guile/doc/ref/gh.texi

@c -*-texinfo-*-
@c This is part of the GNU Guile Reference Manual.
@c Copyright (C)  1996, 1997, 2000, 2001, 2002, 2003, 2004
@c   Free Software Foundation, Inc.
@c See the file guile.texi for copying conditions.

@page
@node GH
@section GH: A Portable C to Scheme Interface
@cindex libguile - gh
@cindex gh
@cindex gh - reference manual

This chapter shows how to use the GH interface to call Guile from your
application's C code, and to add new Scheme level procedures to Guile
whose behaviour is specified by application specific code written in C.

Note, however, that the GH interface is now deprecated, and developers
are encouraged to switch to using the scm interface instead.  Therefore,
for each GH feature, this chapter also documents how to achieve
the same result using the scm interface.

@menu
* GH deprecation::              Why the GH interface is now deprecated.
* Transitioning away from GH::
* GH preliminaries::            
* Data types and constants defined by GH::  
* Starting and controlling the interpreter::  
* Error messages::              
* Executing Scheme code::       
* Defining new Scheme procedures in C::  
* Converting data between C and Scheme::  
* Type predicates::             
* Equality predicates::         
* Memory allocation and garbage collection::  
* Calling Scheme procedures from C::  
@end menu


@node GH deprecation
@subsection Why the GH Interface is Now Deprecated

Historically, the GH interface was the product of a practical problem
and a neat idea.  The practical problem was that the interface of the
@code{scm_} functions with which Guile itself was written (inherited
from Aubrey Jaffer's SCM) was so closely tied to the (rather arcane)
details of the internal data representation that it was extremely
difficult to write a Guile extension using these functions.  The neat
idea was to define a high level language extension interface in such a
way that other extension language projects, not just Guile, would be
able to provide an implementation of that interface; then applications
using this interface could be compiled with whichever of the various
available implementations they chose.  So the GH interface was created,
and advertised both as the recommended interface for application
developers wishing to use Guile, and as a portable high level interface
that could theoretically be implemented by other extension language
projects.

Time passed, and various things changed.  Crucially, an enormous number
of improvements were made to the @code{scm_} interface that Guile itself
uses in its implementation, with the result that it is now both easy and
comfortable to write a Guile extension with this interface.  At the same
time, the contents of the GH interface were somewhat neglected by the
core Guile developers, such that some key operations --- such as smob
creation and management --- are simply not possible using GH alone.
Finally, the idea of multiple implementations of the GH interface did
not really crystallize (apart, I believe, from a short lived
implementation by the MzScheme project).

For all these reasons, the Guile developers have decided to deprecate
the GH interface --- which means that support for GH will be completely
removed after the next few releases --- and to focus only on the
@code{scm_} interface, with additions to ensure that it is as easy to
use in all respects as GH was.

It remains an open question whether a deep kind of interface portability
would be useful for extension language-based applications, and it may
still be an interesting project to attempt to define a corresponding
GH-like interface, but the Guile developers no longer plan to try to do
this as part of the core Guile project.

@node Transitioning away from GH
@subsection Transitioning away from GH

The following table summarizes how to transition from the GH to the
scm interface.  The replacements that are recommended are not always
completely equivalent to the GH functionality that they should
replace.  Therefore, you should read the reference documentation of
the replacements carefully if you are not yet familiar with them.

@table @asis
@item Header file
Use @code{#include <libguile.h>} instead of @code{#include
<guile/gh.h>}.

@item Compiling and Linking
Use @code{guile-config} to pick up the flags required to compile C or
C++ code that uses @code{libguile}, like so

@smallexample
$(CC) -o prog.o -c prog.c `guile-config compile`
@end smallexample

If you are using libtool to link your executables, just use
@code{-lguile} in your link command.  Libtool will expand this into
the needed linker options automatically.  If you are not using
libtool, use the @code{guile-config} program to query the needed
options explicitly.  A linker command like

@smallexample
$(CC) -o prog prog.o `guile-config link`
@end smallexample

should be all that is needed.  To link shared libraries that will be
used as Guile Extensions, use libtool to control both the compilation
and the link stage.

@item The @code{SCM} type
No change: the scm interface also uses this type to represent an
arbitrary Scheme value.

@item @code{SCM_BOOL_F} and @code{SCM_BOOL_T}
No change.

@item @code{SCM_UNSPECIFIED} and @code{SCM_UNDEFINED}
No change.

@item @code{gh_enter}
Use @code{scm_boot_guile} instead, but note that @code{scm_boot_guile}
has a slightly different calling convention from @code{gh_enter}:
@code{scm_boot_guile}, and the main program function that you specify
for @code{scm_boot_guile} to call, both take an additional @var{closure}
parameter.  @ref{Guile Initialization Functions} for more details.

@item @code{gh_repl}
Use @code{scm_shell} instead.

@item @code{gh_init}
Use @code{scm_init_guile} instead.

@item @code{gh_catch}
Use @code{scm_internal_catch} instead.

@item @code{gh_eval_str}
Use @code{scm_c_eval_string} instead.

@item @code{gh_eval_str_with_catch}
Use @code{scm_c_eval_string} together with @code{scm_internal_catch}
instead.

@item @code{gh_eval_str_with_standard_handler}
Use @code{scm_c_eval_string} together with @code{scm_internal_catch}
and @code{scm_handle_by_message_no_exit} instead.

@item @code{gh_eval_str_with_stack_saving_handler}
Use @code{scm_c_eval_string} together with
@code{scm_internal_stack_catch} and
@code{scm_handle_by_message_no_exit} instead.

