erl_marshal.c

来自「OTP是开放电信平台的简称」· C语言 代码 · 共 1,923 行 · 第 1/4 页

/* ``The contents of this file are subject to the Erlang Public License,
 * Version 1.1, (the "License"); you may not use this file except in
 * compliance with the License. You should have received a copy of the
 * Erlang Public License along with this software. If not, it can be
 * retrieved via the world wide web at http://www.erlang.org/.
 * 
 * Software distributed under the License is distributed on an "AS IS"
 * basis, WITHOUT WARRANTY OF ANY KIND, either express or implied. See
 * the License for the specific language governing rights and limitations
 * under the License.
 * 
 * The Initial Developer of the Original Code is Ericsson Utvecklings AB.
 * Portions created by Ericsson are Copyright 1999, Ericsson Utvecklings
 * AB. All Rights Reserved.''
 * 
 *     $Id$
 */
/*
 * Purpose: Decoding and encoding Erlang terms.
 */  
#include "eidef.h"

#include <stdio.h>
#include <stdlib.h>
#include <ctype.h>
#include <sys/types.h>
#include <string.h>

#include "erl_interface.h"
#include "erl_marshal.h"
#include "erl_eterm.h"
#include "erl_malloc.h"
#include "erl_error.h"
#include "erl_internal.h"

#include "eiext.h" /* replaces external.h */
#include "putget.h"

static int is_string(ETERM* term);
#if defined(VXWORKS) && CPU == PPC860
int erl_fp_compare(unsigned *a, unsigned *b);
static void erl_long_to_fp(long l, unsigned *d);
#endif

/* Used when comparing two encoded byte arrays */
/* this global data is ok (from threading point of view) since it is
 * initialized once and never changed
 */

#define CMP_ARRAY_SIZE 256
/* FIXME problem for threaded ? */
static char cmp_array[CMP_ARRAY_SIZE]; 
static int init_cmp_array_p=1; /* initialize array, the first time */

#if defined(VXWORKS) && CPU == PPC860
#include <limits.h>
#endif

#if defined(__GNUC__)
#  define INLINE __inline__
#elif defined(__WIN32__)
#  define INLINE __inline
#else
#  define INLINE
#endif

static int cmp_floats(double f1, double f2);
static INLINE double to_float(long l);

#define ERL_NUM_CMP 1
#define ERL_REF_CMP 3

#define IS_ERL_NUM(t) (cmp_array[t]==ERL_NUM_CMP)

#define CMP_NUM_CLASS_SIZE 256
static unsigned char cmp_num_class[CMP_NUM_CLASS_SIZE]; 
static int init_cmp_num_class_p=1; /* initialize array, the first time */

#define MK_CMP_NUM_CODE(x,y)    (((x)<<2)|(y))
#define CMP_NUM_CLASS(x)        (cmp_num_class[x] & 0x03)
#define CMP_NUM_CODE(x,y)       (MK_CMP_NUM_CODE(CMP_NUM_CLASS(x),CMP_NUM_CLASS(y)))

#define SMALL 1
#define FLOAT 2
#define BIG   3

#define SMALL_SMALL    MK_CMP_NUM_CODE(SMALL,SMALL)
#define SMALL_FLOAT    MK_CMP_NUM_CODE(SMALL,FLOAT)
#define SMALL_BIG      MK_CMP_NUM_CODE(SMALL,BIG)
#define FLOAT_SMALL    MK_CMP_NUM_CODE(FLOAT,SMALL)
#define FLOAT_FLOAT    MK_CMP_NUM_CODE(FLOAT,FLOAT)
#define FLOAT_BIG      MK_CMP_NUM_CODE(FLOAT,BIG)
#define BIG_SMALL      MK_CMP_NUM_CODE(BIG,SMALL)
#define BIG_FLOAT      MK_CMP_NUM_CODE(BIG,FLOAT)
#define BIG_BIG        MK_CMP_NUM_CODE(BIG,BIG)

