[m-rev.] for review: add mdb `dice' command

[Ian MacLarty]

On Mon, Feb 07, 2005 at 03:25:23PM +1100, Julien Fischer wrote:
> > > > +			SpecialPredId = compare,
> > > > +			Name = "__Compare__"
> > > > +		;
> > > > +			SpecialPredId = initialise,
> > > > +			Name = "__Initialise__"
> > > > +		),
> > > The module compiler/special_pred contains predicates for handling the
> > > names of special preds.  I suggest moving the relevant bits into the
> > > mdbcomp library and using them here.
> > >
> >
> > Okay.  I've moved special_pred_description/2 from
> > compiler/special_pred.m to mdbcomp/prim_data.m, made it a function, and
> > used it in dice.m
> >
> 
> special_pred_description/2 is going to give you a different form
> of the name than that you originally had.  I was thinking of
> special_pred_name_arity/3.
> 

special_pred_name_arity/3 will also give a different name, but the
issue's not very important since compiler generated predicates aren't
included in the trace counts anyway.  I've used special_pred_name_arity
instead since the names it gives are shorter.

> > > The amount of nesting and lack of whitespace here make this block of
> > > code very difficult to understand.  Would it be possible to factor some
> > > of this out into separate functions?
> > >
> >
> > What do you mean by lack of whitespace?  Where would you like to see
> > more whitespace?
> >
> Perhaps lack of whitespace wasn't a good way of putting it - my main point
> was that a great big lump of C code like that is fairly unreadable.
> 

What about this:

    for (module_num = 0; module_num < num_modules; module_num++) {
        module = MR_module_infos[module_num];
        /* 
        ** Check if the label occurs in this module.
        */
        if (MR_streq(Module, module->MR_ml_name)) {
            num_files = module->MR_ml_filename_count;
            for (file_num = 0; file_num < num_files; file_num++) {
                file = module->MR_ml_module_file_layout[file_num];
                num_labels = file->MR_mfl_label_count;
                for (label_num = 0; label_num < num_labels; label_num++) {
                    label = file->MR_mfl_label_layout[label_num];
                    proc = label->MR_sll_entry;
                    id = &proc->MR_sle_user;
                    /*
                    ** Check if the proc name, arity and mode are the same as
		    ** the name, arity and mode for the label we are interested
		    ** in.
                    */
                    if (MR_streq(id->MR_user_name, Name) &&
                            id->MR_user_arity == Arity &&
                            id->MR_user_mode == ModeNo) {
                        /*
                        ** Check if the path/port is the same as the
                        ** label we are interested in.
                        */
                        MR_TRACE_CALL_MERCURY(
                            MR_MDB_same_path_port(
                                (MR_String) MR_label_goal_path(label),
                                label->MR_sll_port, PathPort, &are_same);
                        );
                        if (are_same) {
                            /*
                            ** We have found the label, so return the 
                            ** filename and line number.
                            */
                            filename = file->MR_mfl_filename;
                            LineNo = file->MR_mfl_label_lineno[label_num];
                            SUCCESS_INDICATOR = MR_TRUE;
                            MR_TRACE_USE_HP(
                                MR_make_aligned_string(FileName,
                                    (MR_String) filename);
                            );
                            goto end;
                        }
                    }
                }
            }
        }
    }

Note I also removed the call to MR_find_context, since it wasn't really
necessary.

> > > > + at sp 1
> > > > + at item dice [-p at var{filename}] [-f at var{filename}] [-n at var{num}] [-s[pPfFsS]+] [-o @var{filename}]
> > > > + at kindex dice (mdb command)
> > > > +Display a program dice on the screen.
> > > > + at sp 1
> > > > +A dice is a comparison between some successful test runs of the program and a
> > > > +failing test run.  Before using the @samp{dice} command one or more passing
> > > > +execution summaries and one failing execution summary should be generated.
> > > > +This can be done by compiling the program with deep tracing enabled (either by
> > > > +compiling in the .debug or .decldebug grades or with the @samp{--trace deep}
> > >
> > > The .decldebug grade is not currently documented.
> > >
> >
> > Okay, I commented out the .decldebug bit.
> >
> Might it not be better to just document the .decldebug grades since they
> are now useful for the subterm dependency tracking anyway?  Perhaps,
> as a separate change though.
> 

Yes that's a seperate change.  I noticed that the .decldebug grade
documentation is commented out in several places, so I need to go
through and uncomment it.

Ian.
--------------------------------------------------------------------------
mercury-reviews mailing list
post:  mercury-reviews at cs.mu.oz.au
administrative address: owner-mercury-reviews at cs.mu.oz.au
unsubscribe: Address: mercury-reviews-request at cs.mu.oz.au Message: unsubscribe
subscribe:   Address: mercury-reviews-request at cs.mu.oz.au Message: subscribe
--------------------------------------------------------------------------