@item @code{gh_eval_file} or @code{gh_load}
Use @code{scm_c_primitive_load} instead.

@item @code{gh_eval_file_with_catch}
Use @code{scm_c_primitive_load} together with
@code{scm_internal_catch} instead.

@item @code{gh_eval_file_with_standard_handler}
Use @code{scm_c_primitive_load} together with
@code{scm_internal_catch} and @code{scm_handle_by_message_no_exit}
instead.

@item  @code{gh_new_procedure}
@itemx @code{gh_new_procedure0_0}
@itemx @code{gh_new_procedure0_1}
@itemx @code{gh_new_procedure0_2}
@itemx @code{gh_new_procedure1_0}
@itemx @code{gh_new_procedure1_1}
@itemx @code{gh_new_procedure1_2}
@itemx @code{gh_new_procedure2_0}
@itemx @code{gh_new_procedure2_1}
@itemx @code{gh_new_procedure2_2}
@itemx @code{gh_new_procedure3_0} 
@itemx @code{gh_new_procedure4_0} 
@itemx @code{gh_new_procedure5_0} 
Use @code{scm_c_define_gsubr} instead, but note that the arguments are
in a different order: for @code{scm_c_define_gsubr} the C function
pointer is the last argument.  @ref{A Sample Guile Extension} for an
example.

@item @code{gh_defer_ints} and @code{gh_allow_ints}
Use @code{SCM_CRITICAL_SECTION_START} and
@code{SCM_CRITICAL_SECTION_END} instead.  Note that these macros are
used without parentheses, as in @code{SCM_DEFER_INTS;}.

@item @code{gh_bool2scm}
Use @code{scm_from_bool} instead.

@item @code{gh_int2scm}
Use @code{scm_from_int} instead.

@item @code{gh_ulong2scm}
Use @code{scm_from_ulong} instead.

@item @code{gh_long2scm}
Use @code{scm_from_long} instead.

@item @code{gh_double2scm}
Use @code{scm_make_real} instead.

@item @code{gh_char2scm}
Use @code{SCM_MAKE_CHAR} instead.

@item @code{gh_str2scm}
Use @code{scm_from_locale_stringn} instead.

@item @code{gh_str02scm}
Use @code{scm_from_locale_string} instead.

@item @code{gh_set_substr}
Use @code{scm_string_copy_x}.

@item @code{gh_symbol2scm}
Use @code{scm_from_locale_symbol} instead.

@item  @code{gh_ints2scm}
@itemx @code{gh_doubles2scm}
@itemx @code{gh_chars2byvect}
@itemx @code{gh_shorts2svect}
@itemx @code{gh_longs2ivect}
@itemx @code{gh_ulongs2uvect}
@itemx @code{gh_floats2fvect}
@itemx @code{gh_doubles2dvect}
Use the uniform numeric vector function, @xref{Uniform Numeric
Vectors}.

@item @code{gh_scm2bool}
Use @code{scm_is_true} or @code{scm_to_bool} instead.

@item @code{gh_scm2int}
Use @code{scm_to_int} instead.

@item @code{gh_scm2ulong}
Use @code{scm_to_ulong} instead.

@item @code{gh_scm2long}
Use @code{scm_to_long} instead.

@item @code{gh_scm2double}
Use @code{scm_to_double} instead.

@item @code{gh_scm2char}
Use @code{scm_to_char} instead.

@item @code{gh_scm2newstr}
Use @code{scm_to_locale_string} or similar instead.

@item @code{gh_get_substr}
Use @code{scm_c_substring} together with @code{scm_to_locale_string}
or similar instead.

@item @code{gh_symbol2newstr}
Use @code{scm_symbol_to_string} together with @code{scm_to_locale_string} or similar instead.

@item @code{gh_scm2chars}
Use @code{scm_from_locale_string} (or similar) or the uniform numeric
vector functions (@pxref{Uniform Numeric Vectors}) instead.

@item  @code{gh_scm2shorts}
@itemx @code{gh_scm2longs}
@itemx @code{gh_scm2floats}
@itemx @code{gh_scm2doubles}
Use the uniform numeric vector function, @xref{Uniform Numeric
Vectors}.

@item @code{gh_boolean_p}
Use @code{scm_is_bool} instead.

@item @code{gh_symbol_p}
Use @code{scm_is_symbol} instead.

@item @code{gh_char_p}
Replace @code{gh_char_p (@var{obj})} with
@example
scm_is_true (scm_char_p (@var{obj}))
@end example

@item @code{gh_vector_p}
Replace @code{gh_vector_p (@var{obj})} with
@example
scm_is_true (scm_vector_p (@var{obj}))
@end example

@item @code{gh_pair_p}
Replace @code{gh_pair_p (@var{obj})} with
@example
scm_is_true (scm_pair_p (@var{obj}))
@end example

@item @code{gh_number_p}
Use @code{scm_is_number} instead.

@item @code{gh_string_p}
Use @code{scm_is_string} instead.

@item @code{gh_procedure_p}
Replace @code{gh_procedure_p (@var{obj})} by
@example
scm_is_true (scm_procedure_p (@var{obj}))
@end example

@item @code{gh_list_p}
Replace @code{gh_list_p (@var{obj})} with
@example
scm_is_true (scm_list_p (@var{obj}))
@end example

@item @code{gh_inexact_p}
Replace @code{gh_inexact_p (@var{obj})} with
@example
scm_is_true (scm_inexact_p (@var{obj}))
@end example

@item @code{gh_exact_p}
Replace @code{gh_exact_p (@var{obj})} with
@example
scm_is_true (scm_exact_p (@var{obj}))
@end example

@item @code{gh_eq_p}
Use @code{scm_is_eq} instead.

