Hello together!
I have a 2 questions:
1. In the GPC ToDo-list is a point for an AT&T <-> Intel assembler converter. Eike Lange and I discusses to implement this feature in pindent (see anouncement earlier this month). But we are both not very good assembler programmers. The questions: is there is anybody (with very good experience in both assembler languages) who want help us to implement this feature in pindent so please feel free to contribute to this little project. We only talk about inline-assembler and pindent doesn't eat .s-files. Contact us in the GPC- or pindent- (pindent@mond.dida.physik.uni-essen.de) mailing-list. Thanx
2. Today I've look in the GPC ToDo-list and I see the indentation program request. If there is any reason to have this point in the list now? Because pindent is in its first version availible (http://mond.dida.physik.uni-essen.de/~pindent/) and Eike and I working on the next version.
Thorsten
Hello!
Thorsten Hindermann wrote:
- In the GPC ToDo-list is a point for an AT&T <-> Intel assembler
converter. Eike Lange and I discusses to implement this feature in pindent (see anouncement earlier this month). But we are both not very good assembler programmers. The questions: is there is anybody (with very good experience in both assembler languages) who want help us to implement this feature in pindent so please feel free to contribute to this little project. We only talk about inline-assembler and pindent doesn't eat .s-files. Contact us in the GPC- or pindent- (pindent@mond.dida.physik.uni-essen.de) mailing-list. Thanx
I don't quite understand what the pindent program is intended to do: Read Pascal source with Intel inline assembler and output reformatted Pascal source with AT&T inline assembler? IMO both tasks are quite different (while they require similar techniques) and should be in two separate programs.
- Today I've look in the GPC ToDo-list and I see the indentation
program request. If there is any reason to have this point in the list now? Because pindent is in its first version availible (http://mond.dida.physik.uni-essen.de/~pindent/) and Eike and I working on the next version.
Indeed. So we should move that item from the "to do" section to the "done" section. :-) Or shall we reformulate the item to ask for contributions to the pindent project?
Peter
Hi Peter!
Peter Gerwinski wrote:
I don't quite understand what the pindent program is intended to do: Read Pascal source with Intel inline assembler and output reformatted Pascal source with AT&T inline assembler? IMO both tasks are quite different (while they require similar techniques) and should be in two separate programs.
No, if pindent detect the keyword asm the assembler statments between "(" and ")" are pushed to a stack like the pascal source. The next step is to interpret/convert the statments in one directions in case of the command-line switches (AT&T -> Intel or Intel -> AT&T). One switch for AT&T-syntax to Intel the other for Intel to AT&T. When all work is done (pascal and inline-assembler) the stack(s) are written into the output file. But in the moment Eike and I discus this feature. We both not good assembler programmers so we look for anybody who will help us or give an advice for this feature to implement it in pindent or not. On the other side if pindent read the pascal source why not read the inline-assembler and make the interpret/convert operation? What do you and the GPC-community think about this feature in pindent? Let us know your opinions to this feature.
Indeed. So we should move that item from the "to do" section to the "done" section. :-) Or shall we reformulate the item to ask for contributions to the pindent project?
No, the first step is to move that item from the "to do" to the "done" section :-)!
Thorsten
Thorsten Hindermann wrote:
Peter Gerwinski wrote:
I don't quite understand what the pindent program is intended to do: Read Pascal source with Intel inline assembler and output reformatted Pascal source with AT&T inline assembler? IMO both tasks are quite different (while they require similar techniques) and should be in two separate programs.
No, if pindent detect the keyword asm the assembler statments between "(" and ")" are pushed to a stack like the pascal source. The next step is to interpret/convert the statments in one directions in case of the command-line switches (AT&T -> Intel or Intel -> AT&T). One switch for AT&T-syntax to Intel the other for Intel to AT&T. When all work is done (pascal and inline-assembler) the stack(s) are written into the output file. But in the moment Eike and I discus this feature. We both not good assembler programmers so we look for anybody who will help us or give an advice for this feature to implement it in pindent or not. On the other side if pindent read the pascal source why not read the inline-assembler and make the interpret/convert operation? What do you and the GPC-community think about this feature in pindent? Let us know your opinions to this feature.
I agree with Peter. These are really two different tasks that should be done by different programs (keyword modularity). Of course, the programs might share some code (units etc.)...
Frank
-
Frank Heckenbach -
Peter Gerwinski -
Thorsten Hindermann