void erl_init_marshal(void)
{
  if (init_cmp_array_p) {
    memset(cmp_array, 0, CMP_ARRAY_SIZE);
    cmp_array[ERL_SMALL_INTEGER_EXT] = 1;
    cmp_array[ERL_INTEGER_EXT]       = 1;
    cmp_array[ERL_FLOAT_EXT]         = 1;
    cmp_array[ERL_SMALL_BIG_EXT]     = 1;
    cmp_array[ERL_LARGE_BIG_EXT]     = 1;
    cmp_array[ERL_ATOM_EXT]          = 2;
    cmp_array[ERL_REFERENCE_EXT]     = 3;
    cmp_array[ERL_NEW_REFERENCE_EXT] = 3;
    cmp_array[ERL_FUN_EXT]           = 4;
    cmp_array[ERL_NEW_FUN_EXT]       = 4;
    cmp_array[ERL_PORT_EXT]          = 5;
    cmp_array[ERL_PID_EXT]           = 6;
    cmp_array[ERL_SMALL_TUPLE_EXT]   = 7;
    cmp_array[ERL_LARGE_TUPLE_EXT]   = 7;
    cmp_array[ERL_NIL_EXT]           = 8;
    cmp_array[ERL_STRING_EXT]        = 9;
    cmp_array[ERL_LIST_EXT]          = 9;
    cmp_array[ERL_BINARY_EXT]        = 10;
    init_cmp_array_p = 0;
  }
  if (init_cmp_num_class_p) {
    memset(cmp_num_class, 0, CMP_NUM_CLASS_SIZE);
    cmp_num_class[ERL_SMALL_INTEGER_EXT] = SMALL;
    cmp_num_class[ERL_INTEGER_EXT]       = SMALL;
    cmp_num_class[ERL_FLOAT_EXT]         = FLOAT;
    cmp_num_class[ERL_SMALL_BIG_EXT]     = BIG;
    cmp_num_class[ERL_LARGE_BIG_EXT]     = BIG;
    init_cmp_num_class_p = 0;
  }
}

/*==============================================================
 * Marshalling routines.
 *==============================================================
 */

/* 
 * The actual ENCODE engine.
 * Returns 0 on success, otherwise 1.
 */
int erl_encode_it(ETERM *ep, unsigned char **ext, int dist)
{
    int i;
    unsigned int u;
    
    switch(ERL_TYPE(ep)) 
    {
    case ERL_ATOM:
	i =  ep->uval.aval.len;
	*(*ext)++ = ERL_ATOM_EXT;
	*(*ext)++ = (i >>8) &0xff;
	*(*ext)++ = i &0xff;
	memcpy((void *) *ext, (const void *) ep->uval.aval.a, i);
	*ext += i;
	return 0;

    case ERL_INTEGER:
	i = ep->uval.ival.i;
	/* ERL_SMALL_BIG */
	if ((i > ERL_MAX) || (i < ERL_MIN)) { 
	    *(*ext)++ = ERL_SMALL_BIG_EXT;
	    *(*ext)++ = 4;		/* four bytes */
	    if ((*(*ext)++ = ((i>>31) & 0x01))) /* sign byte  */ 
	      i = -i;
	    *(*ext)++ = i  & 0xff;	/* LSB first  */
	    *(*ext)++ = (i >> 8) & 0xff;
	    *(*ext)++ = (i >> 16) & 0xff;
	    *(*ext)++ = (i >> 24) & 0x7f; /* Don't include the sign bit */
	    return 0;
	} 
	/* SMALL_INTEGER */
	if ((i < 256) && (i >= 0)) {
	    *(*ext)++ = ERL_SMALL_INTEGER_EXT;
	    *(*ext)++ = i & 0xff;
	    return 0;
	}
	/* INTEGER */
	*(*ext)++ = ERL_INTEGER_EXT;
	*(*ext)++ = (i >> 24) & 0xff;
	*(*ext)++ = (i >> 16) & 0xff;
	*(*ext)++ = (i >> 8) & 0xff;
	*(*ext)++ = i  & 0xff;
	return 0;