@item @code{gh_eqv_p}
Replace @code{gh_eqv_p (@var{x}, @var{y})} with
@example
scm_is_true (scm_eqv_p (@var{x}, @var{y}))
@end example

@item @code{gh_equal_p}
Replace @code{gh_equal_p (@var{x}, @var{y})} with
@example
scm_is_true (scm_equal_p (@var{x}, @var{y}))
@end example

@item @code{gh_string_equal_p}
Replace @code{gh_string_equal_p (@var{x}, @var{y})} with
@example
scm_is_true (scm_string_equal_p (@var{x}, @var{y}))
@end example

@item @code{gh_null_p}
Use @code{scm_is_null} instead.

@item @code{gh_not}
Use @code{scm_not} instead.

@item @code{gh_make_string}
Use @code{scm_make_string} instead.

@item @code{gh_string_length}
Use @code{scm_string_length} instead.

@item @code{gh_string_ref}
Use @code{scm_string_ref} instead.

@item @code{gh_string_set_x}
Use @code{scm_string_set_x} instead.

@item @code{gh_substring}
Use @code{scm_substring} instead.

@item @code{gh_string_append}
Use @code{scm_string_append} instead.

@item @code{gh_cons}
Use @code{scm_cons} instead.

@item @code{gh_car} and @code{gh_cdr}
Use @code{scm_car} and @code{scm_cdr} instead.

@item @code{gh_cxxr} and @code{gh_cxxxr}
(Where each x is either @samp{a} or @samp{d}.)  Use the corresponding
@code{scm_cxxr} or @code{scm_cxxxr} function instead.

@item @code{gh_set_car_x} and @code{gh_set_cdr_x}
Use @code{scm_set_car_x} and @code{scm_set_cdr_x} instead.

@item @code{gh_list}
Use @code{scm_list_n} instead.

@item @code{gh_length}
Replace @code{gh_length (@var{lst})} with
@example
scm_to_size_t (scm_length (@var{lst}))
@end example

@item @code{gh_append}
Use @code{scm_append} instead.

@item @code{gh_append2}, @code{gh_append3}, @code{gh_append4}
Replace @code{gh_append@var{N} (@var{l1}, @dots{}, @var{lN})} by
@example
scm_append (scm_list_n (@var{l1}, @dots{}, @var{lN}, SCM_UNDEFINED))
@end example

@item @code{gh_reverse}
Use @code{scm_reverse} instead.

@item @code{gh_list_tail} and @code{gh_list_ref}
Use @code{scm_list_tail} and @code{scm_list_ref} instead.

@item @code{gh_memq}, @code{gh_memv} and @code{gh_member}
Use @code{scm_memq}, @code{scm_memv} and @code{scm_member} instead.

@item @code{gh_assq}, @code{gh_assv} and @code{gh_assoc}
Use @code{scm_assq}, @code{scm_assv} and @code{scm_assoc} instead.

@item @code{gh_make_vector}
Use @code{scm_make_vector} instead.

@item @code{gh_vector} or @code{gh_list_to_vector}
Use @code{scm_vector} instead.

@item @code{gh_vector_ref} and @code{gh_vector_set_x}
Use @code{scm_vector_ref} and @code{scm_vector_set_x} instead.

@item @code{gh_vector_length}
Use @code{scm_c_vector_length} instead.

@item @code{gh_uniform_vector_length}
Use @code{scm_c_uniform_vector_length} instead.

@item @code{gh_uniform_vector_ref}
Use @code{scm_c_uniform_vector_ref} instead.

@item @code{gh_vector_to_list}
Use @code{scm_vector_to_list} instead.

@item @code{gh_apply}
Use @code{scm_apply_0} instead.

@item  @code{gh_call0}
@itemx @code{gh_call1}
@itemx @code{gh_call2}
@itemx @code{gh_call3}
Use @code{scm_call_0}, @code{scm_call_1}, etc instead.

@item  @code{gh_display}
@itemx @code{gh_write}
@itemx @code{gh_newline}
Use @code{scm_display (obj, scm_current_output_port ())} instead, etc.

@item @code{gh_lookup}
Use @code{scm_variable_ref (scm_c_lookup (name))} instead.

@item @code{gh_module_lookup}
Use @code{scm_variable_ref (scm_c_module_lookup (module, name))} instead.

@end table

@node GH preliminaries
@subsection GH preliminaries

To use gh, you must have the following toward the beginning of your C
source:
@smallexample
#include <guile/gh.h>
@end smallexample
@cindex gh - headers

When you link, you will have to add at least @code{-lguile} to the list
of libraries.  If you are using more of Guile than the basic Scheme
interpreter, you will have to add more libraries.
@cindex gh - linking


@node Data types and constants defined by GH
@subsection Data types and constants defined by GH

The following C constants and data types are defined in gh:

@code{SCM} is a C data type used to store all Scheme data, no matter what the
Scheme type.  Values are converted between C data types and the SCM type
with utility functions described below (@pxref{Converting data between C
and Scheme}).  [FIXME: put in references to Jim's essay and so forth.]

@defvr Constant SCM_BOOL_T
@defvrx Constant SCM_BOOL_F
The @emph{Scheme} values returned by many boolean procedures in
libguile.

