File:  [gforth] / gforth / prim
Revision 1.30: download - view: text, annotated - select for diffs
Sun May 9 21:21:55 1999 UTC (20 years, 4 months ago) by anton
Branches: MAIN
CVS tags: HEAD
factored out name-too-short? in (').
fixed potential portability bug in read-line (check result of ferror).
added M: and M; to indentation lists in gforth.el

    1: \ Gforth primitives
    2: 
    3: \ Copyright (C) 1995,1996,1997,1998 Free Software Foundation, Inc.
    4: 
    5: \ This file is part of Gforth.
    6: 
    7: \ Gforth is free software; you can redistribute it and/or
    8: \ modify it under the terms of the GNU General Public License
    9: \ as published by the Free Software Foundation; either version 2
   10: \ of the License, or (at your option) any later version.
   11: 
   12: \ This program is distributed in the hope that it will be useful,
   13: \ but WITHOUT ANY WARRANTY; without even the implied warranty of
   14: \ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
   15: \ GNU General Public License for more details.
   16: 
   17: \ You should have received a copy of the GNU General Public License
   18: \ along with this program; if not, write to the Free Software
   19: \ Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
   20: 
   21: 
   22: \ WARNING: This file is processed by m4. Make sure your identifiers
   23: \ don't collide with m4's (e.g. by undefining them).
   24: \ 
   25: \ 
   26: \ 
   27: \ This file contains primitive specifications in the following format:
   28: \ 
   29: \ forth name	stack effect	category	[pronunciation]
   30: \ [""glossary entry""]
   31: \ C code
   32: \ [:
   33: \ Forth code]
   34: \ 
   35: \ prims2x is pedantic about tabs vs. blanks. The fields of the first
   36: \ line of a primitive are separated by tabs, the stack items in a
   37: \ stack effect by blanks.
   38: \
   39: \ Both pronounciation and stack items (in the stack effect) must
   40: \ conform to the C name syntax or the C compiler will complain.
   41: \ 
   42: \ 
   43: \ These specifications are automatically translated into C-code for the
   44: \ interpreter and into some other files. I hope that your C compiler has
   45: \ decent optimization, otherwise the automatically generated code will
   46: \ be somewhat slow. The Forth version of the code is included for manual
   47: \ compilers, so they will need to compile only the important words.
   48: \ 
   49: \ Note that stack pointer adjustment is performed according to stack
   50: \ effect by automatically generated code and NEXT is automatically
   51: \ appended to the C code. Also, you can use the names in the stack
   52: \ effect in the C code. Stack access is automatic. One exception: if
   53: \ your code does not fall through, the results are not stored into the
   54: \ stack. Use different names on both sides of the '--', if you change a
   55: \ value (some stores to the stack are optimized away).
   56: \ 
   57: \ 
   58: \ 
   59: \ The stack variables have the following types:
   60: \ 
   61: \ name matches	type
   62: \ f.*		Bool
   63: \ c.*		Char
   64: \ [nw].*		Cell
   65: \ u.*		UCell
   66: \ d.*		DCell
   67: \ ud.*		UDCell
   68: \ r.*		Float
   69: \ a_.*		Cell *
   70: \ c_.*		Char *
   71: \ f_.*		Float *
   72: \ df_.*		DFloat *
   73: \ sf_.*		SFloat *
   74: \ xt.*		XT
   75: \ wid.*		WID
   76: \ f83name.*	F83Name *
   77: \ 
   78: \ 
   79: \ 
   80: \ In addition the following names can be used:
   81: \ ip	the instruction pointer
   82: \ sp	the data stack pointer
   83: \ rp	the parameter stack pointer
   84: \ lp	the locals stack pointer
   85: \ NEXT	executes NEXT
   86: \ cfa	
   87: \ NEXT1	executes NEXT1
   88: \ FLAG(x)	makes a Forth flag from a C flag
   89: \ 
   90: \ 
   91: \ 
   92: \ Percentages in comments are from Koopmans book: average/maximum use
   93: \ (taken from four, not very representative benchmarks)
   94: \ 
   95: \ 
   96: \ 
   97: \ To do:
   98: \ 
   99: \ throw execute, cfa and NEXT1 out?
  100: \ macroize *ip, ip++, *ip++ (pipelining)?
  101: 
  102: \ these m4 macros would collide with identifiers
  103: undefine(`index')
  104: undefine(`shift')
  105: 
  106: noop	--		gforth
  107: ;
  108: :
  109:  ;
  110: 
  111: lit	-- w		gforth
  112: w = (Cell)NEXT_INST;
  113: INC_IP(1);
  114: :
  115:  r> dup @ swap cell+ >r ;
  116: 
  117: execute		xt --		core
  118: ""Perform the semantics represented by the execution token, @i{xt}.""
  119: ip=IP;
  120: IF_TOS(TOS = sp[0]);
  121: EXEC(xt);
  122: 
  123: perform		a_addr --	gforth
  124: ""Equivalent to @code{@ execute}.""
  125: /* and pfe */
  126: ip=IP;
  127: IF_TOS(TOS = sp[0]);
  128: EXEC(*(Xt *)a_addr);
  129: :
  130:  @ execute ;
  131: 
  132: \+glocals
  133: 
  134: branch-lp+!#	--	gforth	branch_lp_plus_store_number
  135: /* this will probably not be used */
  136: branch_adjust_lp:
  137: lp += (Cell)(IP[1]);
  138: goto branch;
  139: 
  140: \+
  141: 
  142: branch	--		gforth
  143: branch:
  144: SET_IP((Xt *)(((Cell)IP)+(Cell)NEXT_INST));
  145: :
  146:  r> dup @ + >r ;
  147: 
  148: \ condbranch(forthname,restline,code,forthcode)
  149: \ this is non-syntactical: code must open a brace that is closed by the macro
  150: define(condbranch,
  151: $1	$2
  152: $3	SET_IP((Xt *)(((Cell)IP)+(Cell)NEXT_INST));
  153: 	NEXT;
  154: }
  155: else
  156:     INC_IP(1);
  157: $4
  158: 
  159: \+glocals
  160: 
  161: $1-lp+!#	$2_lp_plus_store_number
  162: $3    goto branch_adjust_lp;
  163: }
  164: else
  165:     INC_IP(2);
  166: 
  167: \+
  168: )
  169: 
  170: condbranch(?branch,f --		f83	question_branch,
  171: if (f==0) {
  172:     IF_TOS(TOS = sp[0]);
  173: ,:
  174:  0= dup     \ !f !f
  175:  r> dup @   \ !f !f IP branchoffset
  176:  rot and +  \ !f IP|IP+branchoffset
  177:  swap 0= cell and + \ IP''
  178:  >r ;)
  179: 
  180: \ we don't need an lp_plus_store version of the ?dup-stuff, because it
  181: \ is only used in if's (yet)
  182: 
  183: \+xconds
  184: 
  185: ?dup-?branch	f -- f	new	question_dupe_question_branch
  186: ""The run-time procedure compiled by @code{?DUP-IF}.""
  187: if (f==0) {
  188:   sp++;
  189:   IF_TOS(TOS = sp[0]);
  190:   SET_IP((Xt *)(((Cell)IP)+(Cell)NEXT_INST));
  191:   NEXT;
  192: }
  193: else
  194:   INC_IP(1);
  195: 
  196: ?dup-0=-?branch	f --	new	question_dupe_zero_equals_question_branch
  197: ""The run-time procedure compiled by @code{?DUP-0=-IF}.""
  198: /* the approach taken here of declaring the word as having the stack
  199: effect ( f -- ) and correcting for it in the branch-taken case costs a
  200: few cycles in that case, but is easy to convert to a CONDBRANCH
  201: invocation */
  202: if (f!=0) {
  203:   sp--;
  204:   SET_IP((Xt *)(((Cell)IP)+(Cell)NEXT_INST));
  205:   NEXT;
  206: }
  207: else
  208:   INC_IP(1);
  209: 
  210: \+
  211: 
  212: condbranch((next),--		cmFORTH	paren_next,
  213: if ((*rp)--) {
  214: ,:
  215:  r> r> dup 1- >r
  216:  IF dup @ + >r ELSE cell+ >r THEN ;)
  217: 
  218: condbranch((loop),--		gforth	paren_loop,
  219: Cell index = *rp+1;
  220: Cell limit = rp[1];
  221: if (index != limit) {
  222:     *rp = index;
  223: ,:
  224:  r> r> 1+ r> 2dup =
  225:  IF >r 1- >r cell+ >r
  226:  ELSE >r >r dup @ + >r THEN ;)
  227: 
  228: condbranch((+loop),n --		gforth	paren_plus_loop,
  229: /* !! check this thoroughly */
  230: Cell index = *rp;
  231: /* sign bit manipulation and test: (x^y)<0 is equivalent to (x<0) != (y<0) */
  232: /* dependent upon two's complement arithmetic */
  233: Cell olddiff = index-rp[1];
  234: if ((olddiff^(olddiff+n))>=0   /* the limit is not crossed */
  235:     || (olddiff^n)>=0          /* it is a wrap-around effect */) {
  236: #ifdef i386
  237:     *rp += n;
  238: #else
  239:     *rp = index + n;
  240: #endif
  241:     IF_TOS(TOS = sp[0]);
  242: ,:
  243:  r> swap
  244:  r> r> 2dup - >r
  245:  2 pick r@ + r@ xor 0< 0=
  246:  3 pick r> xor 0< 0= or
  247:  IF    >r + >r dup @ + >r
  248:  ELSE  >r >r drop cell+ >r THEN ;)
  249: 
  250: \+xconds
  251: 
  252: condbranch((-loop),u --		gforth	paren_minus_loop,
  253: /* !! check this thoroughly */
  254: Cell index = *rp;
  255: UCell olddiff = index-rp[1];
  256: if (olddiff>u) {
  257: #ifdef i386
  258:     *rp -= u;
  259: #else
  260:     *rp = index - u;
  261: #endif
  262:     IF_TOS(TOS = sp[0]);
  263: ,)
  264: 
  265: condbranch((s+loop),n --		gforth	paren_symmetric_plus_loop,
  266: ""The run-time procedure compiled by S+LOOP. It loops until the index
  267: crosses the boundary between limit and limit-sign(n). I.e. a symmetric
  268: version of (+LOOP).""
