clisp-cvs digest, Vol 1 #935 - 4 msgs

[<cli...@li...>]

Send clisp-cvs mailing list submissions to
	cli...@li...

To subscribe or unsubscribe via the World Wide Web, visit
	https://lists.sourceforge.net/lists/listinfo/clisp-cvs
or, via email, send a message with subject or body 'help' to
	cli...@li...

You can reach the person managing the list at
	cli...@li...

When replying, please edit your Subject line so it is more specific
than "Re: Contents of clisp-cvs digest..."


CLISP CVS commits for today

Today's Topics:

   1. clisp/src io.d,1.281,1.282 ChangeLog,1.4169,1.4170 (Bruno Haible)
   2. clisp/src intserial.d,NONE,1.1 lisparit.d,1.74,1.75 makemake.in,1.510,1.511 stream.d,1.503,1.504 lispbibl.d,1.609,1.610 genclisph.d,1.167,1.168 ChangeLog,1.4170,1.4171 (Bruno Haible)
   3. clisp/src/po Makefile.devel,1.37,1.38 (Bruno Haible)
   4. clisp/src lispbibl.d,1.610,1.611 genclisph.d,1.168,1.169 intelem.d,1.29,1.30 socket.d,1.89,1.90 ChangeLog,1.4171,1.4172 (Bruno Haible)

--__--__--

Message: 1
From: Bruno Haible <ha...@us...>
To: cli...@li...
Subject: clisp/src io.d,1.281,1.282 ChangeLog,1.4169,1.4170
Date: Sat, 29 Jan 2005 14:50:30 +0000
Reply-To: cli...@li...

Update of /cvsroot/clisp/clisp/src
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv23339/src

Modified Files:
	io.d ChangeLog 
Log Message:
Fix compilation in DEBUG_GCSAFETY mode.


Index: io.d
===================================================================
RCS file: /cvsroot/clisp/clisp/src/io.d,v
retrieving revision 1.281
retrieving revision 1.282
diff -u -d -r1.281 -r1.282
--- io.d	26 Jan 2005 13:25:29 -0000	1.281
+++ io.d	29 Jan 2005 14:50:06 -0000	1.282
@@ -8452,7 +8452,7 @@
   JUSTIFY_SPACE;
   JUSTIFY_LAST(true);
   var object n = (orecordp(obj) # BigReadLabel or Small-Read-Label?
-                  ? TheBigReadLabel(obj)->brl_value
+                  ? (object)TheBigReadLabel(obj)->brl_value
                   : small_read_label_value(obj));
   print_integer(n,10,stream_); # print n in decimal
   JUSTIFY_END_FILL;

Index: ChangeLog
===================================================================
RCS file: /cvsroot/clisp/clisp/src/ChangeLog,v
retrieving revision 1.4169
retrieving revision 1.4170
diff -u -d -r1.4169 -r1.4170
--- ChangeLog	28 Jan 2005 17:22:43 -0000	1.4169
+++ ChangeLog	29 Jan 2005 14:50:09 -0000	1.4170
@@ -1,5 +1,9 @@
 2005-01-28  Bruno Haible  <br...@cl...>
 
+	* io.d (pr_readlabel): Fix C++ compilation.
+
+2005-01-28  Bruno Haible  <br...@cl...>
+
 	* makemake.in (check-fresh-line): Reduce the scope of the test on
 	Woe32.
 



--__--__--

Message: 2
From: Bruno Haible <ha...@us...>
To: cli...@li...
Subject: clisp/src intserial.d,NONE,1.1 lisparit.d,1.74,1.75 makemake.in,1.510,1.511 stream.d,1.503,1.504 lispbibl.d,1.609,1.610 genclisph.d,1.167,1.168 ChangeLog,1.4170,1.4171
Date: Sat, 29 Jan 2005 14:55:13 +0000
Reply-To: cli...@li...

Update of /cvsroot/clisp/clisp/src
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv24689/src

Modified Files:
	lisparit.d makemake.in stream.d lispbibl.d genclisph.d 
	ChangeLog 
Added Files:
	intserial.d 
Log Message:
New primitives for serializing and deserializing integers.


Index: lisparit.d
===================================================================
RCS file: /cvsroot/clisp/clisp/src/lisparit.d,v
retrieving revision 1.74
retrieving revision 1.75
diff -u -d -r1.74 -r1.75
--- lisparit.d	5 Jan 2005 12:12:30 -0000	1.74
+++ lisparit.d	29 Jan 2005 14:55:06 -0000	1.75
@@ -39,6 +39,7 @@
 #include "intsqrt.c"  /* Root, ISQRT */
 #include "intprint.c" /* auxiliary function for output of integers */
 #include "intread.c"  /* auxiliary function for input  of integers */
+#include "intserial.c" /* serialization of integers */
 /* for rational numbers: */
 #include "rational.c" /* rational numbers */
 /* for floats: */

Index: genclisph.d
===================================================================
RCS file: /cvsroot/clisp/clisp/src/genclisph.d,v
retrieving revision 1.167
retrieving revision 1.168
diff -u -d -r1.167 -r1.168
--- genclisph.d	25 Jan 2005 14:35:09 -0000	1.167
+++ genclisph.d	29 Jan 2005 14:55:10 -0000	1.168
@@ -1,6 +1,6 @@
 /*
  * Export CLISP internals for modules
- * Bruno Haible 1994-2004
+ * Bruno Haible 1994-2005
  * Sam Steingold 1998-2005
  */
 
@@ -2221,6 +2221,10 @@
   printf("extern object I_I_plus_I (object x, object y);\n");
   printf("extern object I_I_minus_I (object x, object y);\n");
 #endif
+  printf("extern object LEbytes_to_UI (uintL bytesize, const uintB* bufferptr);\n");
+  printf("extern object LEbytes_to_I (uintL bytesize, const uintB* bufferptr);\n");
+  printf("extern bool UI_to_LEbytes (object obj, uintL bitsize, uintB* bufferptr);\n");
+  printf("extern bool I_to_LEbytes (object obj, uintL bitsize, uintB* bufferptr);\n");
   printf("extern object c_float_to_FF (const ffloatjanus* val_);\n");
   printf("extern void FF_to_c_float (object obj, ffloatjanus* val_);\n");
   printf("extern object c_double_to_DF (const dfloatjanus* val_);\n");