This can cause confusion because they are different from 0 and 1.  In
testing a boolean function in libguile programming, you must always make
sure that you check the spec: @code{gh_} and @code{scm_} functions will
usually return @code{SCM_BOOL_T} and @code{SCM_BOOL_F}, but other C
functions usually can be tested against 0 and 1, so programmers' fingers
tend to just type @code{if (boolean_function()) @{ ... @}}
@end defvr

@defvr Constant SCM_UNSPECIFIED
This is a SCM value that is not the same as any legal Scheme value.  It
is the value that a Scheme function returns when its specification says
that its return value is unspecified.
@end defvr

@defvr Constant SCM_UNDEFINED
This is another SCM value that is not the same as any legal Scheme
value.  It is the value used to mark variables that do not yet have a
value, and it is also used in C to terminate functions with variable
numbers of arguments, such as @code{gh_list()}.
@end defvr


@node Starting and controlling the interpreter
@subsection Starting and controlling the interpreter
@cindex libguile - start interpreter

In almost every case, your first @code{gh_} call will be:

@deftypefun void gh_enter (int @var{argc}, char *@var{argv}[], void (*@var{main_prog})())
Starts up a Scheme interpreter with all the builtin Scheme primitives.
@code{gh_enter()} never exits, and the user's code should all be in the
@code{@var{main_prog}()} function.  @code{argc} and @code{argv} will be
passed to @var{main_prog}.

@deftypefun void main_prog (int @var{argc}, char *@var{argv}[])
This is the user's main program.  It will be invoked by
@code{gh_enter()} after Guile has been started up.
@end deftypefun

Note that you can use @code{gh_repl} inside @code{gh_enter} (in other
words, inside the code for @code{main-prog}) if you want the program to
be controlled by a Scheme read-eval-print loop.
@end deftypefun

@cindex read eval print loop -- from the gh_ interface
@cindex REPL -- from the gh_ interface
A convenience routine which enters the Guile interpreter with the
standard Guile read-eval-print loop (@dfn{REPL}) is:

@deftypefun void gh_repl (int @var{argc}, char *@var{argv}[])
Enters the Scheme interpreter giving control to the Scheme REPL.
Arguments are processed as if the Guile program @file{guile} were being
invoked.

Note that @code{gh_repl} should be used @emph{inside} @code{gh_enter},
since any Guile interpreter calls are meaningless unless they happen in
the context of the interpreter.

Also note that when you use @code{gh_repl}, your program will be
controlled by Guile's REPL (which is written in Scheme and has many
useful features).  Use straight C code inside @code{gh_enter} if you
want to maintain execution control in your C program.
@end deftypefun

You will typically use @code{gh_enter} and @code{gh_repl} when you
want a Guile interpreter enhanced by your own libraries, but otherwise
quite normal.  For example, to build a Guile--derived program that
includes some random number routines @dfn{GSL} (GNU Scientific Library),
you would write a C program that looks like this:

@smallexample
#include <guile/gh.h>
#include <gsl_ran.h>

/* random number suite */
SCM gw_ran_seed(SCM s)
@{
  gsl_ran_seed(gh_scm2int(s));
  return SCM_UNSPECIFIED;
@}

SCM gw_ran_random()
@{
  SCM x;

  x = gh_ulong2scm(gsl_ran_random());
  return x;
@}

SCM gw_ran_uniform()
@{
  SCM x;

  x = gh_double2scm(gsl_ran_uniform());
  return x;
@}
SCM gw_ran_max()
@{
  return gh_double2scm(gsl_ran_max());
@}

void
init_gsl()
@{
  /* random number suite */
  gh_new_procedure("gsl-ran-seed", gw_ran_seed, 1, 0, 0);
  gh_new_procedure("gsl-ran-random", gw_ran_random, 0, 0, 0);
  gh_new_procedure("gsl-ran-uniform", gw_ran_uniform, 0, 0, 0);
  gh_new_procedure("gsl-ran-max", gw_ran_max, 0, 0, 0);
@}

void
main_prog (int argc, char *argv[])
@{
  init_gsl();

  gh_repl(argc, argv);
@}

int
main (int argc, char *argv[])
@{
  gh_enter (argc, argv, main_prog);
@}
@end smallexample

Then, supposing the C program is in @file{guile-gsl.c}, you could
compile it with @kbd{gcc -o guile-gsl guile-gsl.c -lguile -lgsl}.

The resulting program @file{guile-gsl} would have new primitive
procedures @code{gsl-ran-random}, @code{gsl-ran-gaussian} and so forth.


@node Error messages
@subsection Error messages
@cindex libguile - error messages
@cindex error messages in libguile

[FIXME: need to fill this based on Jim's new mechanism]


@node Executing Scheme code
@subsection Executing Scheme code
@cindex libguile - executing Scheme
@cindex executing Scheme

Once you have an interpreter running, you can ask it to evaluate Scheme
code.  There are two calls that implement this:

@deftypefun SCM gh_eval_str (char *@var{scheme_code})
This asks the interpreter to evaluate a single string of Scheme code,
and returns the result of the last expression evaluated.

Note that the line of code in @var{scheme_code} must be a well formed
Scheme expression.  If you have many lines of code before you balance
parentheses, you must either concatenate them into one string, or use
@code{gh_eval_file()}.
@end deftypefun

@deftypefun SCM gh_eval_file (char *@var{fname})
@deftypefunx SCM gh_load (char *@var{fname})
@code{gh_eval_file} is completely analogous to @code{gh_eval_str()},
except that a whole file is evaluated instead of a string.
@code{gh_eval_file} returns @code{SCM_UNSPECIFIED}.