    case ERL_U_INTEGER:
	u = ep->uval.uival.u;
	/* ERL_U_SMALL_BIG */
	if (u > ERL_MAX) {
	*(*ext)++ = ERL_SMALL_BIG_EXT;
	*(*ext)++ = 4;		/* four bytes */
	*(*ext)++ = 0;		/* sign byte  */ 
	*(*ext)++ = u  & 0xff;	/* LSB first  */
	*(*ext)++ = (u >> 8) & 0xff;
	*(*ext)++ = (u >> 16) & 0xff;
	*(*ext)++ = (u >> 24) & 0xff; 
	return 0;
	}
	/* SMALL_INTEGER */
	if ((u < 256) && (u >= 0)) {
	    *(*ext)++ = ERL_SMALL_INTEGER_EXT;
	    *(*ext)++ = u & 0xff;
	    return 0;
	}
	/* INTEGER */
	*(*ext)++ = ERL_INTEGER_EXT;
	*(*ext)++ = (u >> 24) & 0xff;
	*(*ext)++ = (u >> 16) & 0xff;
	*(*ext)++ = (u >> 8) & 0xff;
	*(*ext)++ = u  & 0xff;
	return 0;

    case ERL_PID:
	*(*ext)++ = ERL_PID_EXT;    
	/* First poke in node as an atom */    
	i = strlen(ERL_PID_NODE(ep));
	*(*ext)++ = ERL_ATOM_EXT;
	*(*ext)++ = (i >>8) &0xff;
	*(*ext)++ = i &0xff;
	memcpy(*ext, ERL_PID_NODE(ep), i);
	*ext += i;
	/* And then fill in the integer fields */
	i = ERL_PID_NUMBER(ep);
	*(*ext)++ = (i >>24) &0xff;
	*(*ext)++ = (i >>16) &0xff;
	*(*ext)++ = (i >>8) &0xff;
	*(*ext)++ = i &0xff;
	i = ERL_PID_SERIAL(ep);
	*(*ext)++ = (i >>24) &0xff;
	*(*ext)++ = (i >>16) &0xff;
	*(*ext)++ = (i >>8) &0xff;
	*(*ext)++ = i &0xff;
	*(*ext)++ = ERL_PID_CREATION(ep);
	return 0;
    case ERL_REF: {
	    int len, j;

	    /* Always encode as an extended reference; all
	       participating parties are now expected to be
	       able to decode extended references. */

	    *(*ext)++ = ERL_NEW_REFERENCE_EXT;

	    i = strlen(ERL_REF_NODE(ep));
	    len = ERL_REF_LEN(ep);
	    *(*ext)++ = (len >>8) &0xff;
	    *(*ext)++ = len &0xff;