  269: /* !! check this thoroughly */
  270: Cell index = *rp;
  271: Cell diff = index-rp[1];
  272: Cell newdiff = diff+n;
  273: if (n<0) {
  274:     diff = -diff;
  275:     newdiff = -newdiff;
  276: }
  277: if (diff>=0 || newdiff<0) {
  278: #ifdef i386
  279:     *rp += n;
  280: #else
  281:     *rp = index + n;
  282: #endif
  283:     IF_TOS(TOS = sp[0]);
  284: ,)
  285: 
  286: \+
  287: 
  288: unloop		--	core
  289: rp += 2;
  290: :
  291:  r> rdrop rdrop >r ;
  292: 
  293: (for)	ncount --		cmFORTH		paren_for
  294: /* or (for) = >r -- collides with unloop! */
  295: *--rp = 0;
  296: *--rp = ncount;
  297: :
  298:  r> swap 0 >r >r >r ;
  299: 
  300: (do)	nlimit nstart --		gforth		paren_do
  301: /* or do it in high-level? 0.09/0.23% */
  302: *--rp = nlimit;
  303: *--rp = nstart;
  304: :
  305:  r> swap rot >r >r >r ;
  306: 
  307: (?do)	nlimit nstart --	gforth	paren_question_do
  308: *--rp = nlimit;
  309: *--rp = nstart;
  310: if (nstart == nlimit) {
  311:     IF_TOS(TOS = sp[0]);
  312:     goto branch;
  313:     }
  314: else {
  315:     INC_IP(1);
  316: }
  317: :
  318:   2dup =
  319:   IF   r> swap rot >r >r
  320:        dup @ + >r
  321:   ELSE r> swap rot >r >r
  322:        cell+ >r
  323:   THEN ;				\ --> CORE-EXT
  324: 
  325: \+xconds
  326: 
  327: (+do)	nlimit nstart --	gforth	paren_plus_do
  328: *--rp = nlimit;
  329: *--rp = nstart;
  330: if (nstart >= nlimit) {
  331:     IF_TOS(TOS = sp[0]);
  332:     goto branch;
  333:     }
  334: else {
  335:     INC_IP(1);
  336: }
  337: :
  338:  swap 2dup
  339:  r> swap >r swap >r
  340:  >=
  341:  IF
  342:      dup @ +
  343:  ELSE
  344:      cell+
  345:  THEN  >r ;
  346: 
  347: (u+do)	ulimit ustart --	gforth	paren_u_plus_do
  348: *--rp = ulimit;
  349: *--rp = ustart;
  350: if (ustart >= ulimit) {
  351:     IF_TOS(TOS = sp[0]);
  352:     goto branch;
  353:     }
  354: else {
  355:     INC_IP(1);
  356: }
  357: :
  358:  swap 2dup
  359:  r> swap >r swap >r
  360:  u>=
  361:  IF
  362:      dup @ +
  363:  ELSE
  364:      cell+
  365:  THEN  >r ;
  366: 
  367: (-do)	nlimit nstart --	gforth	paren_minus_do
  368: *--rp = nlimit;
  369: *--rp = nstart;
  370: if (nstart <= nlimit) {
  371:     IF_TOS(TOS = sp[0]);
  372:     goto branch;
  373:     }
  374: else {
  375:     INC_IP(1);
  376: }
  377: :
  378:  swap 2dup
  379:  r> swap >r swap >r
  380:  <=
  381:  IF
  382:      dup @ +
  383:  ELSE
  384:      cell+
  385:  THEN  >r ;
  386: 
  387: (u-do)	ulimit ustart --	gforth	paren_u_minus_do
  388: *--rp = ulimit;
  389: *--rp = ustart;
  390: if (ustart <= ulimit) {
  391:     IF_TOS(TOS = sp[0]);
  392:     goto branch;
  393:     }
  394: else {
  395:     INC_IP(1);
  396: }
  397: :
  398:  swap 2dup
  399:  r> swap >r swap >r
  400:  u<=
  401:  IF
  402:      dup @ +
  403:  ELSE
  404:      cell+
  405:  THEN  >r ;
  406: 
  407: \+
  408: 
  409: \ don't make any assumptions where the return stack is!!
  410: \ implement this in machine code if it should run quickly!
  411: 
  412: i	-- n		core
  413: n = *rp;
  414: :
  415: \ rp@ cell+ @ ;
  416:   r> r> tuck >r >r ;
  417: 
  418: i'	-- w		gforth		i_tick
  419: ""loop end value""
  420: w = rp[1];
  421: :
  422: \ rp@ cell+ cell+ @ ;
  423:   r> r> r> dup itmp ! >r >r >r itmp @ ;
  424: variable itmp
  425: 
  426: j	-- n		core
  427: n = rp[2];
  428: :
  429: \ rp@ cell+ cell+ cell+ @ ;
  430:   r> r> r> r> dup itmp ! >r >r >r >r itmp @ ;
  431: [IFUNDEF] itmp variable itmp [THEN]
  432: 
  433: k	-- n		gforth
  434: n = rp[4];
  435: :
  436: \ rp@ [ 5 cells ] Literal + @ ;
  437:   r> r> r> r> r> r> dup itmp ! >r >r >r >r >r >r itmp @ ;
  438: [IFUNDEF] itmp variable itmp [THEN]
  439: 
  440: \ digit is high-level: 0/0%
  441: 
  442: move	c_from c_to ucount --		core
  443: "" If @i{ucount}>0, copy the contents of @i{ucount} address units
  444: at @i{c-from} to @i{c-to}. @code{move} chooses its copy direction
  445: to avoid problems when @i{c-from}, @i{c-to} overlap.""
  446: memmove(c_to,c_from,ucount);
  447: /* make an Ifdef for bsd and others? */
  448: :
  449:  >r 2dup u< IF r> cmove> ELSE r> cmove THEN ;
  450: 
  451: cmove	c_from c_to u --	string
  452: "" If @i{u}>0, copy the contents of @i{ucount} characters from
  453: data space at @i{c-from} to @i{c-to}. The copy proceeds @code{char}-by-@code{char}
  454: from low address to high address.""
  455: while (u-- > 0)
  456:   *c_to++ = *c_from++;
  457: :
  458:  bounds ?DO  dup c@ I c! 1+  LOOP  drop ;
  459: 
  460: cmove>	c_from c_to u --	string	c_move_up
  461: "" If @i{u}>0, copy the contents of @i{ucount} characters from
  462: data space at @i{c-from} to @i{c-to}. The copy proceeds @code{char}-by-@code{char}
  463: from high address to low address.""
  464: while (u-- > 0)
  465:   c_to[u] = c_from[u];
  466: :
  467:  dup 0= IF  drop 2drop exit  THEN
  468:  rot over + -rot bounds swap 1-
  469:  DO  1- dup c@ I c!  -1 +LOOP  drop ;
  470: 
  471: fill	c_addr u c --	core
  472: "" If @i{u}>0, store character @i{c} in each of @i{u} consecutive
  473: @code{char} addresses in memory, starting at address @i{c-addr}.""
  474: memset(c_addr,c,u);
  475: :
  476:  -rot bounds
  477:  ?DO  dup I c!  LOOP  drop ;
  478: 
  479: compare		c_addr1 u1 c_addr2 u2 -- n	string
  480: ""Compare two strings lexicographically. If they are equal, @i{n} is 0; if
  481: the first string is smaller, @i{n} is -1; if the first string is larger, @i{n}
  482: is 1. Currently this is based on the machine's character
  483: comparison. In the future, this may change to consider the current
  484: locale and its collation order.""
  485: n = memcmp(c_addr1, c_addr2, u1<u2 ? u1 : u2);
  486: if (n==0)
  487:   n = u1-u2;
  488: if (n<0)
  489:   n = -1;
  490: else if (n>0)
  491:   n = 1;
  492: :
  493:  rot 2dup - >r min swap -text dup
  494:  IF    rdrop
  495:  ELSE  drop r@ 0>
  496:        IF    rdrop -1
  497:        ELSE  r> 1 and
  498:        THEN
  499:  THEN ;
  500: 
  501: -text		c_addr1 u c_addr2 -- n	new	dash_text
  502: n = memcmp(c_addr1, c_addr2, u);
  503: if (n<0)
  504:   n = -1;
  505: else if (n>0)
  506:   n = 1;
  507: :
  508:  swap bounds
  509:  ?DO  dup c@ I c@ = WHILE  1+  LOOP  drop 0
  510:  ELSE  c@ I c@ - unloop  THEN  -text-flag ;
  511: : -text-flag ( n -- -1/0/1 )
  512:  dup 0< IF  drop -1  ELSE  0>  1 and  THEN  ;
  513: 
  514: toupper	c1 -- c2	gforth
  515: ""If @i{c1} is a lower-case character (in the current locale), @i{c2}
  516: is the equivalent upper-case character. All other characters are unchanged.""
  517: c2 = toupper(c1);
  518: :
  519:  dup [char] a - [ char z char a - 1 + ] Literal u<  bl and - ;
  520: 
  521: capscomp	c_addr1 u c_addr2 -- n	new
  522: n = memcasecmp(c_addr1, c_addr2, u); /* !! use something that works in all locales */
  523: if (n<0)
  524:   n = -1;
  525: else if (n>0)
  526:   n = 1;
  527: :
  528:  swap bounds
  529:  ?DO  dup c@ I c@ <>
  530:      IF  dup c@ toupper I c@ toupper =
  531:      ELSE  true  THEN  WHILE  1+  LOOP  drop 0
  532:  ELSE  c@ toupper I c@ toupper - unloop  THEN  -text-flag ;
  533: 
  534: -trailing	c_addr u1 -- c_addr u2		string	dash_trailing
  535: ""Adjust the string specified by @i{c-addr, u1} to remove all trailing
  536: spaces. @i{u2} is the length of the modified string.""
  537: u2 = u1;
  538: while (u2>0 && c_addr[u2-1] == ' ')
  539:   u2--;
  540: :
  541:  BEGIN  1- 2dup + c@ bl =  WHILE
  542:         dup  0= UNTIL  ELSE  1+  THEN ;
  543: 
  544: /string		c_addr1 u1 n -- c_addr2 u2	string	slash_string
  545: ""Adjust the string specified by @i{c-addr1, u1} to remove @i{n}
  546: characters from the start of the string.""
  547: c_addr2 = c_addr1+n;
  548: u2 = u1-n;
  549: :
  550:  tuck - >r + r> dup 0< IF  - 0  THEN ;
  551: 
  552: +	n1 n2 -- n		core	plus
  553: n = n1+n2;
  554: 
  555: \ PFE-0.9.14 has it differently, but the next release will have it as follows
  556: under+	n1 n2 n3 -- n n2	gforth	under_plus
  557: ""add @i{n3} to @i{n1} (giving @i{n})""
  558: n = n1+n3;
  559: :
  560:  rot + swap ;
  561: 
  562: -	n1 n2 -- n		core	minus
  563: n = n1-n2;
  564: :
  565:  negate + ;
  566: 
  567: negate	n1 -- n2		core
  568: /* use minus as alias */
  569: n2 = -n1;
  570: :
  571:  invert 1+ ;
  572: 
  573: 1+	n1 -- n2		core		one_plus
  574: n2 = n1+1;
  575: :
  576:  1 + ;
  577: 
  578: 1-	n1 -- n2		core		one_minus
  579: n2 = n1-1;
  580: :
  581:  1 - ;
  582: 
  583: max	n1 n2 -- n	core
  584: if (n1<n2)
  585:   n = n2;
  586: else
  587:   n = n1;
  588: :
  589:  2dup < IF swap THEN drop ;
  590: 
  591: min	n1 n2 -- n	core
  592: if (n1<n2)
  593:   n = n1;
  594: else
  595:   n = n2;
  596: :
  597:  2dup > IF swap THEN drop ;
  598: 
  599: abs	n1 -- n2	core
  600: if (n1<0)
  601:   n2 = -n1;
  602: else
  603:   n2 = n1;
  604: :
  605:  dup 0< IF negate THEN ;
  606: 
  607: *	n1 n2 -- n		core	star
  608: n = n1*n2;
  609: :
  610:  um* drop ;
  611: 
  612: /	n1 n2 -- n		core	slash
  613: n = n1/n2;
  614: :
  615:  /mod nip ;
  616: 
  617: mod	n1 n2 -- n		core
  618: n = n1%n2;
  619: :
  620:  /mod drop ;
  621: 
  622: /mod	n1 n2 -- n3 n4		core		slash_mod
  623: n4 = n1/n2;
  624: n3 = n1%n2; /* !! is this correct? look into C standard! */
  625: :
  626:  >r s>d r> fm/mod ;
  627: 
  628: 2*	n1 -- n2		core		two_star
  629: n2 = 2*n1;
  630: :
  631:  dup + ;
  632: 
  633: 2/	n1 -- n2		core		two_slash
  634: /* !! is this still correct? */
  635: n2 = n1>>1;
  636: :
  637:  dup MINI and IF 1 ELSE 0 THEN
  638:  [ bits/byte cell * 1- ] literal 
  639:  0 DO 2* swap dup 2* >r MINI and 
  640:      IF 1 ELSE 0 THEN or r> swap
  641:  LOOP nip ;
  642: 
  643: fm/mod	d1 n1 -- n2 n3		core		f_m_slash_mod
  644: ""Floored division: @i{d1} = @i{n3}*@i{n1}+@i{n2}, @i{n1}>@i{n2}>=0 or 0>=@i{n2}>@i{n1}.""