@code{gh_load} is identical to @code{gh_eval_file} (it's a macro that
calls @code{gh_eval_file} on its argument).  It is provided to start
making the @code{gh_} interface match the R5RS Scheme procedures
closely.
@end deftypefun


@node Defining new Scheme procedures in C
@subsection Defining new Scheme procedures in C
@cindex libguile - new procedures
@cindex new procedures
@cindex procedures, new
@cindex new primitives
@cindex primitives, new

The real interface between C and Scheme comes when you can write new
Scheme procedures in C.  This is done through the routine


@deftypefn {Libguile high} SCM gh_new_procedure (char *@var{proc_name}, SCM (*@var{fn})(), int @var{n_required_args}, int @var{n_optional_args}, int @var{restp})
@code{gh_new_procedure} defines a new Scheme procedure.  Its Scheme name
will be @var{proc_name}, it will be implemented by the C function
(*@var{fn})(), it will take at least @var{n_required_args} arguments,
and at most @var{n_optional_args} extra arguments.

When the @var{restp} parameter is 1, the procedure takes a final
argument: a list of remaining parameters.

@code{gh_new_procedure} returns an SCM value representing the procedure.

The C function @var{fn} should have the form
@deftypefn {Libguile high} SCM fn (SCM @var{req1}, SCM @var{req2}, ..., SCM @var{opt1},  SCM @var{opt2}, ...,  SCM @var{rest_args})
The arguments are all passed as SCM values, so the user will have to use
the conversion functions to convert to standard C types.

Examples of C functions used as new Scheme primitives can be found in
the sample programs @code{learn0} and @code{learn1}.
@end deftypefn

@end deftypefn

@strong{Rationale:} this is the correct way to define new Scheme
procedures in C.  The ugly mess of arguments is required because of how
C handles procedures with variable numbers of arguments.

@strong{NB:} what about documentation strings?

@cartouche
There are several important considerations to be made when writing the C
routine @code{(*fn)()}.

First of all the C routine has to return type @code{SCM}.

Second, all arguments passed to the C function will be of type
@code{SCM}.

Third: the C routine is now subject to Scheme flow control, which means
that it could be interrupted at any point, and then reentered.  This
means that you have to be very careful with operations such as
allocating memory, modifying static data @dots{}

Fourth: to get around the latter issue, you can use
@code{GH_DEFER_INTS} and @code{GH_ALLOW_INTS}.
@end cartouche

@defmac GH_DEFER_INTS
@defmacx GH_ALLOW_INTS
These macros disable and re-enable Scheme's flow control.  They 
@end defmac


@c [??? have to do this right; maybe using subsections, or maybe creating a
@c section called Flow control issues...]

@c [??? Go into exhaustive detail with examples of the various possible
@c combinations of required and optional args...]


@node Converting data between C and Scheme
@subsection Converting data between C and Scheme
@cindex libguile - converting data
@cindex data conversion
@cindex converting data

Guile provides mechanisms to convert data between C and Scheme.  This
allows new builtin procedures to understand their arguments (which are
of type @code{SCM}) and return values of type @code{SCM}.


@menu
* C to Scheme::                 
* Scheme to C::                 
@end menu

@node C to Scheme
@subsubsection C to Scheme

@deftypefun SCM gh_bool2scm (int @var{x})
Returns @code{#f} if @var{x} is zero, @code{#t} otherwise.
@end deftypefun

@deftypefun SCM gh_ulong2scm (unsigned long @var{x})
@deftypefunx SCM gh_long2scm (long @var{x})
@deftypefunx SCM gh_double2scm (double @var{x})
@deftypefunx SCM gh_char2scm (char @var{x})
Returns a Scheme object with the value of the C quantity @var{x}.
@end deftypefun

@deftypefun SCM gh_str2scm (char *@var{s}, int @var{len})
Returns a new Scheme string with the (not necessarily null-terminated) C
array @var{s} data.
@end deftypefun

@deftypefun SCM gh_str02scm (char *@var{s})
Returns a new Scheme string with the null-terminated C string @var{s}
data.
@end deftypefun

@deftypefun SCM gh_set_substr (char *@var{src}, SCM @var{dst}, int @var{start}, int @var{len})
Copy @var{len} characters at @var{src} into the @emph{existing} Scheme
string @var{dst}, starting at @var{start}.  @var{start} is an index into
@var{dst}; zero means the beginning of the string.

If @var{start} + @var{len} is off the end of @var{dst}, signal an
out-of-range error.
@end deftypefun

@deftypefun SCM gh_symbol2scm (char *@var{name})
Given a null-terminated string @var{name}, return the symbol with that
name.
@end deftypefun

@deftypefun SCM gh_ints2scm (int *@var{dptr}, int @var{n})
@deftypefunx SCM gh_doubles2scm (double *@var{dptr}, int @var{n})
Make a scheme vector containing the @var{n} ints or doubles at memory
location @var{dptr}.
@end deftypefun

@deftypefun SCM gh_chars2byvect (char *@var{dptr}, int @var{n})
@deftypefunx SCM gh_shorts2svect (short *@var{dptr}, int @var{n})
@deftypefunx SCM gh_longs2ivect (long *@var{dptr}, int @var{n})
@deftypefunx SCM gh_ulongs2uvect (ulong *@var{dptr}, int @var{n})
@deftypefunx SCM gh_floats2fvect (float *@var{dptr}, int @var{n})
@deftypefunx SCM gh_doubles2dvect (double *@var{dptr}, int @var{n})
Make a scheme uniform vector containing the @var{n} chars, shorts,
longs, unsigned longs, floats or doubles at memory location @var{dptr}.
@end deftypefun



@node Scheme to C
@subsubsection Scheme to C

@deftypefun int gh_scm2bool (SCM @var{obj})
@deftypefunx {unsigned long} gh_scm2ulong (SCM @var{obj})
@deftypefunx long gh_scm2long (SCM @var{obj})
@deftypefunx double gh_scm2double (SCM @var{obj})
@deftypefunx int gh_scm2char (SCM @var{obj})
These routines convert the Scheme object to the given C type.
@end deftypefun

@deftypefun {char *} gh_scm2newstr (SCM @var{str}, size_t *@var{lenp})
Given a Scheme string @var{str}, return a pointer to a new copy of its
contents, followed by a null byte.  If @var{lenp} is non-null, set
@code{*@var{lenp}} to the string's length.