Index: stream.d
===================================================================
RCS file: /cvsroot/clisp/clisp/src/stream.d,v
retrieving revision 1.503
retrieving revision 1.504
diff -u -d -r1.503 -r1.504
--- stream.d	28 Jan 2005 15:07:43 -0000	1.503
+++ stream.d	29 Jan 2005 14:55:07 -0000	1.504
@@ -4406,65 +4406,12 @@
 # ------------------------------
 
 local maygc object bitbuff_iu_I (object bitbuffer, uintL bitsize, uintL bytesize) {
-  # normalize number in bitbuffer:
   var uintB* bitbufferptr = &TheSbvector(bitbuffer)->data[bytesize-1];
   *bitbufferptr &= (bit(((bitsize-1)%8)+1)-1); # mask High byte
-  var uintL count = bytesize;
-  while ((!(count==0)) && (*bitbufferptr==0)) { count--; bitbufferptr--; }
-  # make number:
-  if # at most oint_data_len Bits ?
-    ((count <= floor(oint_data_len,8))
-     || ((count == floor(oint_data_len,8)+1)
-         && (*bitbufferptr < bit(oint_data_len%8)))) {
-    # yes -> build Fixnum >=0 :
-    var uintL wert = 0;
-    until (count==0) { wert = (wert<<8) | *bitbufferptr--; count--; }
-    return fixnum(wert);
-  }
-  # no -> build Bignum >0 :
   pushSTACK(bitbuffer);
-  var uintL digitcount = floor(count,(intDsize/8));
-  if (((count%(intDsize/8)) > 0) || (*bitbufferptr & bit(7)))
-    digitcount++;
-  # As bitsize < intDsize*uintWC_max,
-  # digitcount <= ceiling((bitsize+1)/intDsize) <= uintWC_max .
-  var object big = allocate_bignum(digitcount,0); # new Bignum >0
-  TheBignum(big)->data[0] = 0; # set highest Digit to 0
-  # fill remaining Digits from right to left,
-  # thereby translate sequence of Bytes into sequence of uintD:
-  bitbuffer = popSTACK();
-  bitbufferptr = &TheSbvector(bitbuffer)->data[0];
-  #if BIG_ENDIAN_P
-  {
-    var uintB* bigptr = (uintB*)(&TheBignum(big)->data[digitcount]);
-    dotimespL(count,count, { *--bigptr = *bitbufferptr++; } );
-  }
-  #else
-  {
-    var uintD* bigptr = &TheBignum(big)->data[digitcount];
-    var uintL count2;
-#define GET_NEXT_BYTE(i)  digit |= ((uintD)(*bitbufferptr++) << (8*i));
-    dotimespL(count2,floor(count,intDsize/8), {
-      var uintD digit = 0;
-      DOCONSTTIMES(intDsize/8,GET_NEXT_BYTE); # GET_NEXT_BYTE(0..intDsize/8-1)
-      *--bigptr = digit;
-    });
-#undef GET_NEXT_BYTE
-    count2 = count % (intDsize/8);
-    if (count2>0) {
-      var uintL shiftcount = 0;
-      var uintD digit = (uintD)(*bitbufferptr++);
-      dotimesL(count2,count2-1, {
-        shiftcount += 8;
-        digit |= ((uintD)(*bitbufferptr++) << shiftcount);
-      });
-      *--bigptr = digit;
-    }
-  }
-  #endif
-  # since (intDsize/8)*(digitcount-1) <= count <= (intDsize/8)*digitcount
-  # everything is filled.
-  return big;
+  var object result = LESbvector_to_UI(bytesize,&STACK_0);
+  skipSTACK(1);
+  return result;
 }
 
 # UP for READ-BYTE on File-Streams of Integers, Type u :
@@ -4475,90 +4422,17 @@
 }
 
 local maygc object bitbuff_is_I (object bitbuffer, uintL bitsize, uintL bytesize) {
-  # normalize number in bitbuffer:
   var uintB* bitbufferptr = &TheSbvector(bitbuffer)->data[bytesize-1];
-  var sintD sign;
   var uintL signbitnr = (bitsize-1)%8;
-  var uintL count = bytesize;
   if (!(*bitbufferptr & bit(signbitnr))) {
-    sign = 0;
     *bitbufferptr &= (bitm(signbitnr+1)-1); # sign-extend High byte
-    # normalize, highest Bit must remain 0:
-    while ((count>=2) && (*bitbufferptr==0)
-           && !(*(bitbufferptr-1) & bit(7))) {
-      count--; bitbufferptr--;
-    }
-    # build number:
-    if # at most oint_data_len+1 Bits, count <2^oint_data_len ?
-      ((count <= floor(oint_data_len,8))
-       || ((count == floor(oint_data_len,8)+1)
-           && (*bitbufferptr < bit(oint_data_len%8)))) {
-      # yes -> build Fixnum >=0:
-      var uintL wert = 0;
-      until (count==0) { wert = (wert<<8) | *bitbufferptr--; count--; }
-      return posfixnum(wert);
-    }
   } else {
-    sign = -1;
     *bitbufferptr |= minus_bitm(signbitnr+1); # sign-extend High byte
-    # normalize, highest Bit must remain 1:
-    while ((count>=2) && (*bitbufferptr==(uintB)(-1))
-           && (*(bitbufferptr-1) & bit(7))) {
-      count--; bitbufferptr--;
-    }
-    # make number:
-    if # at most oint_data_len+1 Bits, count >=-2^oint_data_len ?
-      ((count <= floor(oint_data_len,8))
-       || ((count == floor(oint_data_len,8)+1)
-           && (*bitbufferptr >= (uintB)(-bit(oint_data_len%8))))) {
-      # yes -> build Fixnum <0:
-      var uintL wert = (uintL)(-1);
-      until (count==0) { wert = (wert<<8) | *bitbufferptr--; count--; }
-      return negfixnum(-wbitm(intLsize)+(oint)wert);
-    }
   }
-  # make bignum:
   pushSTACK(bitbuffer);
-  var uintL digitcount = ceiling(count,(intDsize/8));
-  # As bitsize < intDsize*uintWC_max,
-  # digitcount <= ceiling(bitsize/intDsize) <= uintWC_max .
-  var object big = allocate_bignum(digitcount,(sintB)sign);
-  TheBignum(big)->data[0] = sign; # set highest Word to sign
-  # fill the remaining Digits from right to left,
-  # thereby translate sequence of Bytes into sequence of uintD:
-  bitbuffer = popSTACK();
-  bitbufferptr = &TheSbvector(bitbuffer)->data[0];
-  #if BIG_ENDIAN_P
-  {
-    var uintB* bigptr = (uintB*)(TheBignum(big)->data+digitcount);
-    dotimespL(count,count, { *--bigptr = *bitbufferptr++; } );
-  }
-  #else
-  {
-    var uintD* bigptr = TheBignum(big)->data+digitcount;
-    var uintL count2;
-#define GET_NEXT_BYTE(i) digit |= ((uintD)(*bitbufferptr++) << (8*i));
-    dotimespL(count2,floor(count,intDsize/8), {
-      var uintD digit = 0;
-      DOCONSTTIMES(intDsize/8,GET_NEXT_BYTE); # GET_NEXT_BYTE(0..intDsize/8-1)
-      *--bigptr = digit;
-    });
-#undef GET_NEXT_BYTE
-    count2 = count % (intDsize/8);
-    if (count2>0) {
-      var uintL shiftcount = 0;
-      var uintD digit = (uintD)(*bitbufferptr++);
-      dotimesL(count2,count2-1, {
-        shiftcount += 8;
-        digit |= ((uintD)(*bitbufferptr++) << shiftcount);
-      });
-      *--bigptr = digit ^ (sign << (shiftcount+8));
-    }
-  }
-  #endif
-  # since (intDsize/8)*(digitcount-1) < count <= (intDsize/8)*digitcount
-  # everything is filled.
-  return big;
+  var object result = LESbvector_to_I(bytesize,&STACK_0);
+  skipSTACK(1);
+  return result;
 }
 