  645: #ifdef BUGGY_LONG_LONG
  646: DCell r = fmdiv(d1,n1);
  647: n2=r.hi;
  648: n3=r.lo;
  649: #else
  650: /* assumes that the processor uses either floored or symmetric division */
  651: n3 = d1/n1;
  652: n2 = d1%n1;
  653: /* note that this 1%-3>0 is optimized by the compiler */
  654: if (1%-3>0 && (d1<0) != (n1<0) && n2!=0) {
  655:   n3--;
  656:   n2+=n1;
  657: }
  658: #endif
  659: :
  660:  dup >r dup 0< IF  negate >r dnegate r>  THEN
  661:  over       0< IF  tuck + swap  THEN
  662:  um/mod
  663:  r> 0< IF  swap negate swap  THEN ;
  664: 
  665: sm/rem	d1 n1 -- n2 n3		core		s_m_slash_rem
  666: ""Symmetric division: @i{d1} = @i{n3}*@i{n1}+@i{n2}, sign(@i{n2})=sign(@i{d1}) or 0.""
  667: #ifdef BUGGY_LONG_LONG
  668: DCell r = smdiv(d1,n1);
  669: n2=r.hi;
  670: n3=r.lo;
  671: #else
  672: /* assumes that the processor uses either floored or symmetric division */
  673: n3 = d1/n1;
  674: n2 = d1%n1;
  675: /* note that this 1%-3<0 is optimized by the compiler */
  676: if (1%-3<0 && (d1<0) != (n1<0) && n2!=0) {
  677:   n3++;
  678:   n2-=n1;
  679: }
  680: #endif
  681: :
  682:  over >r dup >r abs -rot
  683:  dabs rot um/mod
  684:  r> r@ xor 0< IF       negate       THEN
  685:  r>        0< IF  swap negate swap  THEN ;
  686: 
  687: m*	n1 n2 -- d		core	m_star
  688: #ifdef BUGGY_LONG_LONG
  689: d = mmul(n1,n2);
  690: #else
  691: d = (DCell)n1 * (DCell)n2;
  692: #endif
  693: :
  694:  2dup      0< and >r
  695:  2dup swap 0< and >r
  696:  um* r> - r> - ;
  697: 
  698: um*	u1 u2 -- ud		core	u_m_star
  699: /* use u* as alias */
  700: #ifdef BUGGY_LONG_LONG
  701: ud = ummul(u1,u2);
  702: #else
  703: ud = (UDCell)u1 * (UDCell)u2;
  704: #endif
  705: :
  706:    >r >r 0 0 r> r> [ 8 cells ] literal 0
  707:    DO
  708:        over >r dup >r 0< and d2*+ drop
  709:        r> 2* r> swap
  710:    LOOP 2drop ;
  711: : d2*+ ( ud n -- ud+n c )
  712:    over MINI
  713:    and >r >r 2dup d+ swap r> + swap r> ;
  714: 
  715: um/mod	ud u1 -- u2 u3		core	u_m_slash_mod
  716: #ifdef BUGGY_LONG_LONG
  717: UDCell r = umdiv(ud,u1);
  718: u2=r.hi;
  719: u3=r.lo;
  720: #else
  721: u3 = ud/u1;
  722: u2 = ud%u1;
  723: #endif
  724: :
  725:    0 swap [ 8 cells 1 + ] literal 0
  726:    ?DO /modstep
  727:    LOOP drop swap 1 rshift or swap ;
  728: : /modstep ( ud c R: u -- ud-?u c R: u )
  729:    >r over r@ u< 0= or IF r@ - 1 ELSE 0 THEN  d2*+ r> ;
  730: : d2*+ ( ud n -- ud+n c )
  731:    over MINI
  732:    and >r >r 2dup d+ swap r> + swap r> ;
  733: 
  734: m+	d1 n -- d2		double		m_plus
  735: #ifdef BUGGY_LONG_LONG
  736: d2.lo = d1.lo+n;
  737: d2.hi = d1.hi - (n<0) + (d2.lo<d1.lo);
  738: #else
  739: d2 = d1+n;
  740: #endif
  741: :
  742:  s>d d+ ;
  743: 
  744: d+	d1 d2 -- d		double	d_plus
  745: #ifdef BUGGY_LONG_LONG
  746: d.lo = d1.lo+d2.lo;
  747: d.hi = d1.hi + d2.hi + (d.lo<d1.lo);
  748: #else
  749: d = d1+d2;
  750: #endif
  751: :
  752:  rot + >r tuck + swap over u> r> swap - ;
  753: 
  754: d-	d1 d2 -- d		double		d_minus
  755: #ifdef BUGGY_LONG_LONG
  756: d.lo = d1.lo - d2.lo;
  757: d.hi = d1.hi-d2.hi-(d1.lo<d2.lo);
  758: #else
  759: d = d1-d2;
  760: #endif
  761: :
  762:  dnegate d+ ;
  763: 
  764: dnegate	d1 -- d2		double
  765: /* use dminus as alias */
  766: #ifdef BUGGY_LONG_LONG
  767: d2 = dnegate(d1);
  768: #else
  769: d2 = -d1;
  770: #endif
  771: :
  772:  invert swap negate tuck 0= - ;
  773: 
  774: d2*	d1 -- d2		double		d_two_star
  775: #ifdef BUGGY_LONG_LONG
  776: d2.lo = d1.lo<<1;
  777: d2.hi = (d1.hi<<1) | (d1.lo>>(CELL_BITS-1));
  778: #else
  779: d2 = 2*d1;
  780: #endif
  781: :
  782:  2dup d+ ;
  783: 
  784: d2/	d1 -- d2		double		d_two_slash
  785: #ifdef BUGGY_LONG_LONG
  786: d2.hi = d1.hi>>1;
  787: d2.lo= (d1.lo>>1) | (d1.hi<<(CELL_BITS-1));
  788: #else
  789: d2 = d1>>1;
  790: #endif
  791: :
  792:  dup 1 and >r 2/ swap 2/ [ 1 8 cells 1- lshift 1- ] Literal and
  793:  r> IF  [ 1 8 cells 1- lshift ] Literal + THEN  swap ;
  794: 
  795: and	w1 w2 -- w		core
  796: w = w1&w2;
  797: 
  798: or	w1 w2 -- w		core
  799: w = w1|w2;
  800: :
  801:  invert swap invert and invert ;
  802: 
  803: xor	w1 w2 -- w		core
  804: w = w1^w2;
  805: 
  806: invert	w1 -- w2		core
  807: w2 = ~w1;
  808: :
  809:  MAXU xor ;
  810: 
  811: rshift	u1 n -- u2		core
  812:   u2 = u1>>n;
  813: :
  814:     0 ?DO 2/ MAXI and LOOP ;
  815: 
  816: lshift	u1 n -- u2		core
  817:   u2 = u1<<n;
  818: :
  819:     0 ?DO 2* LOOP ;
  820: 
  821: \ comparisons(prefix, args, prefix, arg1, arg2, wordsets...)
  822: define(comparisons,
  823: $1=	$2 -- f		$6	$3equals
  824: f = FLAG($4==$5);
  825: :
  826:     [ char $1x char 0 = [IF]
  827: 	] IF false ELSE true THEN [
  828:     [ELSE]
  829: 	] xor 0= [
  830:     [THEN] ] ;
  831: 
  832: $1<>	$2 -- f		$7	$3different
  833: f = FLAG($4!=$5);
  834: :
  835:     [ char $1x char 0 = [IF]
  836: 	] IF true ELSE false THEN [
  837:     [ELSE]
  838: 	] xor 0<> [
  839:     [THEN] ] ;
  840: 
  841: $1<	$2 -- f		$8	$3less
  842: f = FLAG($4<$5);
  843: :
  844:     [ char $1x char 0 = [IF]
  845: 	] MINI and 0<> [
  846:     [ELSE] char $1x char u = [IF]
  847: 	]   2dup xor 0<  IF nip ELSE - THEN 0<  [
  848: 	[ELSE]
  849: 	    ] MINI xor >r MINI xor r> u< [
  850: 	[THEN]
  851:     [THEN] ] ;
  852: 
  853: $1>	$2 -- f		$9	$3greater
  854: f = FLAG($4>$5);
  855: :
  856:     [ char $1x char 0 = [IF] ] negate [ [ELSE] ] swap [ [THEN] ]
  857:     $1< ;
  858: 
  859: $1<=	$2 -- f		gforth	$3less_or_equal
  860: f = FLAG($4<=$5);
  861: :
  862:     $1> 0= ;
  863: 
  864: $1>=	$2 -- f		gforth	$3greater_or_equal
  865: f = FLAG($4>=$5);
  866: :
  867:     [ char $1x char 0 = [IF] ] negate [ [ELSE] ] swap [ [THEN] ]
  868:     $1<= ;
  869: 
  870: )
  871: 
  872: comparisons(0, n, zero_, n, 0, core, core-ext, core, core-ext)
  873: comparisons(, n1 n2, , n1, n2, core, core-ext, core, core)
  874: comparisons(u, u1 u2, u_, u1, u2, gforth, gforth, core, core-ext)
  875: 
  876: \ dcomparisons(prefix, args, prefix, arg1, arg2, wordsets...)