This function uses malloc to obtain storage for the copy; the caller is
responsible for freeing it.

Note that Scheme strings may contain arbitrary data, including null
characters.  This means that null termination is not a reliable way to
determine the length of the returned value.  However, the function
always copies the complete contents of @var{str}, and sets @var{*lenp}
to the true length of the string (when @var{lenp} is non-null).
@end deftypefun


@deftypefun void gh_get_substr (SCM str, char *return_str, int *lenp)
Copy @var{len} characters at @var{start} from the Scheme string
@var{src} to memory at @var{dst}.  @var{start} is an index into
@var{src}; zero means the beginning of the string.  @var{dst} has
already been allocated by the caller.

If @var{start} + @var{len} is off the end of @var{src}, signal an
out-of-range error.
@end deftypefun

@deftypefun {char *} gh_symbol2newstr (SCM @var{sym}, int *@var{lenp})
Takes a Scheme symbol and returns a string of the form
@code{"'symbol-name"}.  If @var{lenp} is non-null, the string's length
is returned in @code{*@var{lenp}}.

This function uses malloc to obtain storage for the returned string; the
caller is responsible for freeing it.
@end deftypefun

@deftypefun {char *} gh_scm2chars (SCM @var{vector}, chars *@var{result})
@deftypefunx {short *} gh_scm2shorts (SCM @var{vector}, short *@var{result})
@deftypefunx {long *} gh_scm2longs (SCM @var{vector}, long *@var{result})
@deftypefunx {float *} gh_scm2floats (SCM @var{vector}, float *@var{result})
@deftypefunx {double *} gh_scm2doubles (SCM @var{vector}, double *@var{result})
Copy the numbers in @var{vector} to the array pointed to by @var{result}
and return it.  If @var{result} is NULL, allocate a double array large
enough.

@var{vector} can be an ordinary vector, a weak vector, or a signed or
unsigned uniform vector of the same type as the result array.  For
chars, @var{vector} can be a string or substring.  For floats and
doubles, @var{vector} can contain a mix of inexact and integer values.

If @var{vector} is of unsigned type and contains values too large to fit
in the signed destination array, those values will be wrapped around,
that is, data will be copied as if the destination array was unsigned.
@end deftypefun


@node Type predicates
@subsection Type predicates

These C functions mirror Scheme's type predicate procedures with one
important difference.  The C routines return C boolean values (0 and 1)
instead of @code{SCM_BOOL_T} and @code{SCM_BOOL_F}.

The Scheme notational convention of putting a @code{?} at the end of
predicate procedure names is mirrored in C by placing @code{_p} at the
end of the procedure.  For example, @code{(pair? ...)} maps to
@code{gh_pair_p(...)}.

@deftypefun int gh_boolean_p (SCM @var{val})
Returns 1 if @var{val} is a boolean, 0 otherwise.
@end deftypefun

@deftypefun int gh_symbol_p (SCM @var{val})
Returns 1 if @var{val} is a symbol, 0 otherwise.
@end deftypefun

@deftypefun int gh_char_p (SCM @var{val})
Returns 1 if @var{val} is a char, 0 otherwise.
@end deftypefun

@deftypefun int gh_vector_p (SCM @var{val})
Returns 1 if @var{val} is a vector, 0 otherwise.
@end deftypefun

@deftypefun int gh_pair_p (SCM @var{val})
Returns 1 if @var{val} is a pair, 0 otherwise.
@end deftypefun

@deftypefun int gh_procedure_p (SCM @var{val})
Returns 1 if @var{val} is a procedure, 0 otherwise.
@end deftypefun

@deftypefun int gh_list_p (SCM @var{val})
Returns 1 if @var{val} is a list, 0 otherwise.
@end deftypefun

@deftypefun int gh_inexact_p (SCM @var{val})
Returns 1 if @var{val} is an inexact number, 0 otherwise.
@end deftypefun

@deftypefun int gh_exact_p (SCM @var{val})
Returns 1 if @var{val} is an exact number, 0 otherwise.
@end deftypefun


@node Equality predicates
@subsection Equality predicates

These C functions mirror Scheme's equality predicate procedures with one
important difference.  The C routines return C boolean values (0 and 1)
instead of @code{SCM_BOOL_T} and @code{SCM_BOOL_F}.

The Scheme notational convention of putting a @code{?} at the end of
predicate procedure names is mirrored in C by placing @code{_p} at the
end of the procedure.  For example, @code{(equal? ...)} maps to
@code{gh_equal_p(...)}.

