Re:Multicolumn Printing (Text Flow)
I like Cubist's binary search idea, but I recall that when I originally stepped
through the text by word rather than by char I was stymied by the fact that
a quoted series of words is understood as a single word, even though a proper
page break could fall within that series. I assume a binary search by char
would work. Thanks.
David Epstein
sez [EMAIL PROTECTED]:
Here's another approach to the text overflow calculation:
Nice idea, but do you really need to test every character in sequence? I
was under the impression that Rev doesn't believe in hyphenation, so you
could
speed up this handler by testing words rather than characters; a further
speedup could be gained by starting with the entire field (what if the text
doesn't
actually overflow?) and doing a binary search. With that in mind, here's my
modified version of your handler ('ware the line wraps!)...
function ExtraText2 FldName
-- removes from fld FldName, and returns, the htmlText that overflows size of
fld FldName
put (the number of words in fld FldName) into WordKount
put max (WordKount div 2, 1) into DerDelta
put the height of fld FldName into DerHeight
put the formattedHeight of word 1 to WordKount of fld FldName into
CurrentHeight
if CurrentHeight = DerHeight then return -- no overflow, right?
-- now comes the binary search
put WordKount into HiEnd
put CurrentHeight into MaxH
put DerDelta into LoEnd
put the formattedHeight of word 1 to LoEnd of fld Fldname into MinH
repeat
put max (DerDelta div 2, 1) into DerDelta
if MinH DerHeight then -- even the minimum is too long
put LoEnd into HiEnd
put MinH into MaxH
subtract DerDelta from LoEnd
put the formattedHeight of word 1 to LoEnd of fld Fldname into MinH
else if MinH = DerHeight then -- *just* right!
return the htmlText of word (LoEnd + 1) to -1 of fld FldName
else -- the minimum is too small
if MaxH - MinH = 1 then -- we've discovered the overflow point
return the htmlText of word (LoEnd + 1) to -1 of fld FldName
end if
add DerDelta to LoEnd
end if
put the formattedHeight of word 1 to LoEnd of fld Fldname into MinH
end repeat
end extraText2
___
use-revolution mailing list
[EMAIL PROTECTED]
http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Multicolumn Printing (Text Flow)
Here's another approach to the text overflow calculation: function extraText f -- removes from fld f, and returns, the htmlText that overflows size of fld f repeat with c = 1 to length(fld f) if the formattedHeight of char 1 to c of fld f the height of fld f then put the htmlText of char c to -1 of fld f into myHtml delete char c to -1 of fld f return myHtml end if end repeat return empty end extraText David Epstein 4. Multicolumn Printing (Text Flow)? (Frank D. Engel, Jr.) My best guess is to create separate fields for the left and right columns and somehow figure out where to split the text so that I can fit the text neatly into the two columns. Does anyone know of an easy way to figure out where to split this? ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: selectedText from a palette?
Rich Layne wrote: I trying to make an application that has most of its functionality on a palette-- a bunch of different buttons that will do different stuff to the main data stack. Most things work, but handlers that involve the selectedText dont seem to make it to the main stack, (which is the topStack.) My experience using Metacard 2.5 is that when you click on a palette the selection in the topStack is lost, so the selectedText is empty. I use these handlers in my palette's script as a way of recording and restoring the selection: on mouseEnter global chunkStore get the selectedChunk if it is empty then put empty into chunkStore else put it of word -2 to -1 of the long id of the selectedField into chunkStore end mouseEnter on mouseDown global chunkStore if chunkStore is not empty then do select chunkStore end mouseDown So on mouseUp, the selection is where you want it and references to it should work properly. There is a slight flashing associated with the selection being lost and restored, so I'd be interested to hear other methods. David Epstein ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Windows Printing
