet, but http://hmkdesign.dk/rebol/list-view3.r

my attempt at a list view based on custom faces. more will come during 
the day... basically only fills out values and the sort buttons react 
visibly, but no sorting happens yet.

list is now resizable

new version 0.0.5 uploaded.

single selecting works now.
standard font made a bit smaller.

selecting requires the use of a new face called LIST-TEXT with a 
FEEL that allows for mouse over and single/double clicking in the 
list.
single-click action is stored as a block in LST-ACT
double-click action is stored as a block in DBL-LST-ACT

Demo can be found at http://hmkdesign.dk/rebol/list-view3.r

0.0.7 uploaded...

This one is released under the BSD license.

Changes:
      New: Added /force refinement to UPDATE

      Fix: Fields are rendered now, when filtering with fewer results than 
      the size of the list view.
      New: List now can have an edge of any size.
      New: ALT-LST-ACT action for right click actions.
      New: If HDR-COLS is empty, the header is now not drawn.

viz. moving up/down the list.

Robert: I will look at right click unsorting.


Realtime filtering is really up to the field that provides the string 
and updates the list, not the list itself (more flexible that way). 
I've done it before, where realtime performance could be achieved 
with more than 30000 entries in the list, so it's not a problem to 
do.

sorry for the wait... Version 0.0.8 uploaded

Changes:

      New: Functions to select next, previous, next page, previous page, 
      first and last. Uses FOLLOW.

      New: FOLLOW. Automatically follows the selected entry so it's always 
      in view. Requires FOLLOW? set to TRUE

      New: SCROLL-HERE. Scrolls automatically to the selected entry.
      Fix: Filter result was lost after resize.
      Fix: Sorting was lost after resize.
      Fix: Sorting indication is now persistent after resize.
      Fix: Dual state sorting was accidentally broken.
      New: Initial sorting can now be set before first view
      New: Right-click on header unsorts the list

Demo can be found at http://hmkdesign.dk/rebol/list-view3.r

Are the header names hard coded into the list-view ?

how does your list cope with an empty list ?

I don't know yet, but I plan to just let it draw empty cells. It 
should do the same if you filter for something not in the list

what's the syntax for creating a two column empty list ?

you'd use (for now):

view layout [list-view with [in-cols: [a b]]]

The list uses some standard values that are not correct right now, 
but I'll move those values into the correct place soon. If you for 
example set HDR-COLS to [], the header won't be shown.


The idea is that you have IN-COLS, which is based on the data you 
have in DATA. It's just a block of any words to tell each column 
apart. Then you will have OUT-COLS, which are the columns shown in 
the list view. The ordering of OUT-COLS determines the order that 
the columns are displayed in. HDR-COLS is for now a separate block, 
but it will be combined with OUT-COLS later. This ordering mechanism 
doesn't work yet, but it'll come in very soon.

Example of intended functionality:


in-cols: [person-id age sex height weight name address city country 
phone]
out-cols: [name age city]

Then you could reorder:

out-cols: [age city name]


And after using UPDATE, the list would update on the fly. DATA is 
not manipulated at all.

I tried this

list-view with [in-cols: [user time] data: []]

which doesn't work.

hmm.. that's because the layout is still hardcoded. you'll need to 
define a new layout in LIST-VIEW/LST/SUBFACE for now

the layout is for one row and it must be horizontal. you should also 
use the LIST-TEXT widget.

example:

layout/tight [across space 0 list-text 100 list-text 50]


but.. when I get this done right, it shouldn't be necessary to define 
the layout

uploaded 0.0.11

Changes:

      New: Header can now be defined from the words in OUT-COLS or IN-COLS

      New: No longer need to manually create the subface for the list
      Fix: DATA was not properly initialized

      Fix: List was updated twice during all manipulation functions added 
      in 0.0.10

      New: Added /no-show refinement to FLT-LIST to speed up certain updates 
      that require setting the selected row before showing

repeat i length? out-cols [
          insert tail lo [list-text 100]
        ]

and it automatically expands it to the total width of the list

and the width of the list is determined by what ?

you could: view layout [list-view 600x200]

We don't have a resizable bsd licensed multicolumn list ... so carry 
on :)

so the background behind the list rows is simply 240.240.240

also by manual layout... I think it's actually easier to create the 
manual layout if you want to set a lot of parameters such as:

list-text 75 right bold font-size 17

which is standard LAYOUT dialect

then you can set all parameters of the list-text face

this is also how LIST in VID works

Anytime I press "B", I want SOME_LIST to be modified.

OK. I'm still not sure, but if you want to use SOME_LIST not at the 
head, you can simply move its index elsewhere, using AT.

but if I press the button again, if will act on the original list, 
not the updated list.

yes, it's the same list, because SWAP does not modify the index, 
only the content. you will probably need to use AT in a separate 
operation afterwards to move the index to the desired position.

do you not want the DO block in the layout to be performed after 
FLASHCARD-PARSING? otherwise the questions-list.txt file is not generated 
in time.

since I was new to rebol/view and vid, I didn't know of a better 
way to make two streams of data, one going to questions and another 
to asnwers. The prospect of updating two sides of the same list seemed 
confusing to me hence my split.

there is a typo in line 54: quesiton-list.txt

I see you moving it within the list but the file doesn't dynamically 
get changed using MOVE...

MOVE modifies the list like SWAP does, but does not require any adjustments 
to index.

Is there a standard list of  what goes into the block following the 
REBOL flag? I'm thinking javadocs here. Thanks.

@Henrik, looking at your example code again, I'm not quite sure what 
the refinement qa/1/2 does. I thought it meant that "get list item 
2 at 1 place away" but I'm not sureanymore..

@Henrik, I figured out the problem. I make call next on the input 
at  the end of  the button press. This causes the list to change.

@Henrik, I do need to do a next in the button so that whenever the 
button is pressed, the next question comes up. Is there a way of 
 "holding" the list so it doesn't move around while I try to modify 
it?

Refinements are options, sometimes used in twos or threes, and the 
disadvantage here is that the argument list then can become hard 
to read. It's a good skill to create functions without too many refinements. 
I personally consider refinements to be one of the less stellar parts 
of REBOL.