Annotation of gforth/stuff.fs, revision 1.41
1.1 anton 1: \ miscelleneous words
2:
1.41 ! anton 3: \ Copyright (C) 1996,1997,1998,2000,2003,2004 Free Software Foundation, Inc.
1.1 anton 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
1.16 anton 19: \ Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111, USA.
1.1 anton 20:
1.17 anton 21: require glocals.fs
1.8 anton 22:
1.6 anton 23: ' require alias needs ( ... "name" -- ... ) \ gforth
1.11 crook 24: \G An alias for @code{require}; exists on other systems (e.g., Win32Forth).
1.1 anton 25: \ needs is an F-PC name. we will probably switch to 'needs' in the future
26:
27: \ a little more compiler security
28:
29: \ currently not used by Gforth, but maybe by add-ons e.g., the 486asm
30: AUser CSP
31:
32: : !CSP ( -- )
33: sp@ csp ! ;
34:
35: : ?CSP ( -- )
36: sp@ csp @ <> -22 and throw ;
1.2 anton 37:
1.4 anton 38: \ DMIN and DMAX
39:
1.13 crook 40: : dmin ( d1 d2 -- d ) \ double d-min
1.5 pazsan 41: 2over 2over d> IF 2swap THEN 2drop ;
1.2 anton 42:
1.13 crook 43:
44: : dmax ( d1 d2 -- d ) \ double d-max
1.5 pazsan 45: 2over 2over d< IF 2swap THEN 2drop ;
1.4 anton 46:
47: \ shell commands
48:
1.10 crook 49: 0 Value $? ( -- n ) \ gforth dollar-question
1.12 crook 50: \G @code{Value} -- the exit status returned by the most recently executed
1.9 crook 51: \G @code{system} command.
1.4 anton 52:
1.14 anton 53: : system ( c-addr u -- ) \ gforth
1.36 anton 54: \G Pass the string specified by @var{c-addr u} to the host operating
55: \G system for execution in a sub-shell. The value of the environment
56: \G variable @code{GFORTHSYSTEMPREFIX} (or its default value) is
57: \G prepended to the string (mainly to support using @code{command.com}
58: \G as shell in Windows instead of whatever shell Cygwin uses by
59: \G default; @pxref{Environment variables}).
1.4 anton 60: (system) throw TO $? ;
1.9 crook 61:
1.4 anton 62: : sh ( "..." -- ) \ gforth
1.9 crook 63: \G Parse a string and use @code{system} to pass it to the host
64: \G operating system for execution in a sub-shell.
1.4 anton 65: '# parse cr system ;
66:
1.8 anton 67: \ stuff
68:
1.10 crook 69: : ]L ( compilation: n -- ; run-time: -- n ) \ gforth
70: \G equivalent to @code{] literal}
1.8 anton 71: ] postpone literal ;
72:
1.23 anton 73: [ifundef] in-dictionary?
1.8 anton 74: : in-dictionary? ( x -- f )
75: forthstart dictionary-end within ;
1.23 anton 76: [endif]
1.8 anton 77:
78: : in-return-stack? ( addr -- f )
79: rp0 @ swap - [ forthstart 6 cells + ]L @ u< ;
1.17 anton 80:
81: \ const-does>
82:
83: : compile-literals ( w*u u -- ; run-time: -- w*u ) recursive
1.18 anton 84: \ compile u literals, starting with the bottommost one
1.17 anton 85: ?dup-if
86: swap >r 1- compile-literals
87: r> POSTPONE literal
88: endif ;
89:
90: : compile-fliterals ( r*u u -- ; run-time: -- w*u ) recursive
1.18 anton 91: \ compile u fliterals, starting with the bottommost one
1.17 anton 92: ?dup-if
93: { F: r } 1- compile-fliterals
94: r POSTPONE fliteral
95: endif ;
96:
97: : (const-does>) ( w*uw r*ur uw ur target "name" -- )
1.18 anton 98: \ define a colon definition "name" containing w*uw r*ur as
99: \ literals and a call to target.
1.17 anton 100: { uw ur target }
101: header docol: cfa, \ start colon def without stack junk
102: ur compile-fliterals uw compile-literals
103: target compile, POSTPONE exit reveal ;
104:
1.35 anton 105: : const-does> ( run-time: w*uw r*ur uw ur "name" -- ) \ gforth
106: \G Defines @var{name} and returns.
107: \G
1.18 anton 108: \G @var{name} execution: pushes @var{w*uw r*ur}, then performs the
109: \G code following the @code{const-does>}.