Jacqueline Landman Gay wrote of Windows printing problems: The printing stack has its formatForPrinting set to true. My script fills both fields, prints the card, prints a break, then repeats until the data is gone. The documentation for formatForPrinting warns us never to edit a field with its formatForPrinting set to true. A script that fills both fields is editing a field with formatForPrinting set to true. I experienced the symptoms you described, plus the program frequently simply quitting, while trying to print in Windows using Metacard 2.5. I believe I have now remedied the problem by having my script repeatedly call this command: on ffp whether set the formatForPrinting of stack myPrintLayoutStack to whether end ffp FormatForPrinting needs to be FALSE before the script changes a field's content (including filling it in the first place), and it needs to be TRUE before you try to determine what changes are needed based on how the text fits in the space available (and before you print). For example, here's handler I use to determine what text doesn't fit in a field and needs to be moved to the next column or page: function extraText f -- returns the htmlText that overflows size of fld f ffp true -- SO THAT FORMATTEDHEIGHT WILL BE PROPERLY CALCULATED repeat with c = 1 to length(fld f) if the formattedHeight of char 1 to c of fld f the height of fld f then put the htmlText of char c to -1 of fld f into myHtml ffp false -- BECAUSE I'M ABOUT THE CHANGE THE CONTENT OF FLD F delete char c to -1 of fld f return myHtml end if end repeat ffp false -- READY FOR FURTHER EDITS OF FLD F return empty end extraText Just before you print, set ffp true again. Somewhere (but not in Rev docs, I think) I read advice suggesting a further precaution: close and open the stack and then set formatForPrinting to true. So I added this just before printing: set the destroyStack of stack myPrintLayoutStack to true save stack myPrintLayoutStack close stack myPrintLayoutStack open stack myPrintLayoutStack ffp true This last bit seems a littler superstitious, but since I made these fixes a couple of weeks ago my Windows printing has worked properly. I hope this will help with your difficulties. David Epstein ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: OptionKey madness
Richard Gaskin wrote: My goal was to provide 20 simple keystrokes for user-settable actions, giving them Cmd/Ctrl-1 through 0, and Opt/Alt-1 through 0. Function keys are out since both Win and Mac OSes reserve so many for specific features. Now it seems Option keys may be out as well Have you considered using the Mac CONTROL key to do what your Windows ALT key does? This works: on controlKeyDown what -- for Mac CONTROL key keyShortcut what end controlKeyDown what on optionKeyDown what -- for Windows ALT key keyShortcut what end optionKeyDown (Where a keyShortCut handler does the work intended for Win-Alt-what or Mac Ctrl-What). A commandKeyDown handler catches the use of the Windows control key and the Mac command key. David Epstein ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Selection strategies
Richard Gaskin noted that when text is selected in the field [of the topstack] the palette buttons work great, but if the cursor is inserted with no selection then clicking in the palette window causes the field to lose focus. My palette is a menu of pulldown buttons, and these handlers are in the palette's stack script: on mouseEnter global chunkStore get the selectedChunk if it is empty then put empty into chunkStore else put it of word -2 to -1 of the long id \ of the selectedField into chunkStore end mouseEnter on mouseDown global chunkStore if chunkStore is not empty then do select chunkStore end mouseDown This works properly, although there is a slight flash as the text gets reselected on mousedown, so an approach that avoided that would be welcome. One approach: if I test the selectedText's length and load chunkStore only if that length is 1, the engine keeps the selection without my help for cases where there's more than an insertion point, and the flash is avoided. But in that case my experience is that the insertion point is not properly placed after the command is executed (e.g., a paste command does not place the insertion point after the newly pasted material), whereas the scripts above set the insertion point properly. But there is that flash. I'm using MC 2.5. David Epstein ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
re: Long combo box buttons
To display more than a few lines in a combo box list: set the menuLines of btn myCombo to the number of lines in btn myCombo Works the same way as with option button lists. Tested with Metacard 2.5 Mac OSX. David Epstein ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Poor technique?