 # UP for READ-BYTE on File-Streams of Integers, Type s :
@@ -4578,69 +4452,10 @@
 # stream.
 typedef void wr_by_aux_ix (object stream, uintL bitsize, uintL bytesize);
 
-local void bitbuff_ixu_sub (object stream, object bitbuffer,
-                            uintL bitsize, uintL bytesize, object obj) {
-  ASSERT_wr_int(stream,obj);
-  if (!positivep(obj))
+local inline void bitbuff_ixu_sub (object stream, object bitbuffer,
+                                   uintL bitsize, object obj) {
+  if (UI_to_LEbytes(obj,bitsize,TheSbvector(bitbuffer)->data))
     fehler_bad_integer(stream,obj);
-  # obj is an integer >=0
-  # transfer obj into the bitbuffer:
-  {
-    var uintB* bitbufferptr = TheSbvector(bitbuffer)->data;
-    var uintL count = bytesize;
-    if (posfixnump(obj)) { # obj is a Fixnum >=0
-      var uintL wert = posfixnum_to_L(obj);
-      # check wert < 2^bitsize:
-      if (!((bitsize>=oint_data_len) || (wert < bit(bitsize))))
-        fehler_bad_integer(stream,obj);
-      # store wert in Bitbuffer:
-      until (wert==0) {
-        *bitbufferptr++ = (uint8)wert; wert = wert>>8; count--;
-      }
-    } else { # obj is a Bignum >0
-      var uintL len = (uintL)Bignum_length(obj);
-      # check obj < 2^bitsize:
-      if (!((floor(bitsize,intDsize) >= len)
-            || ((floor(bitsize,intDsize) == len-1)
-                && (TheBignum(obj)->data[0] < bit(bitsize%intDsize)))))
-        fehler_bad_integer(stream,obj);
-      #if BIG_ENDIAN_P
-      {
-        var uintB* ptr = (uintB*)&TheBignum(obj)->data[len];
-        # convert Digit-Length in Byte-Length:
-        len = (intDsize/8)*len;
-#define CHECK_NEXT_BYTE(i)  \
-   if (((uintB*)(&TheBignum(obj)->data[0]))[i] != 0) goto len_ok; len--;
-        DOCONSTTIMES(intDsize/8,CHECK_NEXT_BYTE); # CHECK_NEXT_BYTE(0..intDsize/8-1)
-#undef CHECK_NEXT_BYTE
-      len_ok:
-        # store obj in Bitbuffer:
-        count = count - len;
-        dotimespL(len,len, { *bitbufferptr++ = *--ptr; } );
-      }
-      #else
-      {
-        var uintD* ptr = &TheBignum(obj)->data[len];
-        len--;
-        count -= (intDsize/8)*len;
-        dotimesL(len,len, {
-          var uintD digit = *--ptr;
-          doconsttimes(intDsize/8, {
-            *bitbufferptr++ = (uintB)digit; digit = digit >> 8;
-          });
-        });
-        var uintD digit = *--ptr;
-        doconsttimes(intDsize/8, {
-          if (digit==0) goto ok;
-          *bitbufferptr++ = (uintB)digit; digit = digit >> 8;
-          count--;
-        });
-      ok: ;
-      }
-    #endif
-    }
-    memset(bitbufferptr,0,count);
-  }
 }
 
 # UP for WRITE-BYTE on File-Streams of Integers, Type u :
@@ -4651,79 +4466,15 @@
 local maygc void wr_by_ixu_sub (object stream, object obj, wr_by_aux_ix* finisher) {
   var uintL bitsize = ChannelStream_bitsize(stream);
   var uintL bytesize = ceiling(bitsize,8);
-  bitbuff_ixu_sub(stream,TheStream(stream)->strm_bitbuffer,
-                  bitsize,bytesize,obj);
+  ASSERT_wr_int(stream,obj);
+  bitbuff_ixu_sub(stream,TheStream(stream)->strm_bitbuffer,bitsize,obj);
   (*finisher)(stream,bitsize,bytesize);
 }
 
