Programming question... :)

[Paul Panks]

What is the most efficient way to check for multiple statuses of a particular variable? I thought of:

DIM N%(1,5)

or something like that

i.e.

IF N%(1,1)=-1 THEN PRINT" That isn't valid here.":GOTO Beginning_Loop
IF N%(1,2)=-1 THEN PRINT" There is a magic spell blocking this object's effect.":GOTO Beginning_Loop

The reason I ask all this is I want to cut down on the size of many of my programs, which have some hardcoded routines, i.e.:

IF LOC%(N)=105 then ?"That object is wielded."
IF LOC%(N)=205 then ?"That object is worn."

I seem to be repeating similar lines of code. I want to cut down on that by at least a third or 1/4.

Thanks in advance for any replies. Finally, how to I circumvent the 64KB limit for strings in PB for DOS? I was thinking of using XMS and EMS, but I'm not too knowledgable about those two areas of MS-DOS programming.
I'm trying to convert Westfront PC into something GW-BASIC or QBASIC could handle. I'm using PBDOS to code it in.

Regards,

Dunric
[email protected]
ICQ#12234336

------------------
Few cats act their age, while
most just cough up furballs.

[Matthias Kuhn]

Hi Paul,

a good way to check multiple statuses is:

select case N%

case 1:Rem kill cat
case 0,-1: rem end
case else: rem not good

end select

Your "IF then else" are a little complicated but very save.


Stings over 64k You can get with the latest version of PB:

Dim Virtual BigArray%(1:500000)
or
Dim Virtual bigstring(1:50000) as string*10

over the hole ems space; but they are
as slowly as a quick file-access.

One program to translate Quickbasic code in PB should be
in Your tools folder of PB.

Regards

Matthias Kuhn




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

[Lance Edmonds]

You can also use DIM HUGE, but arrays are still confined to conventional memory. DIM VIRTUAL will create arrays up to 32Mb each, limited by the amount of EMS your memory manager can provide.

DIM VIRTUAL requires PB/DOS 3.5. In PB/DOS 3.2, DIM HUGE limits subscript values to the integer range, but PB/DOS 3.5 does not have this limitation.

------------------
Lance
PowerBASIC Support
mailto:[email protected][email protected]</A>

[Paul Panks]

Hey Lance,

Is version 3.5 still available for sale?

Paul


------------------
Few cats act their age, while
most just cough up furballs.

[Tom Hanlin]

Absolutely.

------------------
Tom Hanlin
PowerBASIC Staff

[David J Walker]

Paul

A cat that wants its breakfast has no snooze button.

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

[Paul Panks]

Yes, but a cat that snoozes gets no breakfast. =)

Paul


------------------
Few cats act their age, while
most just cough up furballs.

[Michael Mattias]

IF N%(1,1)=-1 THEN PRINT" That isn't valid here.":GOTO Beginning_Loop
IF N%(1,2)=-1 THEN PRINT" There is a magic spell blocking this object's effect.":GOTO Beginning_Loop
'The reason I ask all this is I want to cut down on the size of many of my programs, which have some hardcoded routines, i.e.:
IF LOC%(N)=105 then ?"That object is wielded."
IF LOC%(N)=205 then ?"That object is worn."

I seem to be repeating similar lines of code.

This is a fairly common problem, especially when developing a new piece of software.

What I usually do is start out by creating arrays which match the code with the message; I also set up equates for the numeric values
Example:

%LOC_NewYork = 1
%LOC_Boston  = 2
%LOC_Philadelphia = 3
...

DIM CityName (100) AS STRING
CityName (%LOC_NewYork) = "New York"
CityName (%LOC_Boston) = "Boston"
CityName (%LOC_Philadelphia) = "Philadelphia"
...

  ' yadda, yadda, yadda
  ' code to get the city number from data...
    
    PRINT CityName (loc%)

You can do the same kind if thing above with your message. When the variable ou are testing is an array, you just use the array value as the subscript for the text message array.
[code]
PRINT TextMessage (N%(Ghost%, Goblin%))
or
PRINT TextMessage (N%(Widget%))
[code]

This is a lot more work to set up, but I find it more than pays for itself in ease of maintenance and readability.

MCM

[Lance Edmonds]

Yup, I like and often use an array that way too... dead easy.

Of course, if you are using PB/CC or PBDLL (Paul uses at least one of these) you can also use DATA and READ$ in the same manner, and since DATA is local to the sub/function, you can use separate sets of text/data across your app.

Using a string table in a resource file can also be useful, but there is greater overhead (string tables are stored in Unicode and converted to ANSI on the fly), however, if you want to support multiple languages, this method has distinct advantages.


------------------
Lance
PowerBASIC Support
mailto:[email protected][email protected]</A>

[Paul Panks]

Question for Lance:

Is the PB/CC IDE 16-bit or 32-bit? What's the main difference between 16 and 32 bit besides memory capacity per address space?
Or did I totally botch that description? =)

Paul


------------------
Few cats act their age, while
most just cough up furballs.

[Michael Mattias]

Unless my memory really IS going, the IDE is 32-bit.

BUT... the compiler itself (PBCC,EXE) is a 16-bit application (which produces 32-bit code).WHICH IS WHY WE CAN'T USE LFNs FOR #INCLUDE or #COMPILE

MCM

[Lance Edmonds]

Michael is correct - CCEDIT and PBEDIT are 32-bit, but the compilers themselves are 16-bit apps. However, being a 16-bit app does not mean the compiler *cannot* support LFN's in metastatements, etc. It just means that it will have to jump through a couple of small hoops to implement it.

Technically speaking, whether the compiler is 16-bit or 32-bit has almost no relevence to what can be achieved, be it compile-time memory limits or LFN's, etc.

Basically speaking (pun intended) R&D are constantly working at making significant improvements to our compilers... I think you may be impressed with what the future holds!

------------------
Lance
PowerBASIC Support
mailto:[email protected][email protected]</A>