@deftypefun int gh_eq_p (SCM x, SCM y)
Returns 1 if @var{x} and @var{y} are equal in the sense of Scheme's
@code{eq?} predicate, 0 otherwise.
@end deftypefun

@deftypefun int gh_eqv_p (SCM x, SCM y)
Returns 1 if @var{x} and @var{y} are equal in the sense of Scheme's
@code{eqv?} predicate, 0 otherwise.
@end deftypefun

@deftypefun int gh_equal_p (SCM x, SCM y)
Returns 1 if @var{x} and @var{y} are equal in the sense of Scheme's
@code{equal?} predicate, 0 otherwise.
@end deftypefun

@deftypefun int gh_string_equal_p (SCM @var{s1}, SCM @var{s2})
Returns 1 if the strings @var{s1} and @var{s2} are equal, 0 otherwise.
@end deftypefun

@deftypefun int gh_null_p (SCM @var{l})
Returns 1 if @var{l} is an empty list or pair; 0 otherwise.
@end deftypefun


@node Memory allocation and garbage collection
@subsection Memory allocation and garbage collection

@c [FIXME: flesh this out with some description of garbage collection in
@c scm/guile]

@c @deftypefun SCM gh_mkarray (int size)
@c Allocate memory for a Scheme object in a garbage-collector-friendly
@c manner.
@c @end deftypefun


@node Calling Scheme procedures from C
@subsection Calling Scheme procedures from C

Many of the Scheme primitives are available in the @code{gh_}
interface; they take and return objects of type SCM, and one could
basically use them to write C code that mimics Scheme code.

I will list these routines here without much explanation, since what
they do is the same as documented in @ref{Standard procedures, R5RS, ,
r5rs, R5RS}.  But I will point out that when a procedure takes a
variable number of arguments (such as @code{gh_list}), you should pass
the constant @var{SCM_UNDEFINED} from C to signify the end of the list.

@deftypefun SCM gh_define (char *@var{name}, SCM @var{val})
Corresponds to the Scheme @code{(define name val)}: it binds a value to
the given name (which is a C string).  Returns the new object.
@end deftypefun

@heading Pairs and lists

@deftypefun SCM gh_cons (SCM @var{a}, SCM @var{b})
@deftypefunx SCM gh_list (SCM l0, SCM l1, ... , SCM_UNDEFINED)
These correspond to the Scheme @code{(cons a b)} and @code{(list l0 l1
...)} procedures.  Note that @code{gh_list()} is a C macro that invokes
@code{scm_list_n()}.
@end deftypefun

@deftypefun SCM gh_car (SCM @var{obj})
@deftypefunx SCM gh_cdr (SCM @var{obj})
@dots{}

@deftypefunx SCM gh_c[ad][ad][ad][ad]r (SCM @var{obj})
These correspond to the Scheme @code{(caadar ls)} procedures etc @dots{}
@end deftypefun

@deftypefun SCM gh_set_car_x (SCM @var{pair}, SCM @var{value})
Modifies the CAR of @var{pair} to be @var{value}.  This is equivalent to
the Scheme procedure @code{(set-car! ...)}.
@end deftypefun

@deftypefun SCM gh_set_cdr_x (SCM @var{pair}, SCM @var{value})
Modifies the CDR of @var{pair} to be @var{value}.  This is equivalent to
the Scheme procedure @code{(set-cdr! ...)}.
@end deftypefun

@deftypefun {unsigned long} gh_length (SCM @var{ls})
Returns the length of the list.
@end deftypefun

@deftypefun SCM gh_append (SCM @var{args})
@deftypefunx SCM gh_append2 (SCM @var{l1}, SCM @var{l2})
@deftypefunx SCM gh_append3 (SCM @var{l1}, SCM @var{l2}, @var{l3})
@deftypefunx SCM gh_append4 (SCM @var{l1}, SCM @var{l2}, @var{l3}, @var{l4})
@code{gh_append()} takes @var{args}, which is a list of lists
@code{(list1 list2 ...)}, and returns a list containing all the elements
of the individual lists.

A typical invocation of @code{gh_append()} to append 5 lists together
would be
@smallexample
  gh_append(gh_list(l1, l2, l3, l4, l5, SCM_UNDEFINED));
@end smallexample

The functions @code{gh_append2()}, @code{gh_append2()},
@code{gh_append3()} and @code{gh_append4()} are convenience routines to
make it easier for C programs to form the list of lists that goes as an
argument to @code{gh_append()}.
@end deftypefun

@deftypefun SCM gh_reverse (SCM @var{ls})
Returns a new list that has the same elements as @var{ls} but in the
reverse order.  Note that this is implemented as a macro which calls
@code{scm_reverse()}.
@end deftypefun

@deftypefun SCM gh_list_tail (SCM @var{ls}, SCM @var{k})
Returns the sublist of @var{ls} with the last @var{k} elements.
@end deftypefun

@deftypefun SCM gh_list_ref (SCM @var{ls}, SCM @var{k})
Returns the @var{k}th element of the list @var{ls}.
@end deftypefun

@deftypefun SCM gh_memq (SCM @var{x}, SCM @var{ls})
@deftypefunx SCM gh_memv (SCM @var{x}, SCM @var{ls})
@deftypefunx SCM gh_member (SCM @var{x}, SCM @var{ls})
These functions return the first sublist of @var{ls} whose CAR is
@var{x}.  They correspond to @code{(memq x ls)}, @code{(memv x ls)} and
@code{(member x ls)}, and hence use (respectively) @code{eq?},
@code{eqv?} and @code{equal?} to do comparisons.

If @var{x} does not appear in @var{ls}, the value @code{SCM_BOOL_F} (not
the empty list) is returned.

Note that these functions are implemented as macros which call
@code{scm_memq()}, @code{scm_memv()} and @code{scm_member()}
respectively.
@end deftypefun

@deftypefun SCM gh_assq (SCM @var{x}, SCM @var{alist})
@deftypefunx SCM gh_assv (SCM @var{x}, SCM @var{alist})
@deftypefunx SCM gh_assoc (SCM @var{x}, SCM @var{alist})
These functions search an @dfn{association list} (list of pairs)
@var{alist} for the first pair whose CAR is @var{x}, and they return
that pair.

If no pair in @var{alist} has @var{x} as its CAR, the value
@code{SCM_BOOL_F} (not the empty list) is returned.