-local void bitbuff_ixs_sub (object stream, object bitbuffer,
-                            uintL bitsize, uintL bytesize, object obj) {
-  ASSERT_wr_int(stream,obj);
-  # obj is an integer
-  # transfer obj into the bitbuffer:
-  {
-    var uintB* bitbufferptr = TheSbvector(bitbuffer)->data;
-    var uintL count = bytesize;
-    var uintL sign = (sintL)R_sign(obj);
-    if (fixnump(obj)) {
-      # obj is a Fixnum
-      var uintL wert = fixnum_to_L(obj); # >=0 or <0, according to sign
-      # check 0 <= wert < 2^(bitsize-1) resp. -2^(bitsize-1) <= wert < 0:
-      wert = wert^sign;
-      if (!((bitsize>oint_data_len) || (wert < bit(bitsize-1))))
-        fehler_bad_integer(stream,obj);
-      # store wert^sign in Bitbuffer:
-      until (wert == 0) {
-        *bitbufferptr++ = (uint8)(wert^sign); wert = wert>>8; count--;
-      }
-      memset(bitbufferptr,(uint8)sign,count);
-    } else {
-      # obj is a Bignum
-      var uintL len = (uintL)Bignum_length(obj);
-      # check -2^(bitsize-1) <= obj < 2^(bitsize-1):
-      if (!((floor(bitsize,intDsize) >= len)
-            || ((bitsize > intDsize*(len-1))
-                && ((TheBignum(obj)->data[0] ^ (uintD)sign) <
-                    bit((bitsize%intDsize)-1)))))
-        fehler_bad_integer(stream,obj);
-      #if BIG_ENDIAN_P
-      {
-        var uintB* ptr = (uintB*)&TheBignum(obj)->data[len];
-        # convert Digit-Length in Byte-Length:
-        len = (intDsize/8)*len;
-#define CHECK_NEXT_BYTE(i)  \
-   if (((uintB*)(&TheBignum(obj)->data[0]))[i] != (uintB)sign) goto len_ok; \
-   len--;
-        DOCONSTTIMES(intDsize/8,CHECK_NEXT_BYTE); # CHECK_NEXT_BYTE(0..intDsize/8-1)
-#undef CHECK_NEXT_BYTE
-      len_ok:
-        # store obj in Bitbuffer:
-        count = count - len;
-        dotimespL(len,len, { *bitbufferptr++ = *--ptr; } );
-      }
-      #else
-      {
-        var uintD* ptr = &TheBignum(obj)->data[len];
-        len--;
-        count -= (intDsize/8)*len;
-        dotimesL(len,len, {
-          var uintD digit = *--ptr;
-          doconsttimes(intDsize/8, {
-            *bitbufferptr++ = (uintB)digit; digit = digit >> 8;
-          });
-        });
-        var sintD digit = *--ptr;
-        doconsttimes(intDsize/8, {
-          if (digit == (sintD)sign) goto ok;
-          *bitbufferptr++ = (uintB)digit; digit = digit >> 8;
-          count--;
-        });
-      ok: ;
-      }
-      #endif
-      memset(bitbufferptr,(uintB)sign,count);
-    }
-  }
+local inline void bitbuff_ixs_sub (object stream, object bitbuffer,
+                                   uintL bitsize, object obj) {
+  if (I_to_LEbytes(obj,bitsize,TheSbvector(bitbuffer)->data))
+    fehler_bad_integer(stream,obj);
 }
 
 # UP for WRITE-BYTE on File-Streams of Integers, Type s :
@@ -4734,8 +4485,8 @@
 local maygc void wr_by_ixs_sub (object stream, object obj, wr_by_aux_ix* finisher) {
   var uintL bitsize = ChannelStream_bitsize(stream);
   var uintL bytesize = ceiling(bitsize,8);
-  bitbuff_ixs_sub(stream,TheStream(stream)->strm_bitbuffer,
-                  bitsize,bytesize,obj);
+  ASSERT_wr_int(stream,obj);
+  bitbuff_ixs_sub(stream,TheStream(stream)->strm_bitbuffer,bitsize,obj);
   (*finisher)(stream,bitsize,bytesize);
 }
 
@@ -17087,10 +16838,10 @@
   # Copy the integer's data into the buffer.
   switch (eltype.kind) {
     case eltype_iu:
-      bitbuff_ixu_sub(STACK_3,bitbuffer,bitsize,bytesize,obj);
+      bitbuff_ixu_sub(STACK_3,bitbuffer,bitsize,obj);
       break;
     case eltype_is:
-      bitbuff_ixs_sub(STACK_3,bitbuffer,bitsize,bytesize,obj);
+      bitbuff_ixs_sub(STACK_3,bitbuffer,bitsize,obj);
       break;
     default: NOTREACHED;
   }

Index: ChangeLog
===================================================================
RCS file: /cvsroot/clisp/clisp/src/ChangeLog,v
retrieving revision 1.4170
retrieving revision 1.4171
diff -u -d -r1.4170 -r1.4171
--- ChangeLog	29 Jan 2005 14:50:09 -0000	1.4170
+++ ChangeLog	29 Jan 2005 14:55:10 -0000	1.4171
@@ -1,5 +1,26 @@
 2005-01-28  Bruno Haible  <br...@cl...>
 
+	* lispbibl.d (FAKE_8BIT_VECTOR): New macro.
+	(LEbytes_to_UI, LESbvector_to_UI, LEbytes_to_I, LESbvector_to_I,
+	UI_to_LEbytes, I_to_LEbytes): New declarations.
+	* intserial.d: New file, mostly extracted from stream.d.
+	* lisparit.d: Include intserial.c.
+	* stream.d (bitbuff_iu_I): Use LESbvector_to_UI.
+	(bitbuff_is_I): Use LESbvector_to_I.
+	(bitbuff_ixu_sub): Remove bytesize argument. Use UI_to_LEbytes. Don't
+	call ASSERT_wr_int here.
+	(wr_by_ixu_sub): Call ASSERT_wr_int here. Update.
+	(bitbuff_ixs_sub): Remove bytesize argument. Use I_to_LEbytes. Don't
+	call ASSERT_wr_int here.
+	(wr_by_ixs_sub): Call ASSERT_wr_int here. Update.
+	(WRITE-INTEGER): Update.
+	* genclisph.d (main): Emit declarations for LEbytes_to_UI, LEbytes_to_I,
+	UI_to_LEbytes, I_to_LEbytes.
+	* makemake.in (LISPARIT_SUBFILES): Add intserial.
+	* po/Makefile.devel (DSOURCES): Add intserial.
+
+2005-01-28  Bruno Haible  <br...@cl...>
+
 	* io.d (pr_readlabel): Fix C++ compilation.
 
 2005-01-28  Bruno Haible  <br...@cl...>

Index: makemake.in
===================================================================
RCS file: /cvsroot/clisp/clisp/src/makemake.in,v
retrieving revision 1.510
retrieving revision 1.511
diff -u -d -r1.510 -r1.511
--- makemake.in	28 Jan 2005 17:22:41 -0000	1.510
+++ makemake.in	29 Jan 2005 14:55:06 -0000	1.511
@@ -1630,7 +1630,7 @@
   ERROR_INCLUDES=$ERROR_INCLUDES' errwin32'
 fi
 
-LISPARIT_SUBFILES=' aridecl arilev0 arilev1 intelem intlog intplus intcomp intbyte intmal intdiv intgcd int2adic intsqrt intprint intread rational'
+LISPARIT_SUBFILES=' aridecl arilev0 arilev1 intelem intlog intplus intcomp intbyte intmal intdiv intgcd int2adic intsqrt intprint intread intserial rational'
 LISPARIT_SUBFILES=$LISPARIT_SUBFILES' sfloat ffloat dfloat lfloat flo_konv flo_rest realelem realrand realtran compelem comptran'
 LISPARIT_INCLUDES=$LISPARIT_SUBFILES' arilev1c arilev1e arilev1i'
 