  877: define(dcomparisons,
  878: $1=	$2 -- f		$6	$3equals
  879: #ifdef BUGGY_LONG_LONG
  880: f = FLAG($4.lo==$5.lo && $4.hi==$5.hi);
  881: #else
  882: f = FLAG($4==$5);
  883: #endif
  884: 
  885: $1<>	$2 -- f		$7	$3different
  886: #ifdef BUGGY_LONG_LONG
  887: f = FLAG($4.lo!=$5.lo || $4.hi!=$5.hi);
  888: #else
  889: f = FLAG($4!=$5);
  890: #endif
  891: 
  892: $1<	$2 -- f		$8	$3less
  893: #ifdef BUGGY_LONG_LONG
  894: f = FLAG($4.hi==$5.hi ? $4.lo<$5.lo : $4.hi<$5.hi);
  895: #else
  896: f = FLAG($4<$5);
  897: #endif
  898: 
  899: $1>	$2 -- f		$9	$3greater
  900: #ifdef BUGGY_LONG_LONG
  901: f = FLAG($4.hi==$5.hi ? $4.lo>$5.lo : $4.hi>$5.hi);
  902: #else
  903: f = FLAG($4>$5);
  904: #endif
  905: 
  906: $1<=	$2 -- f		gforth	$3less_or_equal
  907: #ifdef BUGGY_LONG_LONG
  908: f = FLAG($4.hi==$5.hi ? $4.lo<=$5.lo : $4.hi<=$5.hi);
  909: #else
  910: f = FLAG($4<=$5);
  911: #endif
  912: 
  913: $1>=	$2 -- f		gforth	$3greater_or_equal
  914: #ifdef BUGGY_LONG_LONG
  915: f = FLAG($4.hi==$5.hi ? $4.lo>=$5.lo : $4.hi>=$5.hi);
  916: #else
  917: f = FLAG($4>=$5);
  918: #endif
  919: 
  920: )
  921: 
  922: \+dcomps
  923: 
  924: dcomparisons(d, d1 d2, d_, d1, d2, double, gforth, double, gforth)
  925: dcomparisons(d0, d, d_zero_, d, DZERO, double, gforth, double, gforth)
  926: dcomparisons(du, ud1 ud2, d_u_, ud1, ud2, gforth, gforth, double-ext, gforth)
  927: 
  928: \+
  929: 
  930: within	u1 u2 u3 -- f		core-ext
  931: f = FLAG(u1-u2 < u3-u2);
  932: :
  933:  over - >r - r> u< ;
  934: 
  935: sp@	-- a_addr		gforth		sp_fetch
  936: a_addr = sp+1;
  937: 
  938: sp!	a_addr --		gforth		sp_store
  939: sp = a_addr;
  940: /* works with and without TOS caching */
  941: 
  942: rp@	-- a_addr		gforth		rp_fetch
  943: a_addr = rp;
  944: 
  945: rp!	a_addr --		gforth		rp_store
  946: rp = a_addr;
  947: 
  948: \+floating
  949: 
  950: fp@	-- f_addr	gforth	fp_fetch
  951: f_addr = fp;
  952: 
  953: fp!	f_addr --	gforth	fp_store
  954: fp = f_addr;
  955: 
  956: \+
  957: 
  958: ;s	--		gforth	semis
  959: ""The primitive compiled by @code{EXIT}.""
  960: SET_IP((Xt *)(*rp++));
  961: 
  962: >r	w --		core	to_r
  963: *--rp = w;
  964: :
  965:  (>r) ;
  966: : (>r)  rp@ cell+ @ rp@ ! rp@ cell+ ! ;
  967: 
  968: r>	-- w		core	r_from
  969: w = *rp++;
  970: :
  971:  rp@ cell+ @ rp@ @ rp@ cell+ ! (rdrop) rp@ ! ;
  972: Create (rdrop) ' ;s A,
  973: 
  974: rdrop	--		gforth
  975: rp++;
  976: :
  977:  r> r> drop >r ;
  978: 
  979: 2>r	w1 w2 --	core-ext	two_to_r
  980: *--rp = w1;
  981: *--rp = w2;
  982: :
  983:  swap r> swap >r swap >r >r ;
  984: 
  985: 2r>	-- w1 w2	core-ext	two_r_from
  986: w2 = *rp++;
  987: w1 = *rp++;
  988: :
  989:  r> r> swap r> swap >r swap ;
  990: 
  991: 2r@	-- w1 w2	core-ext	two_r_fetch
  992: w2 = rp[0];
  993: w1 = rp[1];
  994: :
  995:  i' j ;
  996: 
  997: 2rdrop	--		gforth	two_r_drop
  998: rp+=2;
  999: :
 1000:  r> r> drop r> drop >r ;
 1001: 
 1002: over	w1 w2 -- w1 w2 w1		core
 1003: :
 1004:  sp@ cell+ @ ;
 1005: 
 1006: drop	w --		core
 1007: :
 1008:  IF THEN ;
 1009: 
 1010: swap	w1 w2 -- w2 w1		core
 1011: :
 1012:  >r (swap) ! r> (swap) @ ;
 1013: Variable (swap)
 1014: 
 1015: dup	w -- w w		core
 1016: :
 1017:  sp@ @ ;
 1018: 
 1019: rot	w1 w2 w3 -- w2 w3 w1	core	rote
 1020: :
 1021: [ defined? (swap) [IF] ]
 1022:     (swap) ! (rot) ! >r (rot) @ (swap) @ r> ;
 1023: Variable (rot)
 1024: [ELSE] ]
 1025:     >r swap r> swap ;
 1026: [THEN]
 1027: 
 1028: -rot	w1 w2 w3 -- w3 w1 w2	gforth	not_rote
 1029: :
 1030:  rot rot ;
 1031: 
 1032: nip	w1 w2 -- w2		core-ext
 1033: :
 1034:  swap drop ;
 1035: 
 1036: tuck	w1 w2 -- w2 w1 w2	core-ext
 1037: :
 1038:  swap over ;
 1039: 
 1040: ?dup	w -- w			core	question_dupe
 1041: if (w!=0) {
 1042:   IF_TOS(*sp-- = w;)
 1043: #ifndef USE_TOS
 1044:   *--sp = w;
 1045: #endif
 1046: }
 1047: :
 1048:  dup IF dup THEN ;
 1049: 
 1050: pick	u -- w			core-ext
 1051: w = sp[u+1];
 1052: :
 1053:  1+ cells sp@ + @ ;
 1054: 
 1055: 2drop	w1 w2 --		core	two_drop
 1056: :
 1057:  drop drop ;
 1058: 
 1059: 2dup	w1 w2 -- w1 w2 w1 w2	core	two_dupe
 1060: :
 1061:  over over ;
 1062: 
 1063: 2over	w1 w2 w3 w4 -- w1 w2 w3 w4 w1 w2	core	two_over
 1064: :
 1065:  3 pick 3 pick ;
 1066: 
 1067: 2swap	w1 w2 w3 w4 -- w3 w4 w1 w2	core	two_swap
 1068: :
 1069:  rot >r rot r> ;
 1070: 
 1071: 2rot	w1 w2 w3 w4 w5 w6 -- w3 w4 w5 w6 w1 w2	double-ext	two_rote
 1072: :
 1073:  >r >r 2swap r> r> 2swap ;
 1074: 
 1075: 2nip	w1 w2 w3 w4 -- w3 w4	gforth	two_nip
 1076: :
 1077:  2swap 2drop ;
 1078: 
 1079: 2tuck	w1 w2 w3 w4 -- w3 w4 w1 w2 w3 w4	gforth	two_tuck
 1080: :
 1081:  2swap 2over ;
 1082: 
 1083: \ toggle is high-level: 0.11/0.42%
 1084: 
 1085: @	a_addr -- w		core	fetch
 1086: "" Read from the cell at address @i{a-addr}, and return its contents, @i{w}.""
 1087: w = *a_addr;
 1088: 
 1089: !	w a_addr --		core	store
 1090: "" Write the value @i{w} to the cell at address @i{a-addr}.""
 1091: *a_addr = w;
 1092: 
 1093: +!	n a_addr --		core	plus_store
 1094: "" Add @i{n} to the value stored in the cell at address @i{a-addr}.""
 1095: *a_addr += n;
 1096: :
 1097:  tuck @ + swap ! ;
 1098: 
 1099: c@	c_addr -- c		core	c_fetch
 1100: "" Read from the char at address @i{c-addr}, and return its contents, @i{c}.""
 1101: c = *c_addr;
 1102: :
 1103: [ bigendian [IF] ]
 1104:     [ cell>bit 4 = [IF] ]
 1105: 	dup [ 0 cell - ] Literal and @ swap 1 and
 1106: 	IF  $FF and  ELSE  8>>  THEN  ;
 1107:     [ [ELSE] ]
 1108: 	dup [ cell 1- ] literal and
 1109: 	tuck - @ swap [ cell 1- ] literal xor
 1110:  	0 ?DO 8>> LOOP $FF and
 1111:     [ [THEN] ]
 1112: [ [ELSE] ]
 1113:     [ cell>bit 4 = [IF] ]
 1114: 	dup [ 0 cell - ] Literal and @ swap 1 and
 1115: 	IF  8>>  ELSE  $FF and  THEN
 1116:     [ [ELSE] ]
 1117: 	dup [ cell  1- ] literal and 
 1118: 	tuck - @ swap
 1119: 	0 ?DO 8>> LOOP 255 and
 1120:     [ [THEN] ]
 1121: [ [THEN] ]
 1122: ;
 1123: : 8>> 2/ 2/ 2/ 2/  2/ 2/ 2/ 2/ ;
 1124: 
 1125: c!	c c_addr --		core	c_store
 1126: "" Write the value @i{c} to the char at address @i{c-addr}.""
 1127: *c_addr = c;
 1128: :
 1129: [ bigendian [IF] ]
 1130:     [ cell>bit 4 = [IF] ]
 1131: 	tuck 1 and IF  $FF and  ELSE  8<<  THEN >r
 1132: 	dup -2 and @ over 1 and cells masks + @ and
 1133: 	r> or swap -2 and ! ;
 1134: 	Create masks $00FF , $FF00 ,
 1135:     [ELSE] ]
 1136: 	dup [ cell 1- ] literal and dup 
 1137: 	[ cell 1- ] literal xor >r
 1138: 	- dup @ $FF r@ 0 ?DO 8<< LOOP invert and
 1139: 	rot $FF and r> 0 ?DO 8<< LOOP or swap ! ;
 1140:     [THEN]
 1141: [ELSE] ]
 1142:     [ cell>bit 4 = [IF] ]
 1143: 	tuck 1 and IF  8<<  ELSE  $FF and  THEN >r
 1144: 	dup -2 and @ over 1 and cells masks + @ and
 1145: 	r> or swap -2 and ! ;
 1146: 	Create masks $FF00 , $00FF ,
 1147:     [ELSE] ]
 1148: 	dup [ cell 1- ] literal and dup >r
 1149: 	- dup @ $FF r@ 0 ?DO 8<< LOOP invert and
 1150: 	rot $FF and r> 0 ?DO 8<< LOOP or swap ! ;
 1151:     [THEN]
 1152: [THEN]
 1153: : 8<< 2* 2* 2* 2*  2* 2* 2* 2* ;
 1154: 
 1155: 2!	w1 w2 a_addr --		core	two_store
 1156: "" Write the value @i{w1, w2} to the double at address @i{a-addr}.""
 1157: a_addr[0] = w2;
 1158: a_addr[1] = w1;
 1159: :
 1160:  tuck ! cell+ ! ;
 1161: 
 1162: 2@	a_addr -- w1 w2		core	two_fetch
 1163: "" Read from the double at address @i{a-addr}, and return its contents, @i{w1, w2}.""
 1164: w2 = a_addr[0];
 1165: w1 = a_addr[1];
 1166: :
 1167:  dup cell+ @ swap @ ;
 1168: 
 1169: cell+	a_addr1 -- a_addr2	core	cell_plus
 1170: "" Increment @i{a-addr1} by the number of address units corresponding to the size of
 1171: one cell, to give @i{a-addr2}.""