1.17 anton 110: here >r 0 POSTPONE literal
111: POSTPONE (const-does>)
112: POSTPONE ;
113: noname : POSTPONE rdrop
1.32 anton 114: latestxt r> cell+ ! \ patch the literal
1.17 anton 115: ; immediate
1.19 anton 116:
1.20 anton 117: \ !! rewrite slurp-file using slurp-fid
1.34 anton 118: : slurp-file ( c-addr1 u1 -- c-addr2 u2 ) \ gforth
1.20 anton 119: \G @var{c-addr1 u1} is the filename, @var{c-addr2 u2} is the file's contents
1.19 anton 120: r/o bin open-file throw >r
121: r@ file-size throw abort" file too large"
122: dup allocate throw swap
123: 2dup r@ read-file throw over <> abort" could not read whole file"
124: r> close-file throw ;
125:
1.34 anton 126: : slurp-fid { fid -- addr u } \ gforth
1.20 anton 127: \G @var{addr u} is the content of the file @var{fid}
128: 0 0 begin ( awhole uwhole )
129: dup 1024 + dup >r extend-mem ( anew awhole uwhole R: unew )
130: rot r@ fid read-file throw ( awhole uwhole uread R: unew )
131: r> 2dup =
132: while ( awhole uwhole uread unew )
133: 2drop
134: repeat
135: - + dup >r resize throw r> ;
1.24 anton 136:
1.25 anton 137: \ ]] ... [[
138:
139: : compile-literal ( n -- )
140: postpone literal ;
141:
142: : [[ ( -- )
143: \G switch from postpone state to compile state
144: \ this is only a marker; it is never really interpreted
145: compile-only-error ; immediate
146:
147: : postponer ( c-addr u -- )
148: 2dup find-name dup if ( c-addr u nt )
149: nip nip name>comp
150: 2dup [comp'] [[ d= if
151: 2drop ['] compiler is parser
152: else
153: postpone,
154: endif
155: else
156: drop
157: 2dup snumber? dup if
158: 0> IF
159: swap postpone literal postpone compile-literal
160: THEN
161: postpone Literal postpone compile-literal
162: 2drop
163: ELSE
164: drop no.extensions
165: THEN
166: then ;
167:
168: : ]] ( -- )
169: \ switch into postpone state
170: ['] postponer is parser state on ; immediate restrict
171:
1.26 anton 172: \ f.rdp
173:
1.27 anton 174: : push-right ( c-addr u1 u2 cfill -- )
1.26 anton 175: \ move string at c-addr u1 right by u2 chars (without exceeding
176: \ the original bound); fill the gap with cfill
1.27 anton 177: >r over min dup >r rot dup >r ( u1 u2 c-addr R: cfill u2 c-addr )
1.26 anton 178: dup 2swap /string cmove>
179: r> r> r> fill ;
180:
1.27 anton 181: : f>buf-rdp-try { f: rf c-addr ur nd up um1 -- um2 }
182: \ um1 is the mantissa length to try, um2 is the actual mantissa length
183: c-addr ur um1 /string '0 fill
184: rf c-addr um1 represent if { nexp fsign }
1.26 anton 185: nd nexp + up >=
186: ur nd - 1- dup { beforep } fsign + nexp 0 max >= and if
187: \ fixed-point notation
1.27 anton 188: c-addr ur beforep nexp - dup { befored } '0 push-right
1.26 anton 189: c-addr beforep 1- befored min dup { beforez } 0 max bl fill
190: fsign if
191: '- c-addr beforez 1- 0 max + c!
192: endif
1.27 anton 193: c-addr ur beforep /string 1 '. push-right
194: nexp nd +
1.26 anton 195: else \ exponential notation
1.27 anton 196: c-addr ur 1 /string 1 '. push-right
1.26 anton 197: fsign if
1.27 anton 198: c-addr ur 1 '- push-right
1.26 anton 199: endif
200: nexp 1- s>d tuck dabs <<# #s rot sign 'E hold #> { explen }
1.27 anton 201: ur explen - 1- fsign + { mantlen }
202: mantlen 0< if \ exponent too large
1.26 anton 203: drop c-addr ur '* fill
204: else
205: c-addr ur + 0 explen negate /string move
206: endif
1.27 anton 207: #>> mantlen
1.26 anton 208: endif
209: else \ inf or nan
210: if \ negative
1.27 anton 211: c-addr ur 1 '- push-right
1.26 anton 212: endif
1.27 anton 213: drop ur
1.26 anton 214: \ !! align in some way?
1.27 anton 215: endif
216: 1 max ur min ;
217:
1.29 anton 218: : f>buf-rdp ( rf c-addr +nr +nd +np -- ) \ gforth
1.28 anton 219: \G Convert @i{rf} into a string at @i{c-addr nr}. The conversion
220: \G rules and the meanings of @i{nr nd np} are the same as for
1.27 anton 221: \G @code{f.rdp}.