Index: lispbibl.d
===================================================================
RCS file: /cvsroot/clisp/clisp/src/lispbibl.d,v
retrieving revision 1.609
retrieving revision 1.610
diff -u -d -r1.609 -r1.610
--- lispbibl.d	28 Jan 2005 14:09:10 -0000	1.609
+++ lispbibl.d	29 Jan 2005 14:55:07 -0000	1.610
@@ -8656,6 +8656,20 @@
 #endif
 # used by STREAM, PATHNAME
 
+# Macro: Wraps a GC-invariant uintB* pointer in a fake simple-8bit-vector.
+# FAKE_8BIT_VECTOR(ptr)
+# > uintB* ptr: pointer to GC-invariant data
+# < gcv_object_t obj: a fake simple-8bit-vector,
+#                     with TheSbvector(obj)->data == ptr,
+#                     that must *not* be stored in GC-visible locations
+#ifdef TYPECODES
+  #define FAKE_8BIT_VECTOR(ptr)  \
+    type_pointer_object(0, (const char*)(ptr) - offsetofa(sbvector_,data))
+#else
+  #define FAKE_8BIT_VECTOR(ptr)  \
+    fake_gcv_object((aint)((const char*)(ptr) - offsetofa(sbvector_,data)) + varobject_bias)
+#endif
+
 #if !defined(UNICODE) || defined(HAVE_SMALL_SSTRING)
 # UP, provides 8-bit character string
 # allocate_s8string(len)
@@ -14479,6 +14493,57 @@
 extern uintL I_integer_length (object x);
 # is used by ARRAY
 
+# Converts a little-endian byte sequence to an unsigned integer.
+# > bytesize: number of given 8-bit bytes of the integer,
+#             < intDsize/8*uintWC_max
+# > bufferptr: address of bytesize bytes of memory
+# < result: an integer >= 0 with I_integer_length(result) <= 8*bytesize
+extern maygc object LEbytes_to_UI (uintL bytesize, const uintB* bufferptr);
+
+# Converts a little-endian byte sequence to an unsigned integer.
+# > bytesize: number of given 8-bit bytes of the integer,
+#             < intDsize/8*uintWC_max
+# > *buffer_: address of a simple-8bit-vector (or of a fake)
+#             containing bytesize bytes of memory
+# < result: an integer >= 0 with I_integer_length(result) <= 8*bytesize
+extern maygc object LESbvector_to_UI (uintL bytesize, const gcv_object_t* buffer_);
+# is used by STREAM
+
+# Converts a little-endian byte sequence to an integer.
+# > bytesize: number of given 8-bit bytes of the integer, > 0,
+#             < intDsize/8*uintWC_max
+# > bufferptr: address of bytesize bytes of memory
+# < result: an integer with I_integer_length(result) < 8*bytesize
+extern maygc object LEbytes_to_I (uintL bytesize, const uintB* bufferptr);
+
+# Converts a little-endian byte sequence to an integer.
+# > bytesize: number of given 8-bit bytes of the integer, > 0,
+#             < intDsize/8*uintWC_max
+# > *buffer_: address of a simple-8bit-vector (or of a fake)
+#             containing bytesize bytes of memory
+# < result: an integer with I_integer_length(result) < 8*bytesize
+extern maygc object LESbvector_to_I (uintL bytesize, const gcv_object_t* buffer_);
+# is used by STREAM
+
+# Converts an unsigned integer to a little-endian byte sequence.
+# > obj: an integer
+# > bitsize: maximum number of bits of the integer
+# > bufferptr: pointer to bytesize = ceiling(bitsize,8) bytes of memory
+# < false and bufferptr[0..bytesize-1] filled, if obj >= 0 and
+#                                              I_integer_length(obj) <= bitsize;
+#   true, if obj is out of range
+extern bool UI_to_LEbytes (object obj, uintL bitsize, uintB* bufferptr);
+# is used by STREAM
+
+# Converts an integer to a little-endian byte sequence.
+# > obj: an integer
+# > bitsize: maximum number of bits of the integer, including the sign bit
+# > bufferptr: pointer to bytesize = ceiling(bitsize,8) bytes of memory
+# < false and bufferptr[0..bytesize-1] filled, if I_integer_length(obj) < bitsize;
+#   true, if obj is out of range
+extern bool I_to_LEbytes (object obj, uintL bitsize, uintB* bufferptr);
+# is used by STREAM
+
 # c_float_to_FF(&val) converts an IEEE-single-float val into an single-float.
 # can trigger GC
 extern maygc object c_float_to_FF (const ffloatjanus* val_);

--- NEW FILE: intserial.d ---
/* Serialization of integers into a little-endian sequence of bytes */

# Converts a little-endian byte sequence to an unsigned integer.
# > bytesize: number of given 8-bit bytes of the integer,
#             < intDsize/8*uintWC_max
# > bufferptr: address of bytesize bytes of memory
# < result: an integer >= 0 with I_integer_length(result) <= 8*bytesize
global maygc object LEbytes_to_UI (uintL bytesize, const uintB* bufferptr) {
  var gcv_object_t fake = FAKE_8BIT_VECTOR(bufferptr);
  return LESbvector_to_UI(bytesize,&fake);
}