	    *(*ext)++ = ERL_ATOM_EXT;
	    *(*ext)++ = (i >>8) &0xff;
	    *(*ext)++ = i &0xff;
	    memcpy(*ext, ERL_REF_NODE(ep), i);
	    *ext += i;
	    *(*ext)++ = ERL_REF_CREATION(ep);
	    /* Then the integer fields */
	    for (j = 0; j < ERL_REF_LEN(ep); j++) {
		i = ERL_REF_NUMBERS(ep)[j];
		*(*ext)++ = (i >>24) &0xff;
		*(*ext)++ = (i >>16) &0xff;
		*(*ext)++ = (i >>8) &0xff;
		*(*ext)++ = i &0xff;
	    }
	}
	return 0;
    case ERL_PORT:
	*(*ext)++ = ERL_PORT_EXT;
	/* First poke in node as an atom */
	i = strlen(ERL_PORT_NODE(ep));
	*(*ext)++ = ERL_ATOM_EXT;
	*(*ext)++ = (i >>8) &0xff;
	*(*ext)++ = i &0xff;
	memcpy(*ext, ERL_PORT_NODE(ep), i);
	*ext += i;
	/* Then the integer fields */
	i = ERL_PORT_NUMBER(ep);
	*(*ext)++ = (i >>24) &0xff;
	*(*ext)++ = (i >>16) &0xff;
	*(*ext)++ = (i >>8) &0xff;
	*(*ext)++ = i &0xff;
	*(*ext)++ = ERL_PORT_CREATION(ep);
	return 0;
    case ERL_EMPTY_LIST:
	*(*ext)++ = ERL_NIL_EXT;
	break;
    case ERL_LIST:
	i = is_string(ep);
	if (0 < i && i < 0x10000) { /* String. */
	    *(*ext)++ = ERL_STRING_EXT;
	    *(*ext)++ = (i >>8) &0xff;
	    *(*ext)++ = i &0xff;
	    while (ERL_TYPE(ep) == ERL_LIST) {
		*(*ext)++ = HEAD(ep)->uval.ival.i;
		ep = TAIL(ep);
	    }
	    break;
	} else {		/* List. */
	    i = erl_length(ep);
	    *(*ext)++ = ERL_LIST_EXT;
	    *(*ext)++ = (i >>24) &0xff;
	    *(*ext)++ = (i >>16) &0xff;
	    *(*ext)++ = (i >>8) &0xff;
	    *(*ext)++ = i &0xff;
	    while (ERL_TYPE(ep) == ERL_LIST) {
		if (erl_encode_it(HEAD(ep), ext, dist))
		    return 1;
		ep = TAIL(ep);
	    }
	    i = erl_encode_it(ep, ext, dist);
	    return i;
	}
    case ERL_TUPLE:
	i = ep->uval.tval.size;
	if (i <= 0xff) {
	    *(*ext)++ = ERL_SMALL_TUPLE_EXT;
	    *(*ext)++ = i & 0xff;
	}
	else {
	    *(*ext)++ = ERL_LARGE_TUPLE_EXT;
	    *(*ext)++ = (i >> 24) & 0xff;
	    *(*ext)++ = (i >> 16 ) & 0xff;
	    *(*ext)++ = (i >> 8) & 0xff;
	    *(*ext)++ = i & 0xff;
	}
	for (i=0; i<ep->uval.tval.size; i++)
	    if (erl_encode_it(ep->uval.tval.elems[i], ext, dist))
		return 1;
	break;
    case ERL_FLOAT:
	*(*ext)++ = ERL_FLOAT_EXT;
	memset(*ext, 0, 31);
	sprintf((char *) *ext, "%.20e", ep->uval.fval.f);
	*ext += 31;
	break;
    case ERL_BINARY:
	*(*ext)++ = ERL_BINARY_EXT;
	i = ep->uval.bval.size;
	*(*ext)++ = (i >> 24) & 0xff;
	*(*ext)++ = (i >> 16) & 0xff;
	*(*ext)++ = (i >> 8) & 0xff;
	*(*ext)++ = i  & 0xff;
	memcpy((char *) *ext, (char*) ep->uval.bval.b, i);
	*ext += i;
	break;
    case ERL_FUNCTION:
	if (ERL_FUN_ARITY(ep) != -1) {
	    unsigned char *size_p = *ext + 1;
	    *(*ext)++ = ERL_NEW_FUN_EXT;
	    *ext += 4;
	    i = ERL_FUN_ARITY(ep);
	    put8(*ext, i);
	    memcpy(*ext, ERL_FUN_MD5(ep), 16);
	    *ext += 16;
	    i = ERL_FUN_NEW_INDEX(ep);
	    put32be(*ext, i);
	    i = ERL_CLOSURE_SIZE(ep);
	    put32be(*ext, i);
	    erl_encode_it(ERL_FUN_MODULE(ep), ext, dist);
	    erl_encode_it(ERL_FUN_INDEX(ep), ext, dist);
	    erl_encode_it(ERL_FUN_UNIQ(ep), ext, dist);
	    erl_encode_it(ERL_FUN_CREATOR(ep), ext, dist);
	    for (i = 0; i < ERL_CLOSURE_SIZE(ep); i++)
		erl_encode_it(ep->uval.funcval.closure[i], ext, dist);
	    if (size_p != NULL) {
		i = *ext - size_p;
		put32be(size_p, i);
	    }
	} else {
	    *(*ext)++ = ERL_FUN_EXT;
	    i = ERL_CLOSURE_SIZE(ep);
	    *(*ext)++ = (i >> 24) & 0xff;
	    *(*ext)++ = (i >> 16) & 0xff;
	    *(*ext)++ = (i >> 8) & 0xff;
	    *(*ext)++ = i  & 0xff;
	    erl_encode_it(ERL_FUN_CREATOR(ep), ext, dist);
	    erl_encode_it(ERL_FUN_MODULE(ep), ext, dist);
	    erl_encode_it(ERL_FUN_INDEX(ep), ext, dist);
	    erl_encode_it(ERL_FUN_UNIQ(ep), ext, dist);
	    for (i = 0; i < ERL_CLOSURE_SIZE(ep); i++)
		erl_encode_it(ep->uval.funcval.closure[i], ext, dist);
	}
	break;
    default:
	return 1;
    }
    return 0;
}

