Re: strange behavior of EU to C translator
Jerry Story writes:
> I tried the EU to C translator on my Diet Monger @$$ Kicker
> program, for LCC-WIN. The translator produced almost
> 500 C files. Most of them were 320K. main_328.c was
> 13 megs. init_.c was 15 megs. Way too much C code.
> The total Euphoria code, mostly data, is about 2 megs.
> What's happening?
I tried it too, and with the beta-1 Translator
I got similar results (460 .c files).
The beta-2 (under development) Translator was
somewhat better. init_.c was reduced from 16Mb
to 6.8Mb. main_.h was cut from 4.7Mb to 1.9Mb.
Some of the other .c files were shrunk a bit.
Still pretty bad.
The reason is that your program source code
contains something like 350,000 floating-point
numbers, plus 15,000 or so strings, plus thousands
of complex sequences.
The way I've set things up, each of these items
needs a few lines of C code to declare it and
initialize it. With the interpreter, you can just
list all the numbers 10 or so per line, very compactly.
I think Euman ran into similar problems.
I'll see what I can do to improve the situation,
but you could solve the problem yourself by
reading the data from a data file (or EDS database)
rather than initializing the data with Euphoria
declarations inside your program source.
Regards,
Rob Craig
Rapid Deployment Software
http://www.RapidEuphoria.com
|
Not Categorized, Please Help
|
|
