Right$() or Str$() bug?

[Rod]

Can anyone see why Liberty gets hung up on this?


for n=1 to 10
n$=right$("000"+str$(int(rnd(0)*200)),3)
next


[tsh73]

Interesting.
It depends on randomize, so with randomize 0.1 it dies on 12 th number
I tried to print these numbers
If n= 12 then print rnd(0)*200
but I see nothing interesting

But moving rnd out of formula ( assigning to a variable) helps.

Also
Number in error message is string length under right$.
And removing right$ removes error.

Also
Saving string into a var and then making right$ removes error.

Also
removing str$ and int does not help
n$=right$("000";rnd(0)*200,3)
Still dies.

Left$ and mid$ seems work OK.

[Rod]

This is my error log, My guess is that either the loop index or string index is being compromised. One for Carl. I am pretty sure I have used such code in the past.


Error log timestamp Saturday 26/11/22 12:55:11

Runtime error: index: 6 is outside of collection bounds

Error(Exception)>>defaultAction
Error(Exception)>>activateHandler: <anUndefinedObject>
Error(Exception)>>handle
Error(Exception)>>signal
Error class(Exception class)>>signal: <'index: 6 is outside ...'>
BasicProgram(Object)>>error: <'index: 6 is outside ...'>
BasicProgram>>terminateRun: <anError>
[] in BasicProgram>>errorHandlerBlock
ExceptionHandler>>evaluateResponseBlock: <aBlockClosure> for: <anError>
[] in ExceptionHandler>>handle:
ProtectedFrameMarker(BlockClosure)>>setUnwind: <aBlockClosure>
BlockClosure>>invisibleEnsure: <aBlockClosure>
ExceptionHandler>>handle: <anError>
ExceptionHandler>>findHandler: <anError>
Error(Exception)>>activateHandler: <anExceptionHandler>
Error(Exception)>>handle
Error(Exception)>>signal
Error class(Exception class)>>signal: <'index: 6 is outside ...'>
String(Object)>>error: <'index: 6 is outside ...'>
String(IndexedCollection)>>errorInBounds: <6>
String(IndexedCollection)>>checkIndex: <6>
String>>at: <6>
String>>replaceFrom: <1> to: <3> with: <'00089'> startingAt: <4>
String(IndexedCollection)>>copyFrom: <4> to: <6>
[] in BasicRight class>>using:program:
BasicVariableAssigner>>value
[] in BasicProgram>>begin
ExceptionHandler>>evaluateProtectedBlock: <aBlockClosure>
[] in ExceptionHandler>>activateDuring:
ProtectedFrameMarker(BlockClosure)>>setUnwind: <aBlockClosure>
BlockClosure>>invisibleEnsure: <aBlockClosure>
ExceptionHandler>>activateDuring: <aBlockClosure>
ExceptionHandler class>>handle: <anError class> with: <aBlockClosure> during: <aBlockClosure>
BlockClosure>>on: <anError class> do: <aBlockClosure>
BasicProgram>>begin
BasicProgram>>run
[] in BasicSourcePane>>run
BasicCompiler class>>compileFromString: <'
for n=1 to 10
n$=...'> notify: <aBasicSourcePane> ifSuccess: <aBlockClosure>
BasicSourcePane>>run
BasicSourceWindow>>run
BasicEditorModel>>run: <aToolbarButton>
BasicEditorModel(Object)>>perform: <#run:> with: <aToolbarButton>
ToolbarButton>>triggerClickedEvent
Message>>perform
NotificationManager>>runPendingEvents
NotificationManager>>runEventLoop
Message>>perform
Message>>evaluate
Process>>safelyEvaluate: <aMessage>
Process>>evaluate: <aMessage>


[tsh73]

Just for a record,
it shows same problems in LB 4.04 and JB 2, JB 1.01

[pierre]

for n=1 to 10
n$=right$("000"+str$(int(rnd(0)*200)),3)
print n$
next

runs without any problem in RunBASIC and in LB 5 alpha

[xxgeek]

hmm, very interesting.

As you'll see in the code many different experiments have been conducted.

NOTE: This code creates a file named testRandom.txt in DefaultDir$

A couple of possible "theories" to question.
Just thoughts and more questions. The type a new LB programmer might ask.

1. Does n$ (or should I say int(rnd(0)*200) become too long and gets converted to an e num eg .00000000.......123 e^
and right$ or int() doesn't know what to do with it?
I ask because during experiments I did see some of those numbers.

2. Does right$ use up too much memory trying to read a number longer than a set limit on strings (set by LB)?
I ask because according to tsh73 and confirmed here.

Number in error message is string length under right$.
And removing right$ removes error.

Also
Saving string into a var and then making right$ removes error.

---------

The lines in testRamdom.txt with numberHmmmm happen only when there is NO error.

on error goes straight to [errorr] without completing n$ -that's expected.

If n$ could be captured and investigated some more light could be shed.
Anyone able to capture n$ when an error occurs?


on error goto [errorr]
[start]
open "testRandom.txt" for append as #1
[newNloop]
   #1 "Starting a new n loop"
while errorr <1200
   for n=1 to 100
      x = x + 1
        'n$=right$(rnd(0)*200,3)
        'n$=right$("000";( (rnd(0) )*200),3)
        'n$=right$(str$(int(rnd(0)*200)),3)
        'n$=right$("000"; str$((   (int(   (rnd(0)*200  ) ) )  ))  , 3)
        'n$=right$( "000"+str$(  int( (rnd(0)   ) *200)) ,3)
        #1 right$("000"+str$(int(rnd(0)*200)), 3)+ "Hmmm"
        '#1 "n$ = right$+000+str$(int(rnd(0)*200)) = ";right$("000"+str$(int(rnd(0)*200)), 3)
        'print right$(n$, 3) + "right$(n$,3) = ";
        'print mid$(n$, 4, 3),mid$(n$, 4, 3) + "mid$(n$, 4, 3) = "
        'print left$(n$, (len(n$))-len(n$) ) + right$(n$, 3) + "left$(n$, (len(n$))-len(n$) ) + right$(n$, 3)"
                 'n$=right$("000"+str$(int(rnd(0)*200)),3)
        'print "n = ";n
        'print "n$ = ";n$
        'print "x = " ;x
        '#1 "n = ";n : #1 "n$ = ";n$
     scan
   next
   scan
 wend
   scan
   goto [newNloop]
[errorr]
    errorr = errorr + 1
    if errorr > 1000 then close #1 : end
    #1 "Errorr - n = ";n
    #1 "Errorr - n$ = ";n$
    '#1 "n$ = right$+000+str$(int(rnd(0)*200)) = ";right$("000"+str$(int(rnd(0)*200)), 3)
        '#1 "right$(n$,3) = ";right$(n$, 3);right$(n$, 3)
        '#1 "mid$(n$, 4, 3) = ";mid$(n$, 4, 3)
        'n$=right$("000"+str$(int(rnd(0)*200)),3)
                 #1 "x = " ;x
        #1 "n = ";n' : #1 "n$ = ";n$
#1 errorr;" errors out of ";x;" attempts"
close #1
goto [start]