 1172: a_addr2 = a_addr1+1;
 1173: :
 1174:  cell + ;
 1175: 
 1176: cells	n1 -- n2		core
 1177: "" @i{n2} is the number of address units corresponding to @i{n1} cells.""
 1178: n2 = n1 * sizeof(Cell);
 1179: :
 1180:  [ cell
 1181:  2/ dup [IF] ] 2* [ [THEN]
 1182:  2/ dup [IF] ] 2* [ [THEN]
 1183:  2/ dup [IF] ] 2* [ [THEN]
 1184:  2/ dup [IF] ] 2* [ [THEN]
 1185:  drop ] ;
 1186: 
 1187: char+	c_addr1 -- c_addr2	core	char_plus
 1188: "" Increment @i{c-addr1} by the number of address units corresponding to the size of
 1189: one char, to give @i{c-addr2}.""
 1190: c_addr2 = c_addr1 + 1;
 1191: :
 1192:  1+ ;
 1193: 
 1194: (chars)		n1 -- n2	gforth	paren_chars
 1195: n2 = n1 * sizeof(Char);
 1196: :
 1197:  ;
 1198: 
 1199: count	c_addr1 -- c_addr2 u	core
 1200: "" If @i{c-add1} is the address of a counted string return the length of
 1201: the string, @i{u}, and the address of its first character, @i{c-addr2}.""
 1202: u = *c_addr1;
 1203: c_addr2 = c_addr1+1;
 1204: :
 1205:  dup 1+ swap c@ ;
 1206: 
 1207: (f83find)	c_addr u f83name1 -- f83name2	new	paren_f83find
 1208: for (; f83name1 != NULL; f83name1 = (struct F83Name *)(f83name1->next))
 1209:   if ((UCell)F83NAME_COUNT(f83name1)==u &&
 1210:       memcasecmp(c_addr, f83name1->name, u)== 0 /* or inline? */)
 1211:     break;
 1212: f83name2=f83name1;
 1213: :
 1214:     BEGIN  dup WHILE  (find-samelen)  dup  WHILE
 1215: 	>r 2dup r@ cell+ char+ capscomp  0=
 1216: 	IF  2drop r>  EXIT  THEN
 1217: 	r> @
 1218:     REPEAT  THEN  nip nip ;
 1219: : (find-samelen) ( u f83name1 -- u f83name2/0 )
 1220:     BEGIN  2dup cell+ c@ $1F and <> WHILE  @  dup 0= UNTIL  THEN ;
 1221: 
 1222: \+hash
 1223: 
 1224: (hashfind)	c_addr u a_addr -- f83name2	new	paren_hashfind
 1225: struct F83Name *f83name1;
 1226: f83name2=NULL;
 1227: while(a_addr != NULL)
 1228: {
 1229:    f83name1=(struct F83Name *)(a_addr[1]);
 1230:    a_addr=(Cell *)(a_addr[0]);
 1231:    if ((UCell)F83NAME_COUNT(f83name1)==u &&
 1232:        memcasecmp(c_addr, f83name1->name, u)== 0 /* or inline? */)
 1233:      {
 1234: 	f83name2=f83name1;
 1235: 	break;
 1236:      }
 1237: }
 1238: :
 1239:  BEGIN  dup  WHILE
 1240:         2@ >r >r dup r@ cell+ c@ $1F and =
 1241:         IF  2dup r@ cell+ char+ capscomp 0=
 1242: 	    IF  2drop r> rdrop  EXIT  THEN  THEN
 1243: 	rdrop r>
 1244:  REPEAT nip nip ;
 1245: 
 1246: (tablefind)	c_addr u a_addr -- f83name2	new	paren_tablefind
 1247: ""A case-sensitive variant of @code{(hashfind)}""
 1248: struct F83Name *f83name1;
 1249: f83name2=NULL;
 1250: while(a_addr != NULL)
 1251: {
 1252:    f83name1=(struct F83Name *)(a_addr[1]);
 1253:    a_addr=(Cell *)(a_addr[0]);
 1254:    if ((UCell)F83NAME_COUNT(f83name1)==u &&
 1255:        memcmp(c_addr, f83name1->name, u)== 0 /* or inline? */)
 1256:      {
 1257: 	f83name2=f83name1;
 1258: 	break;
 1259:      }
 1260: }
 1261: :
 1262:  BEGIN  dup  WHILE
 1263:         2@ >r >r dup r@ cell+ c@ $1F and =
 1264:         IF  2dup r@ cell+ char+ -text 0=
 1265: 	    IF  2drop r> rdrop  EXIT  THEN  THEN
 1266: 	rdrop r>
 1267:  REPEAT nip nip ;
 1268: 
 1269: (hashkey)	c_addr u1 -- u2		gforth	paren_hashkey
 1270: u2=0;
 1271: while(u1--)
 1272:    u2+=(Cell)toupper(*c_addr++);
 1273: :
 1274:  0 -rot bounds ?DO  I c@ toupper +  LOOP ;
 1275: 
 1276: (hashkey1)	c_addr u ubits -- ukey		gforth	paren_hashkey1
 1277: ""ukey is the hash key for the string c_addr u fitting in ubits bits""
 1278: /* this hash function rotates the key at every step by rot bits within
 1279:    ubits bits and xors it with the character. This function does ok in
 1280:    the chi-sqare-test.  Rot should be <=7 (preferably <=5) for
 1281:    ASCII strings (larger if ubits is large), and should share no
 1282:    divisors with ubits.
 1283: */
 1284: unsigned rot = ((char []){5,0,1,2,3,4,5,5,5,5,3,5,5,5,5,7,5,5,5,5,7,5,5,5,5,6,5,5,5,5,7,5,5})[ubits];
 1285: Char *cp = c_addr;
 1286: for (ukey=0; cp<c_addr+u; cp++)
 1287:     ukey = ((((ukey<<rot) | (ukey>>(ubits-rot))) 
 1288: 	     ^ toupper(*cp))
 1289: 	    & ((1<<ubits)-1));
 1290: :
 1291:  dup rot-values + c@ over 1 swap lshift 1- >r
 1292:  tuck - 2swap r> 0 2swap bounds
 1293:  ?DO  dup 4 pick lshift swap 3 pick rshift or
 1294:       I c@ toupper xor
 1295:       over and  LOOP
 1296:  nip nip nip ;
 1297: Create rot-values
 1298:   5 c, 0 c, 1 c, 2 c, 3 c,  4 c, 5 c, 5 c, 5 c, 5 c,
 1299:   3 c, 5 c, 5 c, 5 c, 5 c,  7 c, 5 c, 5 c, 5 c, 5 c,
 1300:   7 c, 5 c, 5 c, 5 c, 5 c,  6 c, 5 c, 5 c, 5 c, 5 c,
 1301:   7 c, 5 c, 5 c,
 1302: 
 1303: \+
 1304: 
 1305: (parse-white)	c_addr1 u1 -- c_addr2 u2	gforth	paren_parse_white
 1306: /* use !isgraph instead of isspace? */
 1307: Char *endp = c_addr1+u1;
 1308: while (c_addr1<endp && isspace(*c_addr1))
 1309:   c_addr1++;
 1310: if (c_addr1<endp) {
 1311:   for (c_addr2 = c_addr1; c_addr1<endp && !isspace(*c_addr1); c_addr1++)
 1312:     ;
 1313:   u2 = c_addr1-c_addr2;
 1314: }
 1315: else {
 1316:   c_addr2 = c_addr1;
 1317:   u2 = 0;
 1318: }
 1319: :
 1320:  BEGIN  dup  WHILE  over c@ bl <=  WHILE  1 /string
 1321:  REPEAT  THEN  2dup
 1322:  BEGIN  dup  WHILE  over c@ bl >   WHILE  1 /string
 1323:  REPEAT  THEN  nip - ;
 1324: 
 1325: aligned		c_addr -- a_addr	core
 1326: "" @i{a-addr} is the first aligned address greater than or equal to @i{c-addr}.""
 1327: a_addr = (Cell *)((((Cell)c_addr)+(sizeof(Cell)-1))&(-sizeof(Cell)));
 1328: :
 1329:  [ cell 1- ] Literal + [ -1 cells ] Literal and ;
 1330: 
 1331: faligned	c_addr -- f_addr	float	f_aligned
 1332: "" @i{f-addr} is the first float-aligned address greater than or equal to @i{c-addr}.""
 1333: f_addr = (Float *)((((Cell)c_addr)+(sizeof(Float)-1))&(-sizeof(Float)));
 1334: :
 1335:  [ 1 floats 1- ] Literal + [ -1 floats ] Literal and ;
 1336: 
 1337: >body		xt -- a_addr	core	to_body
 1338: a_addr = PFA(xt);
 1339: :
 1340:     2 cells + ;
 1341: 
 1342: \+standardthreading
 1343: 
 1344: >code-address		xt -- c_addr		gforth	to_code_address
 1345: ""@i{c-addr} is the code address of the word @i{xt}.""
 1346: /* !! This behaves installation-dependently for DOES-words */
 1347: c_addr = (Address)CODE_ADDRESS(xt);
 1348: :
 1349:     @ ;
 1350: 
 1351: >does-code	xt -- a_addr		gforth	to_does_code
 1352: ""If @i{xt} is the execution token of a defining-word-defined word,
 1353: @i{a-addr} is the start of the Forth code after the @code{DOES>};
 1354: Otherwise @i{a-addr} is 0.""
 1355: a_addr = (Cell *)DOES_CODE(xt);
 1356: :
 1357:     cell+ @ ;
 1358: 
 1359: code-address!		c_addr xt --		gforth	code_address_store
 1360: ""Create a code field with code address @i{c-addr} at @i{xt}.""
 1361: MAKE_CF(xt, c_addr);
 1362: CACHE_FLUSH(xt,(size_t)PFA(0));
 1363: :
 1364:     ! ;
 1365: 
 1366: does-code!	a_addr xt --		gforth	does_code_store
 1367: ""Create a code field at @i{xt} for a defining-word-defined word; @i{a-addr}
 1368: is the start of the Forth code after @code{DOES>}.""
 1369: MAKE_DOES_CF(xt, a_addr);
 1370: CACHE_FLUSH(xt,(size_t)PFA(0));
 1371: :
 1372:     dodoes: over ! cell+ ! ;
 1373: 
 1374: does-handler!	a_addr --	gforth	does_handler_store
 1375: ""Create a @code{DOES>}-handler at address @i{a-addr}. Usually, @i{a-addr} points
 1376: just behind a @code{DOES>}.""
 1377: MAKE_DOES_HANDLER(a_addr);
 1378: CACHE_FLUSH((caddr_t)a_addr,DOES_HANDLER_SIZE);
 1379: :
 1380:     drop ;
 1381: 
 1382: /does-handler	-- n	gforth	slash_does_handler
 1383: ""The size of a @code{DOES>}-handler (includes possible padding).""
 1384: /* !! a constant or environmental query might be better */
 1385: n = DOES_HANDLER_SIZE;
 1386: :
 1387:     2 cells ;
 1388: 
 1389: threading-method	-- n	gforth	threading_method
 1390: ""0 if the engine is direct threaded. Note that this may change during
 1391: the lifetime of an image.""
