PB5.50B1 : Little Question to @Constant$

[Bisonte]

Hello...

As i read in the announcement :

Code: Select all

- Added: @#StringConstant$ syntax support, to get the address of a string constant

I ask me one Question :

"What happens if someone do this ?"

Code: Select all

#StringConstant$ = "Hello"
PokeS(@#StringConstant$, "PureBasic")
Debug #StringConstant$

Legal ? Memory Errors (cause the poked string is longer/shorter..) ?

Is this a bug or a feature ?

[DontTalkToMe]

Variable constants are a reality finally. PB is pushing the boundaries

Code: Select all

#StringConstant$ = "Hello"
#Horror$ = "Horror"

Debug #StringConstant$
Debug #Horror$


PokeS(@#StringConstant$, "PureBasic")

Debug #StringConstant$
Debug #Horror$

The offset of #Horror$ is constant (just after the end of the original #Hello$ ) and so the result is what you would expect.

[Derren]

I need to get me the beta soon, this looks "interesting".

What about this? oO

Code: Select all

#const$ = "hello"
Dim myArray.i(10)
For i = 0 To 10
  myArray(i) = i * (10000) + 999
Next
PokeS(@#const$, "PureBasic Beta Stuff")

For i = 0 To 10
  Debug myArray(i) 
Next

[DontTalkToMe]

What about it ?

I need to get me the beta soon, this looks "interesting".

If you didn't have the beta, how can you test the above code ?

[skywalk]

Kinda confusing? #World$ does not always appear next to #Hello$ in the memory viewer?

Code: Select all

#Hello$ = "Hello"
#World$ = "World"
Debug "#Hello$ = " + #Hello$
Debug "#World$ = " + #World$
ShowMemoryViewer(@#Hello$-128, 256)
;Debug @#Hello$ ;<-- Without these, #World$ does not appear in Memory Viewer?
;Debug @#World$ ;<--
CallDebugger
Debug "PokeS(@#Hello$, '12345', 5, #PB_String_NoZero)"
Debug PokeS(@#Hello$, "12345", 5, #PB_String_NoZero)
ShowMemoryViewer(@#Hello$-128, 256)
Debug "#Hello$ = " + #Hello$
Debug "#World$ = " + #World$

[Derren]

What about it ?

I need to get me the beta soon, this looks "interesting".

If you didn't have the beta, how can you test the above code ?

I can't. That's why I asked^^

[DontTalkToMe]

Looked like you were saying "see what this program does!" oO
In that case, your program runs fine for what I can see.
If you were expecting for the array to be overwritten, that doesn't happen.
If that's what you were thinking because you didn't say.
oh well ....

[Fred]

Hello...

As i read in the announcement :

Code: Select all

- Added: @#StringConstant$ syntax support, to get the address of a string constant

I ask me one Question :

"What happens if someone do this ?"

Code: Select all

#StringConstant$ = "Hello"
PokeS(@#StringConstant$, "PureBasic")
Debug #StringConstant$

Legal ? Memory Errors (cause the poked string is longer/shorter..) ?

Is this a bug or a feature ?

It's the same as patching a literal string constant: you can do it, but it's strongly not recommanded.

Code: Select all

PokeS(@"Hello", "World")
Debug "Hello" ; Wooooot

To avoid this, we could put the literal string datasection in read only so it would raise a IMA, but I don't know if it worths it.

[sq4]

To avoid this, we could put the literal string datasection in read only so it would raise a IMA, but I don't know if it worths it.

Please don't. It would break backward compatibility.

[Bisonte]

To avoid this, we could put the literal string datasection in read only so it would raise a IMA, but I don't know if it worths it.

This is not necessary !

I only wanted to know, what happened... so we have NOT! do something like that...