[open-cobol-list] Re: Team setup (Bernard Giroud)

[Roger While]

This is what I am currently doing :
Minor cleanups ->
  message texts
  replace use of back-tick in messages with quote
Fix OPEN EXTEND problem reported in list

I should be able to commit the above this weekend.

There is the problem with default file error handling.
If there is no FILE STATUS, the program must exit in case of
error. This involves 2 things - activating error handling for the OPEN
and checking in default error handling if we should exit or not
(FILE STATUS defined).
I was hoping to use the file status pointer to check, however, if
FILE STATUS is not defined the pointer gets dynamically allocated.
Looks like another status bit in the file structure.
I'll see if I can get this in also this weekend.

Although not urgent, for better readability and manipulation, I would like
to pass all '.h' and '.c" files through "indent". (And edit out the form 
feeds)
Following options seem to get the best results :
indent -brs -ce -cdw -l96 -br -kr -ut -ts8 -i8 -psl -pcs -bs -sob -ss
I've set the line length to 96 as some constructs end up with
multiple non-readable lines. Aim here is obviously to get it to 80.


Pierre, you can send me your fixes. I will check them and integrate.


> Now we will get that quantity into a long but with a mpz_get_si which
> is designed to return an int. I think the solution is to use mpz_get_ui
> into a signed long in the first case. BTW, it seems that 
mpz_fits_slong_p
> return false in that case, which I don't understand :-> ((.
> The modified routine is at the end of this mail. It seems to be OK,
> and no regression.
> Could you tell me if it seems allright to commit ?

Actually, it is more complicated than that. It's not only this routine.
It would appear that there are some places in the code that
assume max 9 digits in 4-byte binary (max is 10) and 18 digits in
8-byte binary (max is 20).
Also, not to forget the "std=mf" with binary byte allocation from 1 to 8.
Routines display_(add/sub)_int are also missing test on !binary_trunc.
Still working on this.


Re. ISAM
Maybe we should petition IBM to opensource their Informix C-ISAM.
After all, that's what MF uses (albeit a derivative)
There is another free ISAM implementation :
http://mission.base.com/peter/source/

Roger