 1392: #if defined(DOUBLY_INDIRECT)
 1393: n=2;
 1394: #else
 1395: # if defined(DIRECT_THREADED)
 1396: n=0;
 1397: # else
 1398: n=1;
 1399: # endif
 1400: #endif
 1401: :
 1402:  1 ;
 1403: 
 1404: \+
 1405: 
 1406: key-file	wfileid -- n		gforth	paren_key_file
 1407: #ifdef HAS_FILE
 1408: fflush(stdout);
 1409: n = key((FILE*)wfileid);
 1410: #else
 1411: n = key(stdin);
 1412: #endif
 1413: 
 1414: key?-file	wfileid -- n		facility	key_q_file
 1415: #ifdef HAS_FILE
 1416: fflush(stdout);
 1417: n = key_query((FILE*)wfileid);
 1418: #else
 1419: n = key_query(stdin);
 1420: #endif
 1421: 
 1422: \+os
 1423: 
 1424: stdin	-- wfileid	gforth
 1425: wfileid = (Cell)stdin;
 1426: 
 1427: stdout	-- wfileid	gforth
 1428: wfileid = (Cell)stdout;
 1429: 
 1430: stderr	-- wfileid	gforth
 1431: wfileid = (Cell)stderr;
 1432: 
 1433: form	-- urows ucols	gforth
 1434: ""The number of lines and columns in the terminal. These numbers may change
 1435: with the window size.""
 1436: /* we could block SIGWINCH here to get a consistent size, but I don't
 1437:  think this is necessary or always beneficial */
 1438: urows=rows;
 1439: ucols=cols;
 1440: 
 1441: flush-icache	c_addr u --	gforth	flush_icache
 1442: ""Make sure that the instruction cache of the processor (if there is
 1443: one) does not contain stale data at @i{c-addr} and @i{u} bytes
 1444: afterwards. @code{END-CODE} performs a @code{flush-icache}
 1445: automatically. Caveat: @code{flush-icache} might not work on your
 1446: installation; this is usually the case if direct threading is not
 1447: supported on your machine (take a look at your @file{machine.h}) and
 1448: your machine has a separate instruction cache. In such cases,
 1449: @code{flush-icache} does nothing instead of flushing the instruction
 1450: cache.""
 1451: FLUSH_ICACHE(c_addr,u);
 1452: 
 1453: (bye)	n --	gforth	paren_bye
 1454: return (Label *)n;
 1455: 
 1456: (system)	c_addr u -- wretval wior	gforth	peren_system
 1457: #ifndef MSDOS
 1458: int old_tp=terminal_prepped;
 1459: deprep_terminal();
 1460: #endif
 1461: wretval=system(cstr(c_addr,u,1)); /* ~ expansion on first part of string? */
 1462: wior = IOR(wretval==-1 || (wretval==127 && errno != 0));
 1463: #ifndef MSDOS
 1464: if (old_tp)
 1465:   prep_terminal();
 1466: #endif
 1467: 
 1468: getenv	c_addr1 u1 -- c_addr2 u2	gforth
 1469: ""The string @i{c-addr1 u1} specifies an environment variable. The string @i{c-addr2 u2}
 1470: is the host operating system's expansion of that environment variable. If the
 1471: environment variable does not exist, @i{c-addr2 u2} specifies a string 0 characters
 1472: in length.""
 1473: c_addr2 = getenv(cstr(c_addr1,u1,1));
 1474: u2 = (c_addr2 == NULL ? 0 : strlen(c_addr2));
 1475: 
 1476: open-pipe	c_addr u ntype -- wfileid wior	gforth	open_pipe
 1477: wfileid=(Cell)popen(cstr(c_addr,u,1),fileattr[ntype]); /* ~ expansion of 1st arg? */
 1478: wior = IOR(wfileid==0); /* !! the man page says that errno is not set reliably */
 1479: 
 1480: close-pipe	wfileid -- wretval wior		gforth	close_pipe
 1481: wretval = pclose((FILE *)wfileid);
 1482: wior = IOR(wretval==-1);
 1483: 
 1484: time&date	-- nsec nmin nhour nday nmonth nyear	facility-ext	time_and_date
 1485: struct timeval time1;
 1486: struct timezone zone1;
 1487: struct tm *ltime;
 1488: gettimeofday(&time1,&zone1);
 1489: ltime=localtime((time_t *)&time1.tv_sec);
 1490: nyear =ltime->tm_year+1900;
 1491: nmonth=ltime->tm_mon+1;
 1492: nday  =ltime->tm_mday;
 1493: nhour =ltime->tm_hour;
 1494: nmin  =ltime->tm_min;
 1495: nsec  =ltime->tm_sec;
 1496: 
 1497: ms	n --	facility-ext
 1498: struct timeval timeout;
 1499: timeout.tv_sec=n/1000;
 1500: timeout.tv_usec=1000*(n%1000);
 1501: (void)select(0,0,0,0,&timeout);
 1502: 
 1503: allocate	u -- a_addr wior	memory
 1504: ""Allocate @i{u} address units of contiguous data space. The initial
 1505: contents of the data space is undefined. If the allocation is successful,
 1506: @i{a-addr} is the start address of the allocated region and @i{wior}
 1507: is 0. If the allocation fails, @i{a-addr} is undefined and @i{wior}
 1508: is an implementation-defined I/O result code.""
 1509: a_addr = (Cell *)malloc(u?u:1);
 1510: wior = IOR(a_addr==NULL);
 1511: 
 1512: free		a_addr -- wior		memory
 1513: ""Return the region of data space starting at @i{a-addr} to the system.
 1514: The regon must originally have been obtained using @code{allocate} or
 1515: @code{resize}. If the operational is successful, @i{wior} is 0.
 1516: If the operation fails, @i{wior} is an implementation-defined
 1517: I/O result code.""
 1518: free(a_addr);
 1519: wior = 0;
 1520: 
 1521: resize		a_addr1 u -- a_addr2 wior	memory
 1522: ""Change the size of the allocated area at @i{a-addr1} to @i{u}
 1523: address units, possibly moving the contents to a different
 1524: area. @i{a-addr2} is the address of the resulting area.
 1525: If the operational is successful, @i{wior} is 0.
 1526: If the operation fails, @i{wior} is an implementation-defined
 1527: I/O result code. If @i{a-addr1} is 0, Gforth's (but not the Standard)
 1528: @code{resize} @code{allocate}s @i{u} address units.""
 1529: /* the following check is not necessary on most OSs, but it is needed
 1530:    on SunOS 4.1.2. */
 1531: if (a_addr1==NULL)
 1532:   a_addr2 = (Cell *)malloc(u);
 1533: else
 1534:   a_addr2 = (Cell *)realloc(a_addr1, u);
 1535: wior = IOR(a_addr2==NULL);	/* !! Define a return code */
 1536: 
 1537: strerror	n -- c_addr u	gforth
 1538: c_addr = strerror(n);
 1539: u = strlen(c_addr);
 1540: 
 1541: strsignal	n -- c_addr u	gforth
 1542: c_addr = strsignal(n);
 1543: u = strlen(c_addr);
 1544: 
 1545: call-c	w --	gforth	call_c
 1546: ""Call the C function pointed to by @i{w}. The C function has to
 1547: access the stack itself. The stack pointers are exported in the global
 1548: variables @code{SP} and @code{FP}.""
 1549: /* This is a first attempt at support for calls to C. This may change in
 1550:    the future */
 1551: IF_FTOS(fp[0]=FTOS);
 1552: FP=fp;
 1553: SP=sp;
 1554: ((void (*)())w)();
 1555: sp=SP;
 1556: fp=FP;
 1557: IF_TOS(TOS=sp[0]);
 1558: IF_FTOS(FTOS=fp[0]);
 1559: 
 1560: \+
 1561: \+file
 1562: 
 1563: close-file	wfileid -- wior		file	close_file
 1564: wior = IOR(fclose((FILE *)wfileid)==EOF);
 1565: 
 1566: open-file	c_addr u ntype -- wfileid wior	file	open_file
 1567: wfileid = (Cell)fopen(tilde_cstr(c_addr, u, 1), fileattr[ntype]);
 1568: #if defined(GO32) && defined(MSDOS)
 1569: if(wfileid && !(ntype & 1))
 1570:   setbuf((FILE*)wfileid, NULL);
 1571: #endif
 1572: wior =  IOR(wfileid == 0);
 1573: 
 1574: create-file	c_addr u ntype -- wfileid wior	file	create_file
 1575: Cell	fd;
 1576: fd = open(tilde_cstr(c_addr, u, 1), O_CREAT|O_TRUNC|ufileattr[ntype], 0666);
 1577: if (fd != -1) {
 1578:   wfileid = (Cell)fdopen(fd, fileattr[ntype]);
 1579: #if defined(GO32) && defined(MSDOS)
 1580:   if(wfileid && !(ntype & 1))
 1581:     setbuf((FILE*)wfileid, NULL);
 1582: #endif
 1583:   wior = IOR(wfileid == 0);
 1584: } else {
 1585:   wfileid = 0;
 1586:   wior = IOR(1);
 1587: }
 1588: 
 1589: delete-file	c_addr u -- wior		file	delete_file
 1590: wior = IOR(unlink(tilde_cstr(c_addr, u, 1))==-1);
 1591: 
 1592: rename-file	c_addr1 u1 c_addr2 u2 -- wior	file-ext	rename_file
 1593: ""Rename file @i{c_addr1 u1} to new name @i{c_addr2 u2}""
 1594: char *s1=tilde_cstr(c_addr2, u2, 1);
 1595: wior = IOR(rename(tilde_cstr(c_addr1, u1, 0), s1)==-1);
 1596: 
 1597: file-position	wfileid -- ud wior	file	file_position
 1598: /* !! use tell and lseek? */
 1599: ud = LONG2UD(ftell((FILE *)wfileid));
 1600: wior = IOR(UD2LONG(ud)==-1);
 1601: 
 1602: reposition-file	ud wfileid -- wior	file	reposition_file
 1603: wior = IOR(fseek((FILE *)wfileid, UD2LONG(ud), SEEK_SET)==-1);
 1604: 
 1605: file-size	wfileid -- ud wior	file	file_size
 1606: struct stat buf;
 1607: wior = IOR(fstat(fileno((FILE *)wfileid), &buf)==-1);
 1608: ud = LONG2UD(buf.st_size);
 1609: 
 1610: resize-file	ud wfileid -- wior	file	resize_file
 1611: wior = IOR(ftruncate(fileno((FILE *)wfileid), UD2LONG(ud))==-1);
 1612: 
 1613: read-file	c_addr u1 wfileid -- u2 wior	file	read_file