Note that these functions are implemented as macros which call
@code{scm_assq()}, @code{scm_assv()} and @code{scm_assoc()}
respectively.
@end deftypefun


@heading Symbols

@c @deftypefun SCM gh_symbol (SCM str, SCM len)
@c @deftypefunx SCM gh_tmp_symbol (SCM str, SCM len)
@c Takes the given string @var{str} of length @var{len} and returns a
@c symbol corresponding to that string.
@c @end deftypefun


@heading Vectors

@deftypefun SCM gh_make_vector (SCM @var{n}, SCM @var{fill})
@deftypefunx SCM gh_vector (SCM @var{ls})
@deftypefunx SCM gh_vector_ref (SCM @var{v}, SCM @var{i})
@deftypefunx SCM gh_vector_set (SCM @var{v}, SCM @var{i}, SCM @var{val})
@deftypefunx {unsigned long} gh_vector_length (SCM @var{v})
@deftypefunx SCM gh_list_to_vector (SCM @var{ls})
These correspond to the Scheme @code{(make-vector n fill)},
@code{(vector a b c ...)} @code{(vector-ref v i)} @code{(vector-set v i
value)} @code{(vector-length v)} @code{(list->vector ls)} procedures.

The correspondence is not perfect for @code{gh_vector}: this routine
takes a list @var{ls} instead of the individual list elements, thus
making it identical to @code{gh_list_to_vector}.

There is also a difference in gh_vector_length: the value returned is a
C @code{unsigned long} instead of an SCM object.
@end deftypefun


@heading Procedures

@c @deftypefun SCM gh_make_subr (SCM (*@var{fn})(), int @var{req}, int @var{opt}, int @var{restp}, char *@var{sym})
@c Make the C function @var{fn} available to Scheme programs.  The function
@c will be bound to the symbol @var{sym}.  The arguments @var{req},
@c @var{opt} and @var{restp} describe @var{fn}'s calling conventions.  The
@c function must take @var{req} required arguments and may take @var{opt}
@c optional arguments.  Any optional arguments which are not supplied by
@c the caller will be bound to @var{SCM_UNSPECIFIED}.  If @var{restp} is
@c non-zero, it means that @var{fn} may be called with an arbitrary number
@c of arguments, and that any extra arguments supplied by the caller will
@c be passed to @var{fn} as a list.  The @var{restp} argument is exactly
@c like Scheme's @code{(lambda (arg1 arg2 . arglist))} calling convention.
@c 
@c For example, the procedure @code{read-line}, which takes optional
@c @var{port} and @var{handle-delim} arguments, would be declared like so:
@c 
@c @example
@c SCM scm_read_line (SCM port, SCM handle_delim);
@c gh_make_subr (scm_read_line, 0, 2, 0, "read-line");
@c @end example
@c 
@c The @var{req} argument to @code{gh_make_subr} is 0 to indicate that
@c there are no required arguments, so @code{read-line} may be called
@c without any arguments at all.  The @var{opt} argument is 2, to indicate
@c that both the @var{port} and @var{handle_delim} arguments to
@c @code{scm_read_line} are optional, and will be bound to
@c @code{SCM_UNSPECIFIED} if the calling program does not supply them.
@c Because the @var{restp} argument is 0, this function may not be called
@c with more than two arguments.
@c @end deftypefun

@deftypefun SCM gh_apply (SCM proc, SCM args)
Call the Scheme procedure @var{proc}, with the elements of @var{args} as
arguments.  @var{args} must be a proper list.  
@end deftypefun

@deftypefun SCM gh_call0 (SCM proc)
@deftypefunx SCM gh_call1 (SCM proc, SCM arg)
@deftypefunx SCM gh_call2 (SCM proc, SCM arg1, SCM arg2)
@deftypefunx SCM gh_call3 (SCM proc, SCM arg1, SCM arg2, SCM arg3)
Call the Scheme procedure @var{proc} with no arguments
(@code{gh_call0}), one argument (@code{gh_call1}), and so on.  You can
get the same effect by wrapping the arguments up into a list, and
calling @code{gh_apply}; Guile provides these functions for convenience.
@end deftypefun


@deftypefun SCM gh_catch (SCM key, SCM thunk, SCM handler)
@deftypefunx SCM gh_throw (SCM key, SCM args)
Corresponds to the Scheme @code{catch} and @code{throw} procedures,
which in Guile are provided as primitives.
@end deftypefun

@c [FIXME: must add the I/O section in gscm.h]

@deftypefun SCM gh_is_eq (SCM a, SCM b)
@deftypefunx SCM gh_is_eqv (SCM a, SCM b)
@deftypefunx SCM gh_is_equal (SCM a, SCM b)
These correspond to the Scheme @code{eq?}, @code{eqv?} and @code{equal?}
predicates.
@end deftypefun

@deftypefun int gh_obj_length (SCM @var{obj})
Returns the raw object length.
@end deftypefun

@heading Data lookup

For now I just include Tim Pierce's comments from the @file{gh_data.c}
file; it should be organized into a documentation of the two functions
here.

@smallexample
/* Data lookups between C and Scheme

   Look up a symbol with a given name, and return the object to which
   it is bound.  gh_lookup examines the Guile top level, and
   gh_module_lookup checks the module name space specified by the
   `vec' argument.

   The return value is the Scheme object to which SNAME is bound, or
   SCM_UNDEFINED if SNAME is not bound in the given context. [FIXME:
   should this be SCM_UNSPECIFIED?  Can a symbol ever legitimately be
   bound to SCM_UNDEFINED or SCM_UNSPECIFIED?  What is the difference?
   -twp] */
@end smallexample