Robert Brenstein quoted Scott Raney: open inv is poor technique in MC anyway (it's only there as a SuperCard compatibility feature), as is setting up fields in one stack from another directly. Instead, do the init in a preOpenStack handler, and get the data in the dialog from a well known place (use the dialogData property if you don't already have such a place). Regards, Scott Why is setting up fields in one stack from another directly a worse technique than writing data to a global property and having the preOpenStack handler transfer the data to that stack? I can see that the latter might be convenient in some cases, but is there any important reason not to do the former? Similarly: is it in any way better to set custom properties of the second stack before opening it than to set the text of its fields before opening it? The ability to read and write to fields of unopened stacks is very powerful; Raney's remarks suggests there are some limits on how this should be used. Can anyone clarify this? David Epstein ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Moving more than one line in a listfield
This is a subset of a more elaborate function, so I hope it still works. It can be used to move the selected lines of a list-type field to the top or bottom of the list, or up or down. David Epstein function reListed theList,actPieces,action,myDelim,@newSelec -- theList is the starting list; if return-delimited, pass return as myDelim -- actPieces gets passed the hilitedLines of the field containing theList -- action is 1, 2, 3, or 4, depending on where you want to move the selected lines. -- the function returns the modified list, and loads variable newSelec -- with a comma-delimited list of numbers useful for setting the appropriate -- hilitedLines of the field after the list has been altered. put theList into origList replace , with in actPieces set the itemDelimiter to myDelim put the number of items in theList into z put the number of words in actPieces into p repeat with w = p down to 1 put word w of actPieces into i put myDelim item i of theList before newList delete item i of theList end repeat -- items (lines) specified in actPieces are now listed in newList, and -- have been removed from theList. The rest of the script -- combines those newList and theList in the right order. delete char 1 of newList switch action case 1 -- move newList to the top put rangeList(1,p) into newSelec return newList myDelim theList case 4 -- newList to the bottom put rangeList(z-p+1,z) into newSelec return theList myDelim newList case 2 -- newList up put word 1 of actPieces into slot if slot 3 then put rangeList(1,p) into newSelec return newList myDelim theList else put slot - 1 into slot put rangeList(slot,slot+p-1) into newSelec return item 1 to slot-1 of theList myDelim newList myDelim item slot to -1 of theList end if case 3 -- newList down put word -1 of actPieces into slot if slot = z then -- last selected is at end of list put rangeList(z-p+1,z) into newSelec return theList myDelim newList else put slot + 2 - p into slot put rangeList(slot,slot+p-1) into newSelec return item 1 to slot-1 of theList myDelim newList myDelim item slot to -1 of theList end if end switch end reListed function rangeList a,z -- returns comma delimited series of numbers from a to z put z-a into myCount if myCount = 0 then return a put a into hold repeat with b = 1 to myCount put , a+b after hold end repeat return hold end rangeList ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Moving more than one line in a listfield (fwd)
A clarification: The function I just suggested for this purpose assumes that you want to group the several selected lines, e.g., that when you move lines 4 and 7 up in the list you want them to become lines 3 and 4, not lines 3 and 6. I hope this is what was intended. David Epstein ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Getting a chunk of stylized text?
There are probably faster ways to do this, but try this: function extraText f -- returns the htmlText that overflows size of fld f -- and deletes it from field f repeat with c = 1 to length(fld f) if the formattedHeight of char 1 to c of fld f the height of fld f then put the htmlText of char c to -1 of fld f into myHtml delete char c to -1 of fld f return myHtml end if end repeat return empty end extraText David Epstein Getting a chunk of stylized text? From: Dan Friedman [EMAIL PROTECTED] ... Let's say field A contains more text than can fit in it. The text in field A contains multiple fonts, styles and sizes. How can I get the stylized text that is not being displayed? (and remove it from field A and put it into field B) ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: ranging lists
Since you're dealing with integers, you could run each line of list 1 through a function like function numberSeries x,y -- returns space-delimited list of all integers from x to y to build an expanded version of list 1 that explicitly lists all possible matching numbers in each range. Then send each item n in list 2 to a function: function whichRange n,expandedList set wholeMatches to true get wordOffset(n,expandedList) if it = 0 then return 0 return the number of lines in word 1 to it of expandedList I suspect this will be faster than running each of your 2000 numbers through a series of comparisons. David Epstein I have 2 lists. I need to check whether the values of list 2 fall within the range of the values of list 1. eg. list 1 1 10 20 25 45 55 list 2 1,3,12,23,34,45,52,78 List 1 may have up to 60 different ranges, one range per line. List 2 may have over 2000 items. (these are line numbers of a text database) ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