 1614: /* !! fread does not guarantee enough */
 1615: u2 = fread(c_addr, sizeof(Char), u1, (FILE *)wfileid);
 1616: wior = FILEIO(u2<u1 && ferror((FILE *)wfileid));
 1617: /* !! is the value of ferror errno-compatible? */
 1618: if (wior)
 1619:   clearerr((FILE *)wfileid);
 1620: 
 1621: read-line	c_addr u1 wfileid -- u2 flag wior	file	read_line
 1622: /*
 1623: Cell c;
 1624: flag=-1;
 1625: for(u2=0; u2<u1; u2++)
 1626: {
 1627:    *c_addr++ = (Char)(c = getc((FILE *)wfileid));
 1628:    if(c=='\n') break;
 1629:    if(c==EOF)
 1630:      {
 1631: 	flag=FLAG(u2!=0);
 1632: 	break;
 1633:      }
 1634: }
 1635: wior=FILEIO(ferror((FILE *)wfileid));
 1636: */
 1637: if ((flag=FLAG(!feof((FILE *)wfileid) &&
 1638: 	       fgets(c_addr,u1+1,(FILE *)wfileid) != NULL))) {
 1639:   wior=FILEIO(ferror((FILE *)wfileid)!=0); /* !! ior? */
 1640:   if (wior)
 1641:     clearerr((FILE *)wfileid);
 1642:   u2 = strlen(c_addr);
 1643:   u2-=((u2>0) && (c_addr[u2-1]==NEWLINE));
 1644: }
 1645: else {
 1646:   wior=0;
 1647:   u2=0;
 1648: }
 1649: 
 1650: \+
 1651: \+file
 1652: 
 1653: write-file	c_addr u1 wfileid -- wior	file	write_file
 1654: /* !! fwrite does not guarantee enough */
 1655: {
 1656:   UCell u2 = fwrite(c_addr, sizeof(Char), u1, (FILE *)wfileid);
 1657:   wior = FILEIO(u2<u1 && ferror((FILE *)wfileid));
 1658:   if (wior)
 1659:     clearerr((FILE *)wfileid);
 1660: }
 1661: 
 1662: \+
 1663: 
 1664: emit-file	c wfileid -- wior	gforth	emit_file
 1665: #ifdef HAS_FILE
 1666: wior = FILEIO(putc(c, (FILE *)wfileid)==EOF);
 1667: if (wior)
 1668:   clearerr((FILE *)wfileid);
 1669: #else
 1670: putc(c, stdout);
 1671: #endif
 1672: 
 1673: \+file
 1674: 
 1675: flush-file	wfileid -- wior		file-ext	flush_file
 1676: wior = IOR(fflush((FILE *) wfileid)==EOF);
 1677: 
 1678: file-status	c_addr u -- ntype wior	file-ext	file_status
 1679: char *filename=tilde_cstr(c_addr, u, 1);
 1680: if (access (filename, F_OK) != 0) {
 1681:   ntype=0;
 1682:   wior=IOR(1);
 1683: }
 1684: else if (access (filename, R_OK | W_OK) == 0) {
 1685:   ntype=2; /* r/w */
 1686:   wior=0;
 1687: }
 1688: else if (access (filename, R_OK) == 0) {
 1689:   ntype=0; /* r/o */
 1690:   wior=0;
 1691: }
 1692: else if (access (filename, W_OK) == 0) {
 1693:   ntype=4; /* w/o */
 1694:   wior=0;
 1695: }
 1696: else {
 1697:   ntype=1; /* well, we cannot access the file, but better deliver a legal
 1698: 	    access mode (r/o bin), so we get a decent error later upon open. */
 1699:   wior=0;
 1700: }
 1701: 
 1702: \+
 1703: \+floating
 1704: 
 1705: comparisons(f, r1 r2, f_, r1, r2, gforth, gforth, float, gforth)
 1706: comparisons(f0, r, f_zero_, r, 0., float, gforth, float, gforth)
 1707: 
 1708: d>f		d -- r		float	d_to_f
 1709: #ifdef BUGGY_LONG_LONG
 1710: extern double ldexp(double x, int exp);
 1711: r = ldexp((Float)d.hi,CELL_BITS) + (Float)d.lo;
 1712: #else
 1713: r = d;
 1714: #endif
 1715: 
 1716: f>d		r -- d		float	f_to_d
 1717: #ifdef BUGGY_LONG_LONG
 1718: d.hi = ldexp(r,-(int)(CELL_BITS)) - (r<0);
 1719: d.lo = r-ldexp((Float)d.hi,CELL_BITS);
 1720: #else
 1721: d = r;
 1722: #endif
 1723: 
 1724: f!		r f_addr --	float	f_store
 1725: "" Store the floating-point value @i{r} to address @i{f-addr}.""
 1726: *f_addr = r;
 1727: 
 1728: f@		f_addr -- r	float	f_fetch
 1729: "" Fetch floating-point value @i{r} from address @i{f-addr}.""
 1730: r = *f_addr;
 1731: 
 1732: df@		df_addr -- r	float-ext	d_f_fetch
 1733: "" Fetch the double-precision IEEE floating-point value @i{r} from the address @i{df-addr}.""
 1734: #ifdef IEEE_FP
 1735: r = *df_addr;
 1736: #else
 1737: !! df@
 1738: #endif
 1739: 
 1740: df!		r df_addr --	float-ext	d_f_store
 1741: "" Store the double-precision IEEE floating-point value @i{r} to the address @i{df-addr}.""
 1742: #ifdef IEEE_FP
 1743: *df_addr = r;
 1744: #else
 1745: !! df!
 1746: #endif
 1747: 
 1748: sf@		sf_addr -- r	float-ext	s_f_fetch
 1749: "" Fetch the single-precision IEEE floating-point value @i{r} from the address @i{sf-addr}.""
 1750: #ifdef IEEE_FP
 1751: r = *sf_addr;
 1752: #else
 1753: !! sf@
 1754: #endif
 1755: 
 1756: sf!		r sf_addr --	float-ext	s_f_store
 1757: "" Store the single-precision IEEE floating-point value @i{r} to the address @i{sf-addr}.""
 1758: #ifdef IEEE_FP
 1759: *sf_addr = r;
 1760: #else
 1761: !! sf!
 1762: #endif
 1763: 
 1764: f+		r1 r2 -- r3	float	f_plus
 1765: r3 = r1+r2;
 1766: 
 1767: f-		r1 r2 -- r3	float	f_minus
 1768: r3 = r1-r2;
 1769: 
 1770: f*		r1 r2 -- r3	float	f_star
 1771: r3 = r1*r2;
 1772: 
 1773: f/		r1 r2 -- r3	float	f_slash
 1774: r3 = r1/r2;
 1775: 
 1776: f**		r1 r2 -- r3	float-ext	f_star_star
 1777: ""@i{r3} is @i{r1} raised to the @i{r2}th power.""
 1778: r3 = pow(r1,r2);
 1779: 
 1780: fnegate		r1 -- r2	float
 1781: r2 = - r1;
 1782: 
 1783: fdrop		r --		float
 1784: 
 1785: fdup		r -- r r	float
 1786: 
 1787: fswap		r1 r2 -- r2 r1	float
 1788: 
 1789: fover		r1 r2 -- r1 r2 r1	float
 1790: 
 1791: frot		r1 r2 r3 -- r2 r3 r1	float
 1792: 
 1793: fnip		r1 r2 -- r2	gforth
 1794: 
 1795: ftuck		r1 r2 -- r2 r1 r2	gforth
 1796: 
 1797: float+		f_addr1 -- f_addr2	float	float_plus
 1798: "" Increment @i{f-addr1} by the number of address units corresponding to the size of
 1799: one floating-point number, to give @i{f-addr2}.""
 1800: f_addr2 = f_addr1+1;
 1801: 
 1802: floats		n1 -- n2	float
 1803: ""@i{n2} is the number of address units corresponding to @i{n1} floating-point numbers.""
 1804: n2 = n1*sizeof(Float);
 1805: 
 1806: floor		r1 -- r2	float
 1807: ""Round towards the next smaller integral value, i.e., round toward negative infinity.""
 1808: /* !! unclear wording */
 1809: r2 = floor(r1);
 1810: 
 1811: fround		r1 -- r2	float
 1812: ""Round to the nearest integral value.""
 1813: /* !! unclear wording */
 1814: #ifdef HAVE_RINT
 1815: r2 = rint(r1);
 1816: #else
 1817: r2 = floor(r1+0.5);
 1818: /* !! This is not quite true to the rounding rules given in the standard */
 1819: #endif
 1820: 
 1821: fmax		r1 r2 -- r3	float
 1822: if (r1<r2)
 1823:   r3 = r2;
 1824: else
 1825:   r3 = r1;
 1826: 
 1827: fmin		r1 r2 -- r3	float
 1828: if (r1<r2)
 1829:   r3 = r1;
 1830: else
 1831:   r3 = r2;
 1832: 
 1833: represent		r c_addr u -- n f1 f2	float
 1834: char *sig;
 1835: int flag;
 1836: int decpt;
 1837: sig=ecvt(r, u, &decpt, &flag);
 1838: n=(r==0 ? 1 : decpt);
 1839: f1=FLAG(flag!=0);
 1840: f2=FLAG(isdigit((unsigned)(sig[0]))!=0);
 1841: memmove(c_addr,sig,u);
 1842: 
 1843: >float	c_addr u -- flag	float	to_float
 1844: ""Attempt to convert the character string @i{c-addr u} to
 1845: internal floating-point representation. If the string
 1846: represents a valid floating-point number @i{r} is placed
 1847: on the floating-point stack and @i{flag} is true. Otherwise,
 1848: @i{flag} is false. A string of blanks is a special case
 1849: and represents the flotaing-point number 0.""
 1850: /* real signature: c_addr u -- r t / f */
 1851: Float r;
 1852: char *number=cstr(c_addr, u, 1);
 1853: char *endconv;
 1854: while(isspace((unsigned)(number[--u])) && u>0);
 1855: switch(number[u])
 1856: {
 1857:    case 'd':
 1858:    case 'D':
 1859:    case 'e':
 1860:    case 'E':  break;
 1861:    default :  u++; break;
 1862: }
 1863: number[u]='\0';
 1864: r=strtod(number,&endconv);
 1865: if((flag=FLAG(!(Cell)*endconv)))
 1866: {
 1867:    IF_FTOS(fp[0] = FTOS);
 1868:    fp += -1;
 1869:    FTOS = r;
 1870: }
 1871: else if(*endconv=='d' || *endconv=='D')
 1872: {
 1873:    *endconv='E';
 1874:    r=strtod(number,&endconv);
 1875:    if((flag=FLAG(!(Cell)*endconv)))
 1876:      {
 1877: 	IF_FTOS(fp[0] = FTOS);
 1878: 	fp += -1;
 1879: 	FTOS = r;
 1880:      }
 1881: }
 1882: 
 1883: fabs		r1 -- r2	float-ext
 1884: r2 = fabs(r1);
 1885: 
 1886: facos		r1 -- r2	float-ext
 1887: r2 = acos(r1);
 1888: 
 1889: fasin		r1 -- r2	float-ext
 1890: r2 = asin(r1);
 1891: 
 1892: fatan		r1 -- r2	float-ext
 1893: r2 = atan(r1);
 1894: 
 1895: fatan2		r1 r2 -- r3	float-ext
 1896: ""@i{r1/r2}=tan(@i{r3}). ANS Forth does not require, but probably
 1897: intends this to be the inverse of @code{fsincos}. In gforth it is.""