# Converts a little-endian byte sequence to an unsigned integer.
# > bytesize: number of given 8-bit bytes of the integer,
#             < intDsize/8*uintWC_max
# > *buffer_: address of a simple-8bit-vector (or of a fake)
#             containing bytesize bytes of memory
# < result: an integer >= 0 with I_integer_length(result) <= 8*bytesize
global maygc object LESbvector_to_UI (uintL bytesize, const gcv_object_t* buffer_) {
  # Normalize number in buffer:
  var uintB* bufferptr = &TheSbvector(*buffer_)->data[bytesize-1];
  var uintL count = bytesize;
  while ((!(count==0)) && (*bufferptr==0)) { count--; bufferptr--; }
  # Make number:
  if # at most oint_data_len Bits ?
    ((count <= floor(oint_data_len,8))
     || ((count == floor(oint_data_len,8)+1)
         && (*bufferptr < bit(oint_data_len%8)))) {
    # yes -> build Fixnum >=0 :
    var uintL wert = 0;
    until (count==0) { wert = (wert<<8) | *bufferptr--; count--; }
    return fixnum(wert);
  }
  # no -> build Bignum >0 :
  var uintL digitcount = floor(count,(intDsize/8));
  if (((count%(intDsize/8)) > 0) || (*bufferptr & bit(7)))
    digitcount++;
  # As bitsize < intDsize*uintWC_max,
  # digitcount <= ceiling((bitsize+1)/intDsize) <= uintWC_max .
  var object big = allocate_bignum(digitcount,0); # new Bignum >0
  TheBignum(big)->data[0] = 0; # set highest Digit to 0
  # Fill remaining Digits from right to left,
  # thereby translate sequence of Bytes into sequence of uintD:
  bufferptr = &TheSbvector(*buffer_)->data[0];
  #if BIG_ENDIAN_P
  {
    var uintB* bigptr = (uintB*)(&TheBignum(big)->data[digitcount]);
    dotimespL(count,count, { *--bigptr = *bufferptr++; } );
  }
  #else
  {
    var uintD* bigptr = &TheBignum(big)->data[digitcount];
    var uintL count2;
    #define GET_NEXT_BYTE(i)  digit |= ((uintD)(*bufferptr++) << (8*i));
    dotimespL(count2,floor(count,intDsize/8), {
      var uintD digit = 0;
      DOCONSTTIMES(intDsize/8,GET_NEXT_BYTE); # GET_NEXT_BYTE(0..intDsize/8-1)
      *--bigptr = digit;
    });
    #undef GET_NEXT_BYTE
    count2 = count % (intDsize/8);
    if (count2>0) {
      var uintL shiftcount = 0;
      var uintD digit = (uintD)(*bufferptr++);
      dotimesL(count2,count2-1, {
        shiftcount += 8;
        digit |= ((uintD)(*bufferptr++) << shiftcount);
      });
      *--bigptr = digit;
    }
  }
  #endif
  # Since (intDsize/8)*(digitcount-1) <= count <= (intDsize/8)*digitcount
  # everything is filled.
  return big;
}

# Converts a little-endian byte sequence to an integer.
# > bytesize: number of given 8-bit bytes of the integer, > 0,
#             < intDsize/8*uintWC_max
# > bufferptr: address of bytesize bytes of memory
# < result: an integer with I_integer_length(result) < 8*bytesize
global maygc object LEbytes_to_I (uintL bytesize, const uintB* bufferptr) {
  var gcv_object_t fake = FAKE_8BIT_VECTOR(bufferptr);
  return LESbvector_to_I(bytesize,&fake);
}

# Converts a little-endian byte sequence to an integer.
# > bytesize: number of given 8-bit bytes of the integer, > 0,
#             < intDsize/8*uintWC_max
# > *buffer_: address of a simple-8bit-vector (or of a fake)
#             containing bytesize bytes of memory
# < result: an integer with I_integer_length(result) < 8*bytesize
global maygc object LESbvector_to_I (uintL bytesize, const gcv_object_t* buffer_) {
  # Normalize number in buffer:
  var uintB* bufferptr = &TheSbvector(*buffer_)->data[bytesize-1];
  var sintD sign;
  var uintL count = bytesize;
  if (!(*bufferptr & bit(7))) {
    sign = 0;
    # Normalize, highest Bit must remain 0:
    while ((count>=2) && (*bufferptr==0)
           && !(*(bufferptr-1) & bit(7))) {
      count--; bufferptr--;
    }
    # Make number:
    if # at most oint_data_len+1 Bits, count <2^oint_data_len ?
      ((count <= floor(oint_data_len,8))
       || ((count == floor(oint_data_len,8)+1)
           && (*bufferptr < bit(oint_data_len%8)))) {
      # yes -> build Fixnum >=0:
      var uintL value = 0;
      until (count==0) { value = (value<<8) | *bufferptr--; count--; }
      return posfixnum(value);
    }
  } else {
    sign = -1;
    # Normalize, highest Bit must remain 1:
    while ((count>=2) && (*bufferptr==(uintB)(-1))
           && (*(bufferptr-1) & bit(7))) {
      count--; bufferptr--;
    }
    # Make number:
    if # at most oint_data_len+1 Bits, count >=-2^oint_data_len ?
      ((count <= floor(oint_data_len,8))
       || ((count == floor(oint_data_len,8)+1)
           && (*bufferptr >= (uintB)(-bit(oint_data_len%8))))) {
      # yes -> build Fixnum <0:
      var uintL value = (uintL)(-1);
      until (count==0) { value = (value<<8) | *bufferptr--; count--; }
      return negfixnum(-wbitm(intLsize)+(oint)value);
    }
  }
  # Make bignum:
  var uintL digitcount = ceiling(count,(intDsize/8));
  # As bitsize < intDsize*uintWC_max,
  # digitcount <= ceiling(bitsize/intDsize) <= uintWC_max .
  var object big = allocate_bignum(digitcount,(sintB)sign);
  TheBignum(big)->data[0] = sign; # set highest Word to sign
  # Fill the remaining Digits from right to left,
  # thereby translate sequence of Bytes into sequence of uintD:
  bufferptr = &TheSbvector(*buffer_)->data[0];
  #if BIG_ENDIAN_P
  {
    var uintB* bigptr = (uintB*)(TheBignum(big)->data+digitcount);
    dotimespL(count,count, { *--bigptr = *bufferptr++; } );
  }
  #else
  {
    var uintD* bigptr = TheBignum(big)->data+digitcount;
    var uintL count2;
    #define GET_NEXT_BYTE(i)  digit |= ((uintD)(*bufferptr++) << (8*i));
    dotimespL(count2,floor(count,intDsize/8), {
      var uintD digit = 0;
      DOCONSTTIMES(intDsize/8,GET_NEXT_BYTE); # GET_NEXT_BYTE(0..intDsize/8-1)
      *--bigptr = digit;
    });
    #undef GET_NEXT_BYTE
    count2 = count % (intDsize/8);
    if (count2>0) {
      var uintL shiftcount = 0;
      var uintD digit = (uintD)(*bufferptr++);
      dotimesL(count2,count2-1, {
        shiftcount += 8;
        digit |= ((uintD)(*bufferptr++) << shiftcount);
      });
      *--bigptr = digit ^ (sign << (shiftcount+8));
    }
  }
  #endif
  # Since (intDsize/8)*(digitcount-1) < count <= (intDsize/8)*digitcount
  # everything is filled.
  return big;
}