222: \ first, get the mantissa length, then convert for real. The
223: \ mantissa length is wrong in a few cases because of different
224: \ rounding; In most cases this does not matter, because the
225: \ mantissa is shorter than expected and the final digits are 0;
226: \ but in a few cases the mantissa gets longer. Then it is
227: \ conceivable that you will see a result that is rounded too much.
228: \ However, I have not been able to construct an example where this
229: \ leads to an unexpected result.
230: swap 0 max swap 0 max
231: fdup 2over 2over 2 pick f>buf-rdp-try f>buf-rdp-try drop ;
1.26 anton 232:
1.28 anton 233: : f>str-rdp ( rf +nr +nd +np -- c-addr nr ) \ gforth
234: \G Convert @i{rf} into a string at @i{c-addr nr}. The conversion
235: \G rules and the meanings of @i{nr +nd np} are the same as for
1.26 anton 236: \G @code{f.rdp}. The result in in the pictured numeric output buffer
237: \G and will be destroyed by anything destroying that buffer.
1.28 anton 238: rot holdptr @ 1- 0 rot negate /string ( rf +nd np c-addr nr )
1.26 anton 239: over holdbuf u< -&17 and throw
240: 2tuck 2>r f>buf-rdp 2r> ;
241:
1.28 anton 242: : f.rdp ( rf +nr +nd +np -- ) \ gforth
1.26 anton 243: \G Print float @i{rf} formatted. The total width of the output is
1.30 anton 244: \G @i{nr}. For fixed-point notation, the number of digits after the
245: \G decimal point is @i{+nd} and the minimum number of significant
246: \G digits is @i{np}. @code{Set-precision} has no effect on
247: \G @code{f.rdp}. Fixed-point notation is used if the number of
248: \G siginicant digits would be at least @i{np} and if the number of
249: \G digits before the decimal point would fit. If fixed-point notation
250: \G is not used, exponential notation is used, and if that does not
251: \G fit, asterisks are printed. We recommend using @i{nr}>=7 to avoid
252: \G the risk of numbers not fitting at all. We recommend
253: \G @i{nr}>=@i{np}+5 to avoid cases where @code{f.rdp} switches to
254: \G exponential notation because fixed-point notation would have too
255: \G few significant digits, yet exponential notation offers fewer
256: \G significant digits. We recommend @i{nr}>=@i{nd}+2, if you want to
257: \G have fixed-point notation for some numbers. We recommend
258: \G @i{np}>@i{nr}, if you want to have exponential notation for all
259: \G numbers.
1.26 anton 260: f>str-rdp type ;
261:
262: 0 [if]
263: : testx ( rf ur nd up -- )
264: '| emit f.rdp ;
265:
266: : test ( -- )
267: -0.123456789123456789e-20
268: 40 0 ?do
269: cr
270: fdup 7 3 1 testx
271: fdup 7 3 4 testx
272: fdup 7 3 0 testx
273: fdup 7 7 1 testx
274: fdup 7 5 1 testx
275: fdup 7 0 2 testx
276: fdup 5 2 1 testx
277: fdup 4 2 1 testx
278: fdup 18 8 5 testx
279: '| emit
280: 10e f*
281: loop ;
282: [then]
1.33 anton 283:
284: : f.s ( -- ) \ gforth f-dot-s
285: \G Display the number of items on the floating-point stack,
286: \G followed by a list of the items; TOS is the right-most item.
287: ." <" fdepth 0 .r ." > " fdepth 0 max maxdepth-.s @ min dup 0
288: ?DO dup i - 1- floats fp@ + f@ 16 5 11 f.rdp space LOOP drop ;
1.37 anton 289:
290: \ defer stuff
291:
1.39 pazsan 292: [ifundef] defer@ : defer@ >body @ ; [then]
293:
1.37 anton 294: :noname ' defer@ ;
295: :noname postpone ['] postpone defer@ ;
296: interpret/compile: action-of ( interpretation "name" -- xt; compilation "name" -- ; run-time -- xt ) \ gforth
297: \G @i{Xt} is the XT that is currently assigned to @i{name}.
298:
299: ' action-of
300: comp' action-of drop
301: interpret/compile: what's ( interpretation "name" -- xt; compilation "name" -- ; run-time -- xt ) \ gforth-obsolete
1.38 anton 302: \G Old name of @code{action-of}
1.40 anton 303:
304:
305: : typewhite ( addr n -- ) \ gforth
306: \G Like type, but white space is printed instead of the characters.
307: \ bounds u+do
308: 0 max bounds ?do
309: i c@ #tab = if \ check for tab
310: #tab
311: else
312: bl
313: then
314: emit
315: loop ;
316:
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>