 1898: r3 = atan2(r1,r2);
 1899: 
 1900: fcos		r1 -- r2	float-ext
 1901: r2 = cos(r1);
 1902: 
 1903: fexp		r1 -- r2	float-ext
 1904: r2 = exp(r1);
 1905: 
 1906: fexpm1		r1 -- r2	float-ext
 1907: ""@i{r2}=@i{e}**@i{r1}@minus{}1""
 1908: #ifdef HAVE_EXPM1
 1909: extern double
 1910: #ifdef NeXT
 1911:               const
 1912: #endif
 1913:                     expm1(double);
 1914: r2 = expm1(r1);
 1915: #else
 1916: r2 = exp(r1)-1.;
 1917: #endif
 1918: 
 1919: fln		r1 -- r2	float-ext
 1920: r2 = log(r1);
 1921: 
 1922: flnp1		r1 -- r2	float-ext
 1923: ""@i{r2}=ln(@i{r1}+1)""
 1924: #ifdef HAVE_LOG1P
 1925: extern double
 1926: #ifdef NeXT
 1927:               const
 1928: #endif
 1929:                     log1p(double);
 1930: r2 = log1p(r1);
 1931: #else
 1932: r2 = log(r1+1.);
 1933: #endif
 1934: 
 1935: flog		r1 -- r2	float-ext
 1936: ""The decimal logarithm.""
 1937: r2 = log10(r1);
 1938: 
 1939: falog		r1 -- r2	float-ext
 1940: ""@i{r2}=10**@i{r1}""
 1941: extern double pow10(double);
 1942: r2 = pow10(r1);
 1943: 
 1944: fsin		r1 -- r2	float-ext
 1945: r2 = sin(r1);
 1946: 
 1947: fsincos		r1 -- r2 r3	float-ext
 1948: ""@i{r2}=sin(@i{r1}), @i{r3}=cos(@i{r1})""
 1949: r2 = sin(r1);
 1950: r3 = cos(r1);
 1951: 
 1952: fsqrt		r1 -- r2	float-ext
 1953: r2 = sqrt(r1);
 1954: 
 1955: ftan		r1 -- r2	float-ext
 1956: r2 = tan(r1);
 1957: :
 1958:  fsincos f/ ;
 1959: 
 1960: fsinh		r1 -- r2	float-ext
 1961: r2 = sinh(r1);
 1962: :
 1963:  fexpm1 fdup fdup 1. d>f f+ f/ f+ f2/ ;
 1964: 
 1965: fcosh		r1 -- r2	float-ext
 1966: r2 = cosh(r1);
 1967: :
 1968:  fexp fdup 1/f f+ f2/ ;
 1969: 
 1970: ftanh		r1 -- r2	float-ext
 1971: r2 = tanh(r1);
 1972: :
 1973:  f2* fexpm1 fdup 2. d>f f+ f/ ;
 1974: 
 1975: fasinh		r1 -- r2	float-ext
 1976: r2 = asinh(r1);
 1977: :
 1978:  fdup fdup f* 1. d>f f+ fsqrt f/ fatanh ;
 1979: 
 1980: facosh		r1 -- r2	float-ext
 1981: r2 = acosh(r1);
 1982: :
 1983:  fdup fdup f* 1. d>f f- fsqrt f+ fln ;
 1984: 
 1985: fatanh		r1 -- r2	float-ext
 1986: r2 = atanh(r1);
 1987: :
 1988:  fdup f0< >r fabs 1. d>f fover f- f/  f2* flnp1 f2/
 1989:  r> IF  fnegate  THEN ;
 1990: 
 1991: sfloats		n1 -- n2	float-ext	s_floats
 1992: ""@i{n2} is the number of address units corresponding to @i{n1}
 1993: single-precision IEEE floating-point numbers.""
 1994: n2 = n1*sizeof(SFloat);
 1995: 
 1996: dfloats		n1 -- n2	float-ext	d_floats
 1997: ""@i{n2} is the number of address units corresponding to @i{n1}
 1998: double-precision IEEE floating-point numbers.""
 1999: n2 = n1*sizeof(DFloat);
 2000: 
 2001: sfaligned	c_addr -- sf_addr	float-ext	s_f_aligned
 2002: "" @i{sf-addr} is the first single-float-aligned address greater
 2003: than or equal to @i{c-addr}.""
 2004: sf_addr = (SFloat *)((((Cell)c_addr)+(sizeof(SFloat)-1))&(-sizeof(SFloat)));
 2005: :
 2006:  [ 1 sfloats 1- ] Literal + [ -1 sfloats ] Literal and ;
 2007: 
 2008: dfaligned	c_addr -- df_addr	float-ext	d_f_aligned
 2009: "" @i{df-addr} is the first double-float-aligned address greater
 2010: than or equal to @i{c-addr}.""
 2011: df_addr = (DFloat *)((((Cell)c_addr)+(sizeof(DFloat)-1))&(-sizeof(DFloat)));
 2012: :
 2013:  [ 1 dfloats 1- ] Literal + [ -1 dfloats ] Literal and ;
 2014: 
 2015: \ The following words access machine/OS/installation-dependent
 2016: \   Gforth internals
 2017: \ !! how about environmental queries DIRECT-THREADED,
 2018: \   INDIRECT-THREADED, TOS-CACHED, FTOS-CACHED, CODEFIELD-DOES */
 2019: 
 2020: \ local variable implementation primitives
 2021: \+
 2022: \+glocals
 2023: 
 2024: @local#		-- w	gforth	fetch_local_number
 2025: w = *(Cell *)(lp+(Cell)NEXT_INST);
 2026: INC_IP(1);
 2027: 
 2028: @local0	-- w	new	fetch_local_zero
 2029: w = *(Cell *)(lp+0*sizeof(Cell));
 2030: 
 2031: @local1	-- w	new	fetch_local_four
 2032: w = *(Cell *)(lp+1*sizeof(Cell));
 2033: 
 2034: @local2	-- w	new	fetch_local_eight
 2035: w = *(Cell *)(lp+2*sizeof(Cell));
 2036: 
 2037: @local3	-- w	new	fetch_local_twelve
 2038: w = *(Cell *)(lp+3*sizeof(Cell));
 2039: 
 2040: \+floating
 2041: 
 2042: f@local#	-- r	gforth	f_fetch_local_number
 2043: r = *(Float *)(lp+(Cell)NEXT_INST);
 2044: INC_IP(1);
 2045: 
 2046: f@local0	-- r	new	f_fetch_local_zero
 2047: r = *(Float *)(lp+0*sizeof(Float));
 2048: 
 2049: f@local1	-- r	new	f_fetch_local_eight
 2050: r = *(Float *)(lp+1*sizeof(Float));
 2051: 
 2052: \+
 2053: 
 2054: laddr#		-- c_addr	gforth	laddr_number
 2055: /* this can also be used to implement lp@ */
 2056: c_addr = (Char *)(lp+(Cell)NEXT_INST);
 2057: INC_IP(1);
 2058: 
 2059: lp+!#	--	gforth	lp_plus_store_number
 2060: ""used with negative immediate values it allocates memory on the
 2061: local stack, a positive immediate argument drops memory from the local
 2062: stack""
 2063: lp += (Cell)NEXT_INST;
 2064: INC_IP(1);
 2065: 
 2066: lp-	--	new	minus_four_lp_plus_store
 2067: lp += -sizeof(Cell);
 2068: 
 2069: lp+	--	new	eight_lp_plus_store
 2070: lp += sizeof(Float);
 2071: 
 2072: lp+2	--	new	sixteen_lp_plus_store
 2073: lp += 2*sizeof(Float);
 2074: 
 2075: lp!	c_addr --	gforth	lp_store
 2076: lp = (Address)c_addr;
 2077: 
 2078: >l	w --	gforth	to_l
 2079: lp -= sizeof(Cell);
 2080: *(Cell *)lp = w;
 2081: 
 2082: \+floating
 2083: 
 2084: f>l	r --	gforth	f_to_l
 2085: lp -= sizeof(Float);
 2086: *(Float *)lp = r;
 2087: 
 2088: fpick	u -- r		gforth
 2089: r = fp[u+1]; /* +1, because update of fp happens before this fragment */
 2090: :
 2091:  floats fp@ + f@ ;
 2092: 
 2093: \+
 2094: \+
 2095: 
 2096: \+OS
 2097: 
 2098: define(`uploop',
 2099:        `pushdef(`$1', `$2')_uploop(`$1', `$2', `$3', `$4', `$5')`'popdef(`$1')')
 2100: define(`_uploop',
 2101:        `ifelse($1, `$3', `$5',
 2102: 	       `$4`'define(`$1', incr($1))_uploop(`$1', `$2', `$3', `$4', `$5')')')
 2103: \ argflist(argnum): Forth argument list
 2104: define(argflist,
 2105:        `ifelse($1, 0, `',
 2106:                `uploop(`_i', 1, $1, `format(`u%d ', _i)', `format(`u%d ', _i)')')')
 2107: \ argdlist(argnum): declare C's arguments
 2108: define(argdlist,
 2109:        `ifelse($1, 0, `',
 2110:                `uploop(`_i', 1, $1, `Cell, ', `Cell')')')
 2111: \ argclist(argnum): pass C's arguments
 2112: define(argclist,
 2113:        `ifelse($1, 0, `',
 2114:                `uploop(`_i', 1, $1, `format(`u%d, ', _i)', `format(`u%d', _i)')')')
 2115: \ icall(argnum)
 2116: define(icall,
 2117: `icall$1	argflist($1)u -- uret	gforth
 2118: uret = (SYSCALL(Cell(*)(argdlist($1)))u)(argclist($1));
 2119: 
 2120: ')
 2121: define(fcall,
 2122: `fcall$1	argflist($1)u -- rret	gforth
 2123: rret = (SYSCALL(Float(*)(argdlist($1)))u)(argclist($1));
 2124: 
 2125: ')
 2126: 
 2127: 
 2128: open-lib	c_addr1 u1 -- u2	gforth	open_lib
 2129: #if defined(HAVE_LIBDL) || defined(HAVE_DLOPEN)
 2130: #ifndef RTLD_GLOBAL
 2131: #define RTLD_GLOBAL 0
 2132: #endif
 2133: u2=(UCell) dlopen(cstr(c_addr1, u1, 1), RTLD_GLOBAL | RTLD_LAZY);
 2134: #else
 2135: #  ifdef _WIN32
 2136: u2 = (Cell) GetModuleHandle(cstr(c_addr1, u1, 1));
 2137: #  else
 2138: #warning Define open-lib!
 2139: u2 = 0;
 2140: #  endif
 2141: #endif
 2142: 
 2143: lib-sym	c_addr1 u1 u2 -- u3	gforth	lib_sym
 2144: #if defined(HAVE_LIBDL) || defined(HAVE_DLOPEN)
 2145: u3 = (UCell) dlsym((void*)u2,cstr(c_addr1, u1, 1));
 2146: #else
 2147: #  ifdef _WIN32
 2148: u3 = (Cell) GetProcAddress((HMODULE)u2, cstr(c_addr1, u1, 1));
 2149: #  else
 2150: #warning Define lib-sym!
 2151: u3 = 0;
 2152: #  endif
 2153: #endif
 2154: 
 2155: uploop(i, 0, 7, `icall(i)')
 2156: icall(20)
 2157: uploop(i, 0, 7, `fcall(i)')
 2158: fcall(20)
 2159: 
 2160: \+
 2161: 
 2162: up!	a_addr --	gforth	up_store
 2163: UP=up=(char *)a_addr;
 2164: :
 2165:  up ! ;
 2166: Variable UP

FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>