# Converts an unsigned integer to a little-endian byte sequence.
# > obj: an integer
# > bitsize: maximum number of bits of the integer
# > bufferptr: pointer to bytesize = ceiling(bitsize,8) bytes of memory
# < false and bufferptr[0..bytesize-1] filled, if obj >= 0 and
#                                              I_integer_length(obj) <= bitsize;
#   true, if obj is out of range
global bool UI_to_LEbytes (object obj, uintL bitsize, uintB* bufferptr) {
  if (!positivep(obj))
    return true;
  # obj is an integer >=0.
  var uintL bytesize = ceiling(bitsize,8);
  # Transfer obj into the buffer:
  {
    var uintL count = bytesize;
    if (posfixnump(obj)) { # obj is a Fixnum >=0
      var uintL value = posfixnum_to_L(obj);
      # check value < 2^bitsize:
      if (!((bitsize>=oint_data_len) || (value < bit(bitsize))))
        return true;
      # store value in Bitbuffer:
      until (value==0) {
        *bufferptr++ = (uint8)value; value = value>>8; count--;
      }
    } else { # obj is a Bignum >0
      var uintL len = (uintL)Bignum_length(obj);
      # check obj < 2^bitsize:
      if (!((floor(bitsize,intDsize) >= len)
            || ((floor(bitsize,intDsize) == len-1)
                && (TheBignum(obj)->data[0] < bit(bitsize%intDsize)))))
        return true;
      #if BIG_ENDIAN_P
      {
        var uintB* ptr = (uintB*)&TheBignum(obj)->data[len];
        # convert Digit-Length in Byte-Length:
        len = (intDsize/8)*len;
        #define CHECK_NEXT_BYTE(i)  \
          if (((uintB*)(&TheBignum(obj)->data[0]))[i] != 0) goto len_ok; \
          len--;
        DOCONSTTIMES(intDsize/8,CHECK_NEXT_BYTE); # CHECK_NEXT_BYTE(0..intDsize/8-1)
        #undef CHECK_NEXT_BYTE
      len_ok:
        # store obj in Bitbuffer:
        count = count - len;
        dotimespL(len,len, { *bufferptr++ = *--ptr; } );
      }
      #else
      {
        var uintD* ptr = &TheBignum(obj)->data[len];
        len--;
        count -= (intDsize/8)*len;
        dotimesL(len,len, {
          var uintD digit = *--ptr;
          doconsttimes(intDsize/8, {
            *bufferptr++ = (uintB)digit; digit = digit >> 8;
          });
        });
        var uintD digit = *--ptr;
        doconsttimes(intDsize/8, {
          if (digit==0) goto ok;
          *bufferptr++ = (uintB)digit; digit = digit >> 8;
          count--;
        });
      ok: ;
      }
    #endif
    }
    if (count > 0) {
      begin_system_call();
      memset(bufferptr,0,count);
      end_system_call();
    }
  }
  return false;
}

# Converts an integer to a little-endian byte sequence.
# > obj: an integer
# > bitsize: maximum number of bits of the integer, including the sign bit
# > bufferptr: pointer to bytesize = ceiling(bitsize,8) bytes of memory
# < false and bufferptr[0..bytesize-1] filled, if I_integer_length(obj) < bitsize;
#   true, if obj is out of range
global bool I_to_LEbytes (object obj, uintL bitsize, uintB* bufferptr) {
  # obj is an integer.
  var uintL bytesize = ceiling(bitsize,8);
  # Transfer obj into the buffer:
  {
    var uintL count = bytesize;
    var uintL sign = (sintL)R_sign(obj);
    if (fixnump(obj)) {
      # obj is a Fixnum
      var uintL value = fixnum_to_L(obj); # >=0 or <0, according to sign
      # check 0 <= value < 2^(bitsize-1) resp. -2^(bitsize-1) <= value < 0:
      value = value^sign;
      if (!((bitsize>oint_data_len) || (value < bit(bitsize-1))))
        return true;
      # store value^sign in Bitbuffer:
      until (value == 0) {
        *bufferptr++ = (uint8)(value^sign); value = value>>8; count--;
      }
      if (count > 0) {
        begin_system_call();
        memset(bufferptr,(uint8)sign,count);
        end_system_call();
      }
    } else {
      # obj is a Bignum
      var uintL len = (uintL)Bignum_length(obj);
      # check -2^(bitsize-1) <= obj < 2^(bitsize-1):
      if (!((floor(bitsize,intDsize) >= len)
            || ((bitsize > intDsize*(len-1))
                && ((TheBignum(obj)->data[0] ^ (uintD)sign) <
                    bit((bitsize%intDsize)-1)))))
        return true;
      #if BIG_ENDIAN_P
      {
        var uintB* ptr = (uintB*)&TheBignum(obj)->data[len];
        # convert Digit-Length in Byte-Length:
        len = (intDsize/8)*len;
        #define CHECK_NEXT_BYTE(i)  \
          if (((uintB*)(&TheBignum(obj)->data[0]))[i] != (uintB)sign) goto len_ok; \
          len--;
        DOCONSTTIMES(intDsize/8,CHECK_NEXT_BYTE); # CHECK_NEXT_BYTE(0..intDsize/8-1)
        #undef CHECK_NEXT_BYTE
      len_ok:
        # store obj in Bitbuffer:
        count = count - len;
        dotimespL(len,len, { *bufferptr++ = *--ptr; } );
      }
      #else
      {
        var uintD* ptr = &TheBignum(obj)->data[len];
        len--;
        count -= (intDsize/8)*len;
        dotimesL(len,len, {
          var uintD digit = *--ptr;
          doconsttimes(intDsize/8, {
            *bufferptr++ = (uintB)digit; digit = digit >> 8;
          });
        });
        var sintD digit = *--ptr;
        doconsttimes(intDsize/8, {
          if (digit == (sintD)sign) goto ok;
          *bufferptr++ = (uintB)digit; digit = digit >> 8;
          count--;
        });
      ok: ;
      }
      #endif
      if (count > 0) {
        begin_system_call();
        memset(bufferptr,(uintB)sign,count);
        end_system_call();
      }
    }
  }
  return false;
}



--__--__--

Message: 3
From: Bruno Haible <ha...@us...>
To: cli...@li...
Subject: clisp/src/po Makefile.devel,1.37,1.38
Date: Sat, 29 Jan 2005 14:55:09 +0000
Reply-To: cli...@li...

Update of /cvsroot/clisp/clisp/src/po
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv24689/src/po

Modified Files:
	Makefile.devel 
Log Message:
New primitives for serializing and deserializing integers.


