reading/writing binary data in structured form

reading/writing binary data in structured form

[Werner LEMBERG]

Folks,


does there exist a major mode which allows definition of templates so
that editing of binary data can be done in a structured way?

Looking around in the internet I wasn't very successful in finding
such a binary editor for GNU/Linux at all.  Any ideas what I could
try?  Since I've never used such a thing I would be glad for getting
advices.  I would use such an editor (mode) to modify and create test
fonts.


    Werner

RE: reading/writing binary data in structured form

[Drew Adams]

> does there exist a major mode which allows definition of templates so
> that editing of binary data can be done in a structured way?

hexl-mode

(A better list for this kind of question might be help-gnu-emacs@gnu.org.)

Re: reading/writing binary data in structured form

[Werner LEMBERG]

> > does there exist a major mode which allows definition of templates
> > so that editing of binary data can be done in a structured way?
>
> hexl-mode
>
> (A better list for this kind of question might be
> help-gnu-emacs@gnu.org.)

You are joking, aren't you?  I want *structured* editing, this is, for
example, defining a template

  struct foo {
    int16 bar;
    int32 baz;
  }

which can be used later on as

  foo[32];

and which gets applied to a certain address, and where, for example, I
can modify a value with

  foo[12].bar=1;

or something similar.

A commercial product which looks quite promising is

  http://www.sweetscape.com/010editor/

but this is for MS Windows only.


     Werner

Re: reading/writing binary data in structured form

[Eli Zaretskii]

> Date: Sat, 26 Apr 2008 16:05:13 +0200 (CEST)
> From: Werner LEMBERG <wl@gnu.org>
> 
> does there exist a major mode which allows definition of templates so
> that editing of binary data can be done in a structured way?

Maybe bindat.el (bundled with Emacs) can be a good beginning of a
solution.

RE: reading/writing binary data in structured form

[Drew Adams]

> > > does there exist a major mode which allows definition of templates
> > > so that editing of binary data can be done in a structured way?
> >
> > hexl-mode
>
> You are joking, aren't you?  I want *structured* editing...

No, I wasn't joking; I was trying to help. 

It wasn't clear to me from your "in a structured way" that you meant anything
like what you describe now. You did not mention that you were familiar with
hexl-mode or explain anything about what you wanted that was different.

Re: reading/writing binary data in structured form

[Werner LEMBERG]

> No, I wasn't joking; I was trying to help. 

OK, thanks.

> It wasn't clear to me from your "in a structured way" that you meant
> anything like what you describe now. You did not mention that you
> were familiar with hexl-mode or explain anything about what you
> wanted that was different.

This is probably caused by the fact that English is not my mother
tongue...


    Werner

Re: reading/writing binary data in structured form

[Werner LEMBERG]

[GNU Emacs 23.0.50.2 of 2008-01-31]

> > does there exist a major mode which allows definition of templates
> > so that editing of binary data can be done in a structured way?
> 
> Maybe bindat.el (bundled with Emacs) can be a good beginning of a
> solution.

Aah, yes, this looks very promising!  Actually, I've never seen any
mentioning of this package in emacs-devel (which is probably a good
thing since it means that it works flawlessly) ...

BTW, it's quite hard to find that package IMHO within Emacs; the only
sensible way which comes to my mind is using `C-h p', going to section
`data'.  However, the files mentioned in this section are unsorted
(which I consider a bad thing -- is this a bug?), and there are many
items which are rather unneeded by Joe User (like most of the
url-xxx.el files).

Is there a better way?  Neither `apropos' nor the manual index gives a
sensible result which means that I've basically exhausted the
suggested searching possibilities as given in the `Help' section of
the Emacs info manual.


    Werner

Re: reading/writing binary data in structured form

[Dan Espen]

Werner LEMBERG <wl@gnu.org> writes:

>> > does there exist a major mode which allows definition of templates
>> > so that editing of binary data can be done in a structured way?
>>
>> hexl-mode
>>
>> (A better list for this kind of question might be
>> help-gnu-emacs@gnu.org.)
...
> A commercial product which looks quite promising is
>
>   http://www.sweetscape.com/010editor/
>
> but this is for MS Windows only.

Emacs edits a few image binary file types that I'm aware of, XPM
images for example, but not with user supplied templates.

Re: reading/writing binary data in structured form

[Eli Zaretskii]

> Date: Sun, 27 Apr 2008 00:44:57 +0200 (CEST)
> From: Werner LEMBERG <wl@gnu.org>
> Cc: emacs-devel@gnu.org
> 
> BTW, it's quite hard to find that package IMHO within Emacs; the only
> sensible way which comes to my mind is using `C-h p', going to section
> `data'.  However, the files mentioned in this section are unsorted
> (which I consider a bad thing -- is this a bug?), and there are many
> items which are rather unneeded by Joe User (like most of the
> url-xxx.el files).
> 
> Is there a better way?  Neither `apropos' nor the manual index gives a
> sensible result which means that I've basically exhausted the
> suggested searching possibilities as given in the `Help' section of
> the Emacs info manual.