/* 
 * ENCODE an ETERM into a BUFFER, assuming BUFFER is of 
 * enough size. At success return number of bytes written 
 * into it, otherwise return 0.
 */
static int erl_encode3(ETERM *ep, unsigned char *t, int dist)
{
  unsigned char *x = t;
  
  *x++ = ERL_VERSION_MAGIC;
  if (erl_encode_it(ep, &x, dist)) {
#ifdef DEBUG
    erl_err_msg("<ERROR> erl_encode: Error while encoding");
#endif
    return 0;
  }
  return (x - t);

}

/* API */

int erl_encode(ETERM *ep, unsigned char *t)
{
    return erl_encode3(ep, t, 4);
}

/* determine the buffer size that will be required for the eterm */
static int erl_term_len_helper(ETERM *ep, int dist);

/* FIXME hard coded dist version */
int erl_term_len(ETERM *ep)
{
  return 1+erl_term_len_helper(ep, 4);
}

static int erl_term_len_helper(ETERM *ep, int dist)
{
  int len = 0;
  int i;
  unsigned int u;

  if (ep) {
    switch (ERL_TYPE(ep)) {
    case ERL_ATOM:
      i = ep->uval.aval.len;
      len = i + 3;
      break;

    case ERL_INTEGER:
      i = ep->uval.ival.i;
      if ((i > ERL_MAX) || (i < ERL_MIN)) len = 7;
      else if ((i < 256) && (i >= 0)) len = 2; 
      else len = 5;
      break;

    case ERL_U_INTEGER:
      u = ep->uval.uival.u;
      if (u > ERL_MAX) len = 7;
      else if (u  < 256) len = 2;
      else len = 5;
      break;

    case ERL_PID:
      /* 1 + N + 4 + 4 + 1 where N = 3 + strlen */
      i = strlen(ERL_PID_NODE(ep));
      len = 13 + i;
      break;

    case ERL_REF:
      i = strlen(ERL_REF_NODE(ep));
      if (dist >= 4 && ERL_REF_LEN(ep) > 1) {
	  len = 1 + 2 + (i+3) + 1 + ERL_REF_LEN(ep) * 4;
      } else {
	  /* 1 + N + 4 + 1 where N = 3 + strlen */
	  len = 9 + i;
      }
      break;

    case ERL_PORT:
      /* 1 + N + 4 + 1 where N = 3 + strlen */
      i = strlen(ERL_PORT_NODE(ep));
      len = 9 + i;
      break;

    case ERL_EMPTY_LIST:
      len = 1;
      break;

    case ERL_LIST:
      i = is_string(ep);
      if ((i > 0) && (i < 0x10000)) { /* string: 3 + strlen */
	for (len = 3; ERL_TYPE(ep) == ERL_LIST; ep =  TAIL(ep)) {
	  len++;
	}