Index: Makefile.devel
===================================================================
RCS file: /cvsroot/clisp/clisp/src/po/Makefile.devel,v
retrieving revision 1.37
retrieving revision 1.38
diff -u -d -r1.37 -r1.38
--- Makefile.devel	20 Dec 2004 14:15:25 -0000	1.37
+++ Makefile.devel	29 Jan 2005 14:55:06 -0000	1.38
@@ -37,8 +37,8 @@
  misc time predtype symbol \
  lisparit aridecl arilev0 arilev1 arilev1c arilev1e arilev1i intelem \
     intlog intplus intcomp intbyte intmal intdiv intgcd int2adic intsqrt \
-    intprint intread rational sfloat ffloat dfloat lfloat flo_konv flo_rest \
-    realelem realrand realtran compelem comptran \
+    intprint intread intserial rational sfloat ffloat dfloat lfloat flo_konv \
+    flo_rest realelem realrand realtran compelem comptran \
  affi \
  foreign \
  unixaux win32aux \



--__--__--

Message: 4
From: Bruno Haible <ha...@us...>
To: cli...@li...
Subject: clisp/src lispbibl.d,1.610,1.611 genclisph.d,1.168,1.169 intelem.d,1.29,1.30 socket.d,1.89,1.90 ChangeLog,1.4171,1.4172
Date: Sat, 29 Jan 2005 14:57:13 +0000
Reply-To: cli...@li...

Update of /cvsroot/clisp/clisp/src
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv25162/src

Modified Files:
	lispbibl.d genclisph.d intelem.d socket.d ChangeLog 
Log Message:
Remove ill-designed function udigits_to_I.


Index: socket.d
===================================================================
RCS file: /cvsroot/clisp/clisp/src/socket.d,v
retrieving revision 1.89
retrieving revision 1.90
diff -u -d -r1.89 -r1.90
--- socket.d	25 Jan 2005 14:35:06 -0000	1.89
+++ socket.d	29 Jan 2005 14:57:10 -0000	1.90
@@ -337,7 +337,7 @@
 
 local int string_to_addr1 (void* addr, int addrlen, int family, void* ret) {
   *(object*)ret = (addrlen
-                   ? udigits_to_I(addr,addrlen)
+                   ? LEbytes_to_UI(addrlen,(const uintB*)addr)
                    : asciz_to_string((char*)addr,O(misc_encoding)));
   (void)family; /* ignore */
   return 0;

Index: lispbibl.d
===================================================================
RCS file: /cvsroot/clisp/clisp/src/lispbibl.d,v
retrieving revision 1.610
retrieving revision 1.611
diff -u -d -r1.610 -r1.611
--- lispbibl.d	29 Jan 2005 14:55:07 -0000	1.610
+++ lispbibl.d	29 Jan 2005 14:57:10 -0000	1.611
@@ -14442,7 +14442,6 @@
  there must be room for 1 digit below of MSDptr.
  can trigger GC */
 extern maygc object UDS_to_I (uintD* MSDptr, uintC len);
-extern maygc object udigits_to_I (void* digits, uintC len);
 /* is used by modules */
 
 /* Digit Sequence to Integer

Index: intelem.d
===================================================================
RCS file: /cvsroot/clisp/clisp/src/intelem.d,v
retrieving revision 1.29
retrieving revision 1.30
diff -u -d -r1.29 -r1.30
--- intelem.d	25 Jan 2005 09:32:48 -0000	1.29
+++ intelem.d	29 Jan 2005 14:57:10 -0000	1.30
@@ -1357,27 +1357,6 @@
   return NDS_to_I(MSDptr,len);
 }
 
-/* bit sequence --> integer */
-global maygc object udigits_to_I (void* digits, uintC len) {
-  var uintL bn_size = ceiling(len,sizeof(uintD));
-  var uintD total = bn_size * sizeof(uintD);
-  var void *data = (len == total ? digits :
-                    total<=1024 ? alloca(total) : my_malloc(total));
-  if (data != digits) {      /* len is not divisible by sizeof(uintD) */
-    begin_system_call();
-    memset(data,0,total);
-    memcpy((char*)data + total - len,digits,len);
-    end_system_call();
-  }
-  var object ret = UDS_to_I(data,bn_size);
-  if (data != digits && total>1024) {
-    begin_system_call();
-    free(data);
-    end_system_call();
-  }
-  return ret;
-}
-
 /* Digit Sequence to Integer
  DS_to_I(MSDptr,len)
  convert DS MSDptr/len/.. into Integer.

Index: genclisph.d
===================================================================
RCS file: /cvsroot/clisp/clisp/src/genclisph.d,v
retrieving revision 1.168
retrieving revision 1.169
diff -u -d -r1.168 -r1.169
--- genclisph.d	29 Jan 2005 14:55:10 -0000	1.168
+++ genclisph.d	29 Jan 2005 14:57:10 -0000	1.169
@@ -2213,7 +2213,6 @@
   printf("#define I_to_slong  I_to_sint64\n");
 #endif
   printf("extern object UDS_to_I (uintD* MSDptr, uintC len);\n");
-  printf("extern object udigits_to_I (void* digits, uintC len);\n");
   printf("extern object DS_to_I (const uintD* MSDptr, uintC len);\n");
 #if notused
   printf("extern object I_1_plus_I (object x);\n");

Index: ChangeLog
===================================================================
RCS file: /cvsroot/clisp/clisp/src/ChangeLog,v
retrieving revision 1.4171
retrieving revision 1.4172
diff -u -d -r1.4171 -r1.4172
--- ChangeLog	29 Jan 2005 14:55:10 -0000	1.4171
+++ ChangeLog	29 Jan 2005 14:57:10 -0000	1.4172
@@ -1,5 +1,13 @@
 2005-01-28  Bruno Haible  <br...@cl...>
 
+	* lispbibl.d (udigits_to_I): Remove declaration.
+	* intelem.d (udigits_to_I): Remove function.
+	* socket.d (string_to_addr1): Call LEbytes_to_UI instead of
+	udigits_to_I.
+	* genclisph.d (main): Don't emit udigits_to_I declaration.
+
+2005-01-28  Bruno Haible  <br...@cl...>
+
 	* lispbibl.d (FAKE_8BIT_VECTOR): New macro.
 	(LEbytes_to_UI, LESbvector_to_UI, LEbytes_to_I, LESbvector_to_I,
 	UI_to_LEbytes, I_to_LEbytes): New declarations.




--__--__--

_______________________________________________
clisp-cvs mailing list
cli...@li...
https://lists.sourceforge.net/lists/listinfo/clisp-cvs


End of clisp-cvs Digest