OpenEuphoria

Re: Win32Lib/Llama Status

• Posted by Pete Eberlein <xseal at HARBORSIDE.COM> Jun 25, 1999
• 421 views

Rob Craig writes:
> How the binder currently works:
>
> The binder takes a multi-file program, consisting of a main
> file and all necessary include files, and it converts it into
> a single large file. In general, this shouldn't be possible if there
> are local symbols in 2 different files that have the same name.
> However, the binder changes all names to short,
> meaningless names of it's own choosing. fred() in x.e
> might be changed to aa(), while fred() in y.e might be changed
> to bb(), so there's no conflict.
>
> Now, suppose the programmer has routine_id("fred") in his
> program. fred() has been changed to aa(), so this routine_id() call is
> now wrong. The binder will issue a warning, and you must rebind with
> -clear_routines. This will force the binder to leave fred() in x.e and
> fred() in y.e unchanged. But now when the second instance of fred is
> encountered, the binder will issue a warning, and change the 2nd fred
> anyway, in the hope that you aren't calling routine_id for the second
> fred(). If you ignore the warning, you can run, but things will likely
> fail.

I think this could be averted by changing bind in the following way...

get rid of the -clear_routines option.
whenever the routine_id() function is used, replace "routine_id" with
another unique meaningless name.  Then have bind generate a function for
that meaningless name and insert the new function body somewhere before
the place where routine_id() is called.  The code that goes in the
function will translate an unmangled name to the meaningless name for each
routine currently visible at the place where routine_id was called.

For example:

  procedure foo()  --> gets renamed to aa() by bind
  -- ...
  end procedure

  procedure bar()  --> gets renamed to ab() by bind
    integer id
    id = routine_id("foo")  --> routine_id gets renamed to ac() by bind
  end procedure

Bind will output something equivalent to:

  procedure ab()  --> this used to be foo()
  -- ..
  end procedure

  function ac(sequence routine) --> this is a generated routine_id()
    sequence names
    sequence ids
    integer id
    -- list of names currently visible
    names = {"foo"}
    -- only bind will know this list of meaningless names
    ids = {"ab"}
    -- lookup the desired routine name
    id = find(routines, names)
    -- is it valid?
    if id > 0 then
       -- return the routine_id of the actual routine
       return routine_id(ids[id])
    end if
    -- unknown routine
    return -1
  end function

  procedure ab()  --> this used to be bar()
    integer id
    id = ac("foo")  --> this used to be routine_id()
  end procedure


A difficulty I forsee in this method is the problem of inserting the
generated routine_id() functions before the place where they are called.
Bind may have to buffer routines or nested statements, in case they used
routine_id() so that they can be output after the new routine_id()
function.  Also, using routine_id() in many different places can generate
a lot of extra code for each new routine_id() function.

A trick I used in Peu for routine_id was to define routine_id internally
to accept *two* parameters, one for the name and the other for an index
into the symbol table.  And then the parser watches for a call to
routine_id and secretly passes the extra parameter for the currently
visible symbol table.  That way routine_id can figure out the
correct id by looking in the right symbol table based on where routine_id
was called from.  Perhaps bind could do something similar.

> In the redesign, I hope to get away from the concept
> of creating "one" file. The source file
> boundaries would be retained somehow, solving this
> routine_id situation and making other features available,
> such as the option to bind clear, unshrouded source, and
> get proper ex.err dumps etc.

The Peu symbol table keeps track of file boundaries by storing special
symbols for the start and end of each include file.  When searching for a
name in the table, it sees when it enters into an include file so that it
can skip over non-global names.  These special symbols could be
made available to the programmer (or bind) as euphoria keywords such as
"namespace" and "end namespace"

> Regards,
>      Rob Craig
>      Rapid Deployment Software
>      http://members.aol.com/FilesEu/



Regards,
 _______  ______  _______  ______
[    _  \[    _ ][ _   _ ][    _ ]
[/| [_] |[/| [_\][/ | | \][/| [_\]
  |  ___/  |  _]    | |     |  _]
[\| [/]  [\| [_/] [\| |/] [\| [_/]
[_____]  [______] [_____] [______]
xseal at harborside.com  ICQ:13466657
http://www.harborside.com/home/x/xseal/euphoria/