Maybe try `apropos-documentation'.

Re: reading/writing binary data in structured form

[Werner LEMBERG]

> > BTW, it's quite hard to find that package IMHO within Emacs; the
> > only sensible way which comes to my mind is using `C-h p', going
> > to section `data'.  [...]
>
> Maybe try `apropos-documentation'.

Hmm, `apropos-documentation bindat' gives just this result:

  risky-local-variable-p
    Function: Non-nil if SYM could be dangerous as a file-local variable.
  It is dangerous if either of these conditions are met:

   * Its `risky-local-variable' property is non-nil.

   * Its name ends with "hook(s)", "function(s)", "form(s)", "map",
     "program", "command(s)", "predicate(s)", "frame-alist",
     "mode-alist", "font-lock-(syntactic-)keyword*",
     "map-alist", or "bindat-spec".

which is rather useless IMHO.


    Werner

Re: reading/writing binary data in structured form

[Eli Zaretskii]

> Date: Sun, 27 Apr 2008 09:05:53 +0200 (CEST)
> Cc: emacs-devel@gnu.org
> From: Werner LEMBERG <wl@gnu.org>
> 
> > > BTW, it's quite hard to find that package IMHO within Emacs; the
> > > only sensible way which comes to my mind is using `C-h p', going
> > > to section `data'.  [...]
> >
> > Maybe try `apropos-documentation'.
> 
> Hmm, `apropos-documentation bindat' gives just this result:
> 
>   risky-local-variable-p
>     Function: Non-nil if SYM could be dangerous as a file-local variable.
>   It is dangerous if either of these conditions are met:

I guess that's because there are no autoload cookies in bindat.el for
anything whose doc string would be picked up by your apropos command.

Kim, is there a reason not to make bindat-pack and bindat-unpack
autoloaded?

Re: reading/writing binary data in structured form

[Kim F. Storm]

Eli Zaretskii <eliz@gnu.org> writes:

> Kim, is there a reason not to make bindat-pack and bindat-unpack
> autoloaded?

Only that I envisioned that people would explicitly require bindat
if they need the features it provides.

-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: reading/writing binary data in structured form

[Stefan Monnier]

>> Kim, is there a reason not to make bindat-pack and bindat-unpack
>> autoloaded?

> Only that I envisioned that people would explicitly require bindat
> if they need the features it provides.

Indeed.  It doesn't seem to make much sense to autoload part of it.
Maybe we should begin by improving the C-h p answer.


        Stefan

Re: reading/writing binary data in structured form

[Eli Zaretskii]

> From: Stefan Monnier <monnier@iro.umontreal.ca>
> Cc: Eli Zaretskii <eliz@gnu.org>,  emacs-devel@gnu.org
> Date: Sun, 27 Apr 2008 21:29:15 -0400
> 
> >> Kim, is there a reason not to make bindat-pack and bindat-unpack
> >> autoloaded?
> 
> > Only that I envisioned that people would explicitly require bindat
> > if they need the features it provides.
> 
> Indeed.  It doesn't seem to make much sense to autoload part of it.

Yes, but can it hurt?  We could put comments there explaining the
reasons, in case at some point in the future doing so for the sake of
documentation search will not be necessary.

> Maybe we should begin by improving the C-h p answer.

I'm for it, but IMHO "C-h p" cannot be a precise tool, because it is
based on too small amount of info (a few keywords), whereas
apropos-documentation has much more to play with.

Re: reading/writing binary data in structured form

[Stefan Monnier]

>> >> Kim, is there a reason not to make bindat-pack and bindat-unpack
>> >> autoloaded?
>> 
>> > Only that I envisioned that people would explicitly require bindat
>> > if they need the features it provides.
>> 
>> Indeed.  It doesn't seem to make much sense to autoload part of it.

> Yes, but can it hurt?  We could put comments there explaining the
> reasons, in case at some point in the future doing so for the sake of
> documentation search will not be necessary.

It's not very satisfactory.  Maybe we should simply mention the package
in the lispref.

>> Maybe we should begin by improving the C-h p answer.
> I'm for it, but IMHO "C-h p" cannot be a precise tool, because it is
> based on too small amount of info (a few keywords), whereas
> apropos-documentation has much more to play with.

I don't think it "cannot be a precise tool" much more than C-h i cannot
be a precise tool.  But I agree that it requires a non-trivial amount
of work.  It would be worthwhile